Introduction

Top Abstract Introduction Materials & Methods Results & Discussion References

The coadministration of cocaine (benzoylmethylecgonine) and ethanol produces a unique metabolite, cocaethylene (benzoylethylecgonine, ethylcocaine), which possesses pharmacological activity nearly identical to that of cocaine, a longer elimination half-life, and greater toxicity than cocaine (Hearn et al., 1991; Woodward et al., 1991; Katz et al., 1992). This metabolite is formed via a transesterification reaction that is catalyzed by an hepatic carboxylesterase enzyme (Dean et al., 1991; Brzenzinski et al., 1994; Boyer and Petersen, 1992). Recently, a microsomal carboxylesterase enzyme responsible for cocaine hydrolysis has been implicated in this transesterification reaction (Brzenzinski et al., 1994; Boyer and Petersen, 1992). We have hypothesized that cocaethylene, the transesterification product formed from the interaction of cocaine and ethanol, itself participates with ethanol in this interaction. The transesterification of cocaethylene would involve the exchange of a carboxyl ethyl group for the ethyl group of ethanol and, therefore, no change in molecular structure. To monitor this exchange, deuterated ethanol (²H₆-ethanol) was incubated with cocaethylene and the resultant deuterated product (²H₅-cocaethylene) was assayed by GC/MS. To illustrate kinetic changes brought about by the regeneration of parent compound via transesterification, drug was incubated with liver homogenates with and without unlabeled ethanol. A comparison of the ethanolic transesterification of cocaine and cocaethylene was made by determining K_M, V_max, and intrinsic metabolite formation clearance values for both substrates. Further characterization of the enzymatic interaction between cocaethylene and ²H₆-ethanol was accomplished by using specific and nonspecific esterase inhibitors.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results & Discussion References

Chemicals. Cocaine HCl and cocaethylene fumarate were provided by Research Triangle Institute (Chapel Hill, NC) through the generosity of the National Institute on Drug Abuse. Ethanol (absolute ethyl alcohol) was purchased from Quantum Chemical Corporation, USI Division (Tuscola, IL). ²H₆-Ethanol (ethyl-d₅-alcohol-d), physostigmine (eserine), BNPP,² and 4-methylumbelliferyl acetate were purchased from Sigma Chemical Co. (St. Louis, MO).

Rat Liver S9 Fraction Preparation. Male rats (250-275 g; Harlan Sprague Dawley, Indianapolis, IN) were anesthetized with ether, and the entire liver was removed and washed in cold KCl solution (1.15%). The livers were blotted dry, weighed, and mixed with 70 mM KH₂PO₄ buffer solution at a ratio of 1 g liver/2 ml buffer solution. The liver was homogenized by mixing for two 30-sec periods at high speed, using an Omni-Mixer (Omni Corporation International, Waterbury, CT). The homogenized liver was then centrifuged at 4°C for 30 min at 9000g (model J2-21; Beckman, Palo Alto, CA). Aliquots of the resulting supernatant were placed in several polypropylene tubes and stored frozen at 20°C. Esterase activity was determined by the method of Dean et al. (1991), which uses 4-methylumbelliferyl acetate as the substrate. The esterase activity was measured in freshly thawed liver (time 0) and at 1, 2, and 4 hr after incubation at 37°C, to estimate any loss of activity resulting from incubation. The formation of cocaethylene from cocaine and ethanol in rat liver homogenate was measured as a positive control and to assess carboxylesterase activity (data not shown). To rule out nonenzymatic transesterification, cocaethylene was incubated with ethanol, ²H₆-ethanol, or water in buffer solution. The total protein concentration was determined with the use of a commercial product (bicinchoninic acid protein assay; Pierce, Rockford, IL).

Drug Incubation. On the day of the incubation experiment, 1-ml aliquots of 100 µM solutions of cocaethylene in methanol were evaporated to dryness in a SpeedVac concentrator (Savant, Farmingdale, NY). The dry test tube contents were reconstituted with 0.9 ml of 70 mM KH₂PO₄ buffer solution, pH 7.4, and 0.1 ml of either deionized H₂O (control), 5.86% (v/v) ethanol, or 5.86% (v/v) ²H₆-ethanol. At the start of the incubation experiment, 1 ml of 9000g liver supernatant (S9 fraction) was added to each vial containing control, ethanol, or ²H₆-ethanol solution (total volume, 2 ml), and the vials were capped and placed in a 37°C water bath. The final concentrations in each vial were 50 µM drug with buffer only (control), 50 mM ethanol, or 51.3 mM ²H₆-ethanol. Ten samples of 100 µl each were taken from each tube, at the following times: 0, 5, 10, 15, 30, 45, 60, 120, 180, and 240 min. The 100-µl samples were immediately transferred to a vial (placed on ice) containing 100 µl of saturated NaF, to halt enzyme activity. The procedure described above was repeated in buffer solution for all three conditions, but only four samples each (0, 1, 2, and 4 hr) were collected. All experiments were conducted in triplicate and performed using a single homogenate preparation obtained from the livers of four rats.

