Binding of Halothane to Serum Albumin Demonstrated Using Tryptophan Fluorescence

Halothane (2-bromo-2-chloro-1,1,1-trifluoroethane) was obtained from Halocarbon Laboratories (Hackensack, NJ). The thymol preservative was removed before use with an aluminum oxide column, which also removes any water that may be present. [7]Diethyl ether (anhydrous) was supplied by Aldrich Chemical Co. (Milwaukee, WI). Fatty acid-free BSA (lot 119F9306), fatty acid-free human serum albumin (HSA) (lot 42H9313), apomyoglobin (lot 113H8255, from horse skeletal muscle), and L-tryptophan were purchased from Sigma Chemical Co. (St. Louis, MO). All other chemicals used were reagent grade.

The optical isomers of halothane were generously provided by Dr. J. Meinwald of Cornell University (Ithaca, NY). The S-isomer has a reported purity of 97.22% and the R-isomer a purity of 99.90%. The chemical purity is 100%.

The buffer used for all fluorescence experimentation was 130 mM sodium chloride, 20 mM sodium phosphate, pH 7.0. Proteins (5 micro Meter) were equilibrated with halothane (final concentration 12 mM) in gas-tight Hamilton syringes. This concentration is less than the reported water solubility of halothane at 25 degrees Celsius by¹⁹F NMR spectroscopy. [8]The halothane-equilibrated BSA was then diluted with predetermined volumes of plain BSA (treated in the same manner) to achieve the final halothane concentrations (shown in the figures).

BSA and HSA were used without further purification (purity > 96% by electrophoresis). There are two tryptophans in BSA, at positions 134 and 212. The tryptophan at position 134 is thought to be more exposed to the aqueous solvent than is tryptophan 212. [9]HSA has a single tryptophan, at position 214. Equine apomyoglobin has two tryptophans, at positions 7 and 14. [10]Protein concentration after dissolution was determined with a UV/Vis Spectrophotometer Lambda 2 (Perkin-Elmer, Norwalk, CT), taking an extinction coefficient at 279 nm for BSA = 4,500 M sup -1 *symbol* cm sup -1 and an extinction coefficient at 279 nm for HSA = 3,600 M sup -1 *symbol* cm sup -1. [9]The apomyoglobin concentration was determined using an extinction coefficient at 280 nm = 1,600 M sup -1 *symbol* cm sup -1. [10].

All fluorescence measurements were performed with a Fluorescence Spectrophotometer F-4500 (Hitachi, Danbury, CT). Tryptophan was excited at a wavelength of 295 nm and emission spectra recorded with peaks at 344 nm for BSA at pH 7.0. This excitation wavelength was selected to minimize radiationless energy transfer from the tyrosine residues in BSA and HSA, which can contribute to tryptophan emission when an excitation wavelength of 280 nm is used. [9]A 10-mm-pathlength quartz cell with a polytetrafluorethylene stopper was used. Care was taken to avoid the presence of significant air pockets. The cell was placed in a thermostatically controlled cell holder with a temperature of 25.0 plus/minus 0.1 degree Celsius. Both excitation and emission slit widths were 5 nm. Fluorescence (F) measurements were corrected for inner filter effects [11]according to the relation F^corrected= F^observed*symbol* antilog (OD^ex- OD^em/2). Halothane was determined to have an extinction coefficient at 199 nm of 420 plus/minus 10 M sup -1 *symbol* cm sup -1 (data not shown), so inner filter effects were found to be negligible.

For the competition binding studies, fluorescence spectra of a portion of a sample of BSA (5 micro Meter) equilibrated with halothane were first measured. The remaining protein-halothane mixture in the gas-tight syringe was then equilibrated with varying concentrations of diethyl ether, and additional spectra were obtained.

Circular dichroism spectra were generated with an DS Spectrometer (Model 62, Aviv, Lakewood, NJ). BSA (5 micro Meter) in 10 mM potassium phosphate buffer at pH 7.0 was equilibrated with halothane or diethyl ether in gas-tight Hamilton syringes. A 1-mm-pathlength quartz cell with a polytetrafluorethylene stopper was used. The cell holder was temperature-controlled at 25.0 plus/minus 0.1 degree Celsius. The bandwidth was 1.00 nm, with a scan step of 0.5 nm and an average scan time of 1.0 s. The circular dichroism spectra were stored on diskette and later analyzed with the KaleidaGraph program (Abelbeck Software, Reading, PA, 1991).

Best-fit curves were generated with the Peakfit (Jandel Scientific, San Rafael, CA, 1990) and KaleidaGraph programs. Data are expressed as means plus/minus SD. Data points are the averages of at least three experiments conducted with separate samples.