Enzyme Kinetics. The K_M and V_max values for the ethanolic transesterification of cocaine and cocaethylene were obtained from triplicate assays with the rat S9 fraction incubated with eight different concentrations of substrate for 10 min at 37°C. Reactions were stopped by placing the reaction tube on ice and adding 100 µl of saturated NaF solution. The concentrations of ethanol and ²H₆-ethanol were kept constant at 50 mM (a physiologically relevant concentration equal to 0.230 g/100 ml ethanol), and the cocaine and cocaethylene concentrations ranged from 10 to 2500 µM. Three separate K_M and V_max determinations, each in triplicate, were made from the following combinations: cocaine plus ethanol, cocaine plus ²H₆-ethanol, and cocaethylene plus ²H₆-ethanol. Cocaethylene and ²H₅-cocaethylene concentrations were assayed, and the rate of formation was determined by dividing by the incubation time and protein concentration. Substrate concentrations (micromolar) and corresponding rates (picomoles per minute·milligram) from individual experiments were analyzed by nonlinear regression (WinNonlin Scientific Consulting Inc., Apex, NC), fitting the data to the full nonlinear (hyperbolic) form of the Michaelis-Menten equation to obtain estimates of V_max and K_M.

Analytical Procedures. Samples from all experiments were extracted for cocaethylene and ²H₅-cocaethylene with GV-50 Detectabuse solid-phase extraction columns (Biochemical Diagnostics Inc., Edgewood, NY). Columns were conditioned with 2 ml of methanol followed by 2 ml of 0.25 M K₂HPO₄ buffer, pH 9.1, with 7% n-propanol. To each sample was added 100 µl of internal standard solution (25 µM tropacocaine in H₂O) and 3 ml of 0.25 M Na₂HPO₄ buffer, pH 9.1. The tubes were vortex-mixed and then decanted into the preconditioned solid-phase extraction columns. The columns were washed with 4 ml of 0.0625 M Na₂HPO₄ buffer solution and dried under vacuum at 10 in Hg for 10 min. Samples were eluted with 1 ml of methylene chloride/isopropyl alcohol (80:20) solution by gravity and collected in disposable glass test tubes (12 × 75 mm). The eluted extract was evaporated to dryness and then reconstituted with 50 µl of methanol just before GC/MS analysis.

Data Analysis. Concentration-time profiles for cocaethylene and the deuterated exchange product were constructed and kinetic parameters were obtained using WinNonlin (Scientific Consulting). The apparent in vitro t_1/2 values were calculated as 0.693/k, where k is the slope of the log concentration vs. time plot. Statistical analysis of the apparent t_1/2, K_M, and V_max values was accomplished using Bonferroni's t test (SigmaStat Statistical Software for Windows; Jandel Scientific Software, San Rafael, CA). The harmonic mean and "pseudo"-standard deviation of the apparent t_1/2 values were calculated (Lam et al., 1985).

	Results and Discussion

Top Abstract Introduction Materials & Methods Results & Discussion References

The protein concentration of the rat S9 fraction was determined to be ~40 mg/ml. The mean esterase activity of freshly thawed rat S9 homogenate was 305 ± 4 nmol/min·mg of protein. Esterase activity did not significantly change after 1 and 2 hr of incubation (303 ± 14 nmol/min·mg and 293 ± 13 nmol/min·mg, respectively; using Bonferroni's t test for significance). After 4 hr of incubation, a statistically significant decrease in esterase activity was demonstrated (272 ± 5 nmol/min·mg; ~10% decrease from the control conditions). This reduced enzyme activity could affect the validity of the concentrations measured after 4 hr of incubation. No cocaethylene was detected after 4 hr of incubation under control conditions or after the addition of ethanol (fig. 1B). Deletion of the 4-hr value from the plot would not significantly alter the time profile for cocaethylene. Therefore, the significant loss of enzyme activity after 4 hr had little or no impact on the differences in cocaethylene t_1/2 values with and without ethanol.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Cocaethylene and 2H5-cocaethylene concentrations as a function of time. Cocaethylene (50 µM) was incubated with the rat liver S9 fraction for 4 hr at 37°C. A, Unlabeled cocaethylene concentrations () (harmonic mean t1/2 = 12.0 ± 1.1 min) in the presence of 2H6-ethanol (51.3 mM) and the resulting concentrations of formed 2H5-cocaethylene () (harmonic mean t1/2 = 79.0 ± 4.8 min). Dashed line, total cocaethylene concentrations (cocaethylene + 2H5-cocaethylene) () (harmonic mean t1/2 = 72.3 ± 5.4 min) in the presence of 2H6-ethanol. B, Cocaethylene concentrations as a function of time in the absence () (harmonic mean t1/2 = 13.0 ± 1.4 min) and presence of ethanol (50 mM) () (harmonic mean t1/2 = 67.8 ± 0.3 min). Each value is the mean of three experiments and the vertical bars represent the SD.