(Figure 1(curve A)) shows that coequilibration of halothane with BSA (5 micro Meter) at pH 7.0 causes a concentration-dependent decrease in the intrinsic tryptophan fluorescence. No shift in the emission wavelength maximum (344 nm) was observed. Taking F⁰as the fluorescence in the absence of added halothane and F as the fluorescence in its presence, then; Equation 1where Q = the quenched fluorescence. The quenched fluorescence is a function of the maximum fluorescence that can be quenched (Q^max) and the affinity of halothane (dissociation constant, K^d) for its binding site in the vicinity of tryptophan. If, to a first approximation, the binding sites are treated as equal and no fluorescence polarization changes occur, then from mass law considerations it follows [12]that Equation 2. The line through the data points is a best-fit curve derived from Equation 2. The K^d= 1.9 plus/minus 0.2 mM, and Q^max= 0.99 plus/minus 0.04.

To evaluate the importance of the native BSA conformation to halothane binding, experiments were performed at pH 3.0. At this pH, BSA changes from an ellipsoid (normal) to a fully uncoiled (expanded) shape. [9]With excitation at 295 nm, peak fluorescence shifted to 330 nm, with a concomitant decrease in quantum yield, indicating the change in environment of the tryptophan residues under these conditions. Figure 1(curve B) shows that there is a large decrease in the amount of quenching of BSA tryptophan fluorescence compared with that at pH 7.0 (curve A) and that there is apparent loss of saturable binding over the concentration range studied.

Also shown in Figure 1(curve D) is the effect of halothane on apomyoglobin (5 micro Meter) tryptophan fluorescence at pH 7.0. Apomyoglobin has two tryptophan residues, [10]at positions 7 and 14, one of which is buried and the other of which is exposed to the solvent. [13]Addition of halothane causes a linear decrease in the intrinsic apomyoglobin tryptophan fluorescence with lack of saturation over the concentration range studied (to 12 mM halothane). At 12 mM halothane only 12 plus/minus 1% (n = 3) of the total tryptophan fluorescence of apomyoglobin was quenched.

The bromine atom in the halothane is the most likely cause of the quenching reaction, although the chlorine atom may also contribute. [11]As an additional control for the possibility that the quenching is caused by free halothane in solution, the effect of sodium bromide on BSA (5 micro Meter) fluorescence at pH 7.0 was measured. Bromide ion (12 mM) caused a 9.4 plus/minus 0.5% (n = 3) decrease in BSA tryptophan fluorescence, and 200 mM sodium bromide decreased BSA tryptophan fluorescence by only 18.2 plus/minus 2.5% (n = 3). These results indicate only minor effects of free solvated halothane on the observed quenching of BSA fluorescence.

In the converse, control experiment, the effect of halothane (to 12 mM) on free L-tryptophan (10 micro Meter) fluorescence at pH 7.0 was assessed (Figure 1[curve C]). A linear decrease in tryptophan fluorescence was observed, with a maximum quenching of 22 plus/minus 2% (n = 5) at 12 mM halothane.

A Hill plot of halothane binding to BSA at pH 7.0 is shown in Figure 2. A K^d= 1.8 plus/minus 0.1 mM and a Hill number of 1.0 plus/minus 0.1 can be calculated by this approach. Thus, the Hill number shows that there is no cooperativity to the halothane-BSA interaction.

The optical isomers of isoflurane have been reported to have different potencies in activating potassium channels [3]and in binding to BSA. [5]The ability of the two stereoisomers of halothane to quench BSA tryptophan fluorescence was characterized with respect to K^dand Q^max. No significant differences in the ability of the two optical isomers of halothane to bind to BSA was measured by this approach. The calculated dissociation constants were 2.2 plus/minus 0.1 mM for the S-isomer and 2.1 plus/minus 0.2 mM for the R-isomer.

The equilibrium binding free energy (molar standard state) for halothane from the aqueous phase to the BSA interior (Delta G^oB,w) can be calculated as Delta G^oB,w = RTlnK^d, where R is the gas constant and T is the absolute temperature. At 25 degrees Celsius C, Delta G^oB,w = -3.7 kcal/mol. A portion of the free energy of binding is contributed by the simple partitioning of halothane into the protein interior as a result of the hydrophobic effect (Delta G^oB,h). Warncke and Dutton [14]developed an empirical expression for the interaction between quinone and the photosynthetic reaction center protein: Equation 3where Delta G^oB,h and Delta G^oB,w = the standard Gibbs free energies for ligand binding to protein from hydrophobic and aqueous phases, respectively; Delta G^otr= the standard free energy for the transfer of ligand from an aqueous to a hydrophobic phase; and alpha = an empirical constant (0.78), giving the ratio of ligand activities in the hydrophobic and aqueous phases. According to published data for the partitioning of halothane between water and phospholipid phases, [15]Delta G^otr= -2.8 kcal/mol at 25 degrees Celsius, and Delta G^oB,h can be calculated as -1.5 kcal/mol.

(Figure 1(curve A)) shows that halothane is able to quench the fluorescence of both of the tryptophan residues in BSA, indicating the presence of at least two halothane binding sites. To measure the relative affinities of the two halothane binding sites identified by tryptophan fluorescence quenching in BSA, experiments were carried out with HSA, which contains only a single tryptophan, at the conserved position 214. [9]The tryptophan at position 214 in HSA is analogous to the tryptophan at position 212 in BSA and is more buried compared with that at position 134 in BSA. [9] Figure 3shows the concentration-dependent quenching of HSA and BSA tryptophan fluorescence with the addition of halothane. By Equation 2, the K^d= 3.9 plus/minus 0.5 mM and the Q sub max = 1.03 plus/minus 0.04 for HSA (compared with the K^d= 1.8 plus/minus 0.2 mM for BSA).

Assuming that the halothane binding site in the vicinity of tryptophan 212 in BSA is similar to that in HSA, one can then calculate the affinity of the more exposed tryptophan 134 site for halothane in BSA. The K^dof the tryptophan 212 site can be taken as 3.9 mM and the data in Figure 1(curve A) fitted to an equation of the form Equation 4where halothane concentrations are expressed in millimolar units; 0.63 and 0.37 = the experimentally determined contributions [9]of each tryptophan residue to the total fluorescence at 344 nm; and K^d= the dissociation constant of the tryptophan 134 site. This equation allows a value of 1.3 plus/minus 0.1 mM to be calculated for the K^dof the more exposed site. Almost identical chi-squared (0.00987 vs. 0.00998 for a two vs. a one affinity model, respectively) and correlation coefficients (both 0.995) are obtained with both solutions.

Ultraviolet irradiation of halothane can result in the production of a bromide ion and a 1-chloro-2,2,2-trifluoroethyl free radical. [16,17]Indeed, this forms the basis of the halothane photoaffinity labeling technique developed by Eckenhoff and Shuman. [17]To test whether the fluorometric experiment may be causing irreversible binding of halothane to BSA, samples of BSA (5 micro Meter) with halothane (12 mM) were examined fluorometrically and degassed with nitrogen in the same cuvette for 30 min, and a second tryptophan emission spectrum was obtained on the same specimen. Compared with a control sample of BSA treated in the same manner, the BSA exposed to halothane and to ultraviolet light and subsequently degassed retained 95 plus/minus 9% (n = 3) of its tryptophan fluorescence. These results indicate that the observed fluorescence quenching is reversible and that the spectroscopic technique is not leading to covalent modification of the BSA. The reversibility also means that the observed quenching of tryptophan fluorescence is not caused by halothane-induced aggregation or precipitation of BSA.

As shown in Figure 1(curve C), halothane had only a small quenching effect on free L-tryptophan (10 micro Meter) fluorescence in buffer at pH 7.0. Experiments were performed in methanol to examine the effects of higher concentrations of halothane (to 170 mM) on tryptophan fluorescence. This higher concentration range was chosen to approximate the known partitioning of volatile anesthetics into hydrophobic regions of membranes. [18]At these very high concentrations halothane quenched free L-tryptophan fluorescence to the same degree as it did BSA tryptophan fluorescence. Figure 4(curve A) presents the data in the form of a Stern-Volmer plot. The data points were fit using the classic Stern-Volmer equation, Equation 5where K^sv= the collisional quenching constant. K^svis equal to the product of the bimolecular quenching constant and the fluorescence lifetime of the tryptophan in the absence of quencher. A K^sv= 16.4 plus/minus 0.5 M sup -1 was calculated from the slope of the plot.

A linear Stern-Volmer plot can be obtained with either collisional or static quenching. Collisional quenching is a diffusion-controlled process that occurs when a quencher molecule is within a certain minimal distance from the fluorophore at the time of excitation. In contrast, static quenching results from the formation of a complex between the fluorophore and quencher. To differentiate between these quenching mechanisms, the viscosity of the solvent was increased by the addition of glycerol. Figure 4(curve B) shows the Stern-Volmer plot of halothane quenching of tryptophan in 1:1 (volume in volume) glycerol:methanol. A K^sv= 7.6 plus/minus 0.2 M sup -1 was now determined, indicating the collisional nature of the quenching reaction.