Incubation of cocaethylene in buffer only in the presence of ²H₆-ethanol resulted in no ²H₅-cocaethylene formation, ruling out nonenzymatic transesterification. There was, however, evidence of nonenzymatic hydrolysis of cocaethylene. From a starting concentration of 50 µM cocaethylene, there was a significant (p < 0.05) decrease in concentration after 4 hr of incubation (aqueous buffer, 41.5 ± 0.5 µM; ethanol, 41.6 ± 1.4 µM; ²H₆-ethanol, 42.9 ± 0.3 µM).

The concentration-time profiles for unlabeled cocaethylene in the presence of ²H₆-ethanol and for the formed ²H₅-cocaethylene are illustrated in fig. 1A. These data clearly indicate that ethyl ester exchange occurs between unlabeled cocaethylene and ²H₆-ethanol. The exchange process begins immediately upon incubation, inasmuch as measurable concentrations of ²H₅-cocaethylene are seen within 5 min. Fig. 1A also illustrates total cocaethylene concentrations (i.e. obtained from the sum of labeled and unlabeled cocaethylene concentrations). That profile for total cocaethylene was virtually identical to the cocaethylene concentration-time profile measured in the presence of unlabeled ethanol (fig. 1B). Similarly, the unlabeled form of cocaethylene provided nearly identical concentration-time profiles under control conditions (i.e. no ethanol) (fig. 1B) and in the presence of ²H₆-ethanol (fig. 1A). There were no statistically significant differences between those profiles in apparent t_1/2 values (harmonic mean t_1/2, 13.0 ± 1.4 min vs. 12.0 ± 1.1 min). This might appear to be inconsistent, because the disappearance of cocaethylene under control conditions is primarily the result of hydrolysis to benzoylecgonine and metabolism to other metabolites (i.e. ethylecgonine), whereas the disappearance of cocaethylene in the presence of ²H₆-ethanol involves an additional pathway of formed ²H₅-cocaethylene. However, the carboxylesterase enzyme is responsible for both the hydrolysis of cocaethylene to benzoylecgonine and the transesterification to ²H₅-cocaethylene. Because the disappearance of parent compound (unlabeled cocaethylene) is approximately the same under the two conditions, presumably less initial benzoylecgonine would be formed in the presence of ²H₆-ethanol, to compensate for the amount of parent compound being metabolized to ²H₅-cocaethylene. Benzoylecgonine was not measured in this experiment and it might be difficult to illustrate this hypothesis, because the formed ²H₅-cocaethylene would also be hydrolyzed to unlabeled benzoylecgonine. One might expect a similar area under the concentration-time curve (AUC) value for benzoylecgonine but a different profile, with less benzoylecgonine being formed early in the reactions involving ethanol and ²H₆-ethanol vs. control. Fig. 1B illustrates the in vitro disposition of 50 µM cocaethylene with and without 50 mM unlabeled ethanol. The apparent t_1/2 of cocaethylene was increased by ~5-fold in the presence of ethanol (13.0 ± 1.4 vs. 67.8 ± 0.3 min). The significant increase in the apparent t_1/2 value in the presence of ethanol is explained by the ethyl ester formation of additional cocaethylene. This is corroborated by the extent of formation of ²H₅-cocaethylene from ²H₆-ethanol (fig. 1A). The change in the cocaethylene concentration-time profile in the presence of ethanol (fig. 1B) is identical to the change noted in a comparison of unlabeled cocaethylene and total cocaethylene in the presence of labeled ethanol (fig. 1A). The apparent t_1/2 of unlabeled cocaethylene, 12.0 ±1.1 min, was increased to 72.3 ± 5.4 min for total cocaethylene (labeled and unlabeled) in the presence of ²H₆-ethanol, a ~6-fold increase.

The experiments using enzyme inhibitors yielded an "all or nothing" response. Although the concentrations of cocaethylene and ²H₅-cocaethylene were quantitatively determined, they are reported here qualitatively. When cocaethylene was incubated with the S9 fraction and ²H₆-ethanol in the presence of BNPP, a specific carboxylesterase inhibitor, absolutely no formation of the product ²H₅-cocaethylene was detected. Therefore, BNPP, a specific carboxylesterase inhibitor, totally blocked the ethyl ester exchange reaction between cocaethylene and ²H₆-ethanol. As expected, no ²H₅-cocaethylene was detected in the buffer condition (negative control) or when saturated NaF, a nonspecific esterase inhibitor, was added. However, when physostigmine, a cholinesterase inhibitor, was added to the S9 fraction, ²H₆-ethanol, and cocaethylene incubation mixture, the formation product, ²H₅-cocaethylene, was detected at approximately the same concentration as under positive control conditions (the S9 fraction with no inhibitors). Therefore, this ethyl ester exchange is most likely catalyzed by a carboxylesterase enzyme, namely the same transesterification enzyme system responsible for the formation of cocaethylene from cocaine and ethanol.

For the enzyme kinetic analysis, the rate of ethyl ester formation was measured at different substrate concentrations and the data were fit to the Michaelis-Menten equation. The apparent K_M for cocaethylene formation from cocaine and ethanol in the rat S9 fraction was 327 ± 6 µM (fig. 2A). The K_M value for this reaction has been reported by others to be 116 µM (Woodward et al., 1991) and 864 µM (Katz et al., 1992) in humans and mice, respectively. When ²H₆-ethanol was substituted for ethanol, ²H₅-cocaethylene formation from the cocaine substrate yielded a K_M value of 366 ± 25 µM (fig. 2B). Bonferroni's t test determined that these two values were not significantly different from each other, ruling out a kinetic isotope effect with the deuterated ethanol. V_max values for cocaethylene and ²H₅-cocaethylene formation from the cocaine substrate were 305 ± 6 and 319 ± 9 pmol/min·mg, respectively (fig. 2, A and B). The apparent K_M and V_max values for ²H₅-cocaethylene formation from cocaethylene and ²H₆-ethanol were 223 ± 4 µM and 428 ± 12 pmol/min·mg, respectively (fig. 2C). These values were significantly different from the cocaine substrate K_M and V_max values (p < 0.05). From these data it appears that the rat carboxylesterase enzyme more efficiently catalyzes the ethanolic transesterification of cocaethylene, compared with that of cocaine. The lower K_M value (223 vs. 366 µM) indicates that the enzyme has a greater affinity for cocaethylene than for cocaine for ethanolic transesterification. The higher V_max (428 vs. 319 pmol/min·mg) indicates more rapid transesterification of cocaethylene, compared with cocaine, under similar substrate concentration conditions. This conclusion is further supported by estimates of intrinsic metabolite formation clearance (V_max/K_M) values for cocaethylene formation. The values for cocaethylene formation clearance from cocaine and ethanol or ²H₆-ethanol are 0.94 ± 0.01 and 0.87 ± 0.04 µl/min·mg protein, respectively. In contrast, the intrinsic formation clearance estimate for ²H₅-cocaethylene from cocaethylene and ²H₆-ethanol is 1.92 ± 0.03 µl/min·mg protein, a value several times the formation clearance from cocaine.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Rate of metabolite formation as a function of substrate concentration for determination of KM and Vmax values. Substrate and ethanol were incubated with the rat liver S9 fraction at 37°C for 10 min. Substrate concentrations were 10, 25, 50, 100, 250, 500, 1000, and 2500 µM. A, Cocaethylene formed in the presence of cocaine (10-2500 µM) and ethanol (50 mM) (KM = 327 ± 6 µM, Vmax = 305 ± 6 pmol/min·mg). B, 2H5-Cocaethylene formed in the presence of cocaine (10-2500 µM) and 2H6-ethanol (51.3 mM) (KM = 366 ± 25 µM, Vmax = 319 ± 9 pmol/min·mg). C, 2H5-Cocaethylene formed in the presence of cocaethylene (10-2500 µM) and 2H6-ethanol (51.3 mM) (KM = 223 ± 4 µM, Vmax = 428 ± 12 pmol/min·mg). Each experiment was performed in triplicate; each point represents the mean and the vertical bars represent the SD.

The most important and novel aspect of this communication is the observation that cocaethylene undergoes ethyl ester exchange in the presence of ethanol. The prolonged half-life and reduced clearance of cocaethylene in the presence of ethanol are not indications of altered hepatic clearance because of ethanol interaction with metabolizing enzymes. That conclusion is supported by the observation that cocaethylene under control conditions (i.e. no ethanol) (fig. 2B) and unlabeled cocaethylene in the presence of ²H₆-ethanol (fig. 2A) both underwent rapid and nearly identical elimination. Rather, a reversible metabolic system is established, with an excess of ethanol providing a substrate pool of an exchangeable ethyl group. This process would be expressed in a clinical setting as prolonged cocaethylene elimination in the presence of ethanol after cocaethylene has been formed by transesterification. Continued ethanol ingestion should further prolong the pharmacological and toxicological effects of the formed cocaethylene. We are currently examining this interaction using an in vivo animal model and a human liver preparation.