In addition to the proximity of bound halothane to tryptophan residues, fluorescence quenching could result from structural modifications of BSA upon halothane binding. Far-ultraviolet circular dichroism spectroscopy was performed to examine the secondary structure of BSA in the presence of as much as 12 mM halothane. As shown in Figure 5, BSA 5 micro Meter in the absence of halothane exhibited negative absorption bands with maxima at 222 and 208 nm and a positive band with a maximum at 192 nm. Figure 5shows that no significant effect of halothane on the circular dichroism spectrum of BSA was detected. The magnitude of the molar ellipticity for the three absorption bands compares well with previous work on BSA. [19].

As an additional test of whether quenching of tryptophan fluorescence by halothane results primarily from the chemical structure of halothane, rather than changes in the structure of the protein, and therefore whether quenching indicates the anesthetic's location, competition studies were performed with a non-quenching anesthetic. Diethyl ether alone had no effect on BSA tryptophan fluorescence at concentrations to 50 mM (data not shown). In addition, Figure 4(curve C) shows that diethyl ether had no effect on the fluorescence of free L-tryptophan in methanol.

In contrast, if BSA is incubated with halothane and diethyl ether is added, the fluorescence increases, suggesting that halothane can be displaced from its sites on the protein. The concentration-dependent inhibition of halothane binding by diethyl ether, as assessed by fluorescence quenching, is shown in Figure 6. The calculated diethyl ether concentration required for 50% inhibition of halothane binding is 39 plus/minus 7 mM. To test whether diethyl ether causes changes in the secondary structure of BSA at these concentrations that may explain the halothane displacement, circular dichroism spectroscopy was performed on BSA in the presence of diethyl ether (37.5 mM). The spectra for BSA in the presence and absence of diethyl ether are indistinguishable from each other and from those shown in Figure 5.

Using the quenching of native tryptophan fluorescence, we have shown that halothane binds saturably and reversibly to BSA. This approach has not been used previously to study interactions between soluble proteins and volatile anesthetics. The main advantage of this technique is that it provides information about the location of at least some of the site(s) in the protein to which the volatile anesthetic binds. The technique may lend itself to an improved understanding of where volatile anesthetics interact with membrane proteins. In this regard, Augustin and Hasselbach [20]reported that halothane quenched the intrinsic fluorescence of rabbit sarcoplasmic reticulum vesicles.

The calculated K^dfor the halothane-BSA interaction from our study (1.8 mM) compares well with a value of 1.3 mM reported for halothane binding to BSA using¹⁹F NMR spectroscopy. [6]Eckenhoff and Shuman, [5]using photoaffinity labeling, reported a slightly lower K^d, 0.3-0.5 mM, for halothane binding to BSA. These quantitative differences may be due to the inability of the fluorescence approach to examine more than two of the possible four sites for halothane binding in BSA. We are not able to determine the total number of binding sites for halothane per BSA molecule using the fluorescence quenching approach. The finding that halothane quenches the fluorescence of both tryptophan residues suggests that halothane is binding to at least two sites. Halothane binding to other sites in BSA could not be detected using this approach. The finding that the two stereoisomers of halothane interact with BSA with the same affinity indicates that the domains are not structured to confer any discriminating specificity on the binding.

Saturable binding of halothane to BSA was found to be abolished by lowering the pH to 2.5 in the¹⁹F NMR spectroscopy study. [6]Similarly, Eckenhoff and Shuman [5]found that at pH 3.0 photoaffinity labeling decreased by approximately 75% (at a halothane concentration of 3 mM). Our finding that halothane quenching of tryptophan fluorescence is attenuated by approximately 74%, at a halothane concentration of 3 mM (Figure 1), is in good agreement with these studies. Taken together, these data indicate that the uncoiling of BSA alters the environments of the tryptophan residues in such a way that they are either less accessible to halothane or unable to bind halothane to the same degree as the native conformation.

The experiments with apomyoglobin further demonstrate that halothane binding to proteins is conformation dependent. Apomyoglobin also contains two tryptophans. [10]However, the environment of these two indole rings apparently does not favor the binding of halothane with sufficient proximity or concentration to quench tryptophan fluorescence.

Because bromide ions caused only a small decrease in BSA tryptophan fluorescence, the high degree of quenching by halothane of BSA fluorescence is attributed to partitioning of the anesthetic into hydrophobic domains in the vicinity of the indole rings. A similar mechanism has been proposed to explain trichloroethanol quenching of protein tryptophan fluorescence. [21,22]The quenching of free L-tryptophan in methanol shown in Figure 4demonstrates that at sufficiently high concentrations, the degree of fluorescence quenching observed in curve a of Figure 1for BSA can be attained. This apparent elevated halothane concentration in the protein is attributed to the favorable energetics of the anesthetic-protein interaction and to the steric hindrance to dissociation from the tryptophan residues imposed by the surrounding polypeptide.

Binding of halothane to HSA occurred with lower affinity (K^d= 3.9 mM) than the average affinity of the two sites in BSA (K^d= 1.8 mM). Approximating the single tryptophan in HSA to the more buried residue in BSA, the exposed tryptophan site in BSA can be calculated to have a K^d= 1.3 mM. This suggests that halothane binds more easily to water accessible portions of BSA than deeply buried hydrophobic segments in the protein interior, consistent with the proposal of Franks and Lieb that anesthetics bind to hydrophobic pockets in proteins exposed to water. [3].

However, we predict using Equation 3that the free energy of binding of halothane to BSA from a hydrophobic solvent remains favorable. The conclusion is that halothane would still bind to BSA if the protein were dissolved in a hydrophobic phase (such as a lipid membrane), but with a lower affinity (K^d= 80 mM). It is of interest that this K^dapproximates the concentration of halothane that is attained in the lipid portion of membranes during clinical anesthesia. [18].

The precise mechanism of quenching by halothane is not clear. This is important because it may indicate the degree of proximity of halothane to the tryptophan residue. Halothane probably quenches tryptophan fluorescence by spin-orbital coupling of the excited (singlet) indole ring and the 2-bromo atom leading to intersystem crossing to an excited triplet state. [11]The 2-chloro atom may also contribute to the observed quenching of tryptophan fluorescence. To test whether this is indeed the mechanism of fluorescence quenching, the production of triplets may be detected using either electron paramagnetic resonance or phosphorescence spectroscopy. Berlman [23]reported that the electronic orbitals of chromophore and quencher need to overlap for heavy atom collisional quenching to occur. However, bromine atoms have been reported to quench fluorophores in other systems [24,25]at distances of 3-5 Angstrom. This would decrease the molecular resolution of the fluorescence quenching approach, but would still allow the conclusion that halothane binds to BSA within two spheres of radii 3-5 Angstrom centered on each indole ring. For comparison, the BSA molecule is an ellipsoid with the overall dimensions of 42 x 141 Angstrom. [9].

It should be noted that anesthetic-induced structural changes in BSA may cause fluorescence quenching by mechanisms that do not involve direct tryptophan-halothane proximity. A variety of chemical groups present in proteins (including histidine, cysteine, proline and arginine) are capable of quenching tryptophan fluorescence, [26]if structural changes alter their proximity to the indole ring. However, halothane (to 12 mM) has no effect on the far-ultraviolet circular dichroism spectrum of BSA, indicating a lack of major secondary structural changes. This result suggests that halothane interacts with BSA in a manner that does not involve extensive hydrogen-bond disruption. It is, however, not possible to exclude the contribution of small structural changes in the protein to the observed fluorescence quenching.

The experiments showing that diethyl ether can in part displace halothane from its binding sites adds support to the interpretation that the heavy atoms on the halothane are responsible for the observed quenching, rather than protein structural changes. Diethyl ether is able to compete with halothane for its binding sites with a 50% inhibition concentration of 39 plus/minus 7 mM. This is comparable to the value of 46 mM reported previously for diethyl ether inhibition of halothane photoaffinity labeling of BSA. [5]This appears to be a direct competitive interaction, because no secondary structural changes are observed at the reported anesthetic concentrations.

In summary, using a simple and readily available approach, we have demonstrated the saturable and reversible binding of halothane to the water-soluble protein BSA. Our results are in good agreement with published studies using alternative methods. The results suggest that both tryptophan 134 and 212 are part of at least a portion of the halothane binding domains in BSA and therefore represent the highest resolution to which such binding sites for halothane in BSA have been described. Advantages of this method are that it is more sensitive and requires less protein than the¹⁹F NMR and gas chromatography approach [6]and that it is nondestructive compared with the photoaffinity labeling technique. [5]The method presented here should allow kinetic and thermodynamic analysis of the interactions between volatile anesthetics and proteins, furthering our understanding of the types of molecular interactions involved. The main limitation of the technique is that it fails to detect anesthetic binding to proteins at sites not containing aromatic residues. The fluorometric quenching approach may prove useful for studying anesthetic binding to membrane proteins and water-soluble regulatory proteins, which are presumably more relevant candidates as potential sites of anesthetic action.

The authors thank Dr. Franz M. Matschinsky, Department of Biochemistry and Biophysics, for the use of the spectrofluorometer.