Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Approximately 2.5 million Americans are "heavy" abusers of cocaine, and a similar number of Americans are "light" users of cocaine (Das, 1993). Cocaine abuse has a significant pharmacokinetic component because cocaine is generally not abused via the oral route of administration (Jones, 1990). Adverse effects of cocaine overuse include blood pressure elevation, seizures, cardiac arrest, "sudden death", brain damage, psychosis, and reproductive disturbances (Adams et al., 1987; Brody et al., 1990; Isnen and Chokshi, 1991; Rich and Singer, 1991).

Although progress is being made in the development of pharmacological intervention of cocaine abuse, it is important to develop alternative strategies that use new approaches to prevent cocaine addiction and overuse. For example, studies have shown that the administration of purified human butyrylcholinesterase (BuChE) protects animals from the seizures and lethal effects of cocaine (Hoffman et al., 1996; Lynch et al., 1997; Mattes et al., 1997). Other studies have shown that antibodies that selectively bind cocaine are effective in preventing cocaine toxicity to animals (Fox et al., 1996). Another approach has shown that catalytic antibodies that hydrolyze cocaine are capable of detoxicating cocaine in animals (Landry et al., 1998). However, these studies used the administration of an antibody before cocaine and studies to assess the protective effect after cocaine administration have not been done. Antibody-based treatments have been used for many years for drug overdose detoxication, including digoxin (Smith et al., 1976), desipramine (Brunn et al., 1992), phencyclidine (Valentine et al., 1996), colchicine (Sabouraud et al., 1991), and others. Another related approach is the use of active immunization to suppress the psychoactive effects of cocaine (Carrera et al., 1995).

Catalytic antibodies that are elicited against a stable synthetic analog of the transition state of the hydrolytic reaction to be catalyzed could have many advantages over pharmacological intervention of cocaine abuse (Landry et al., 1993; Yang et al., 1996). First, the administration of an antibody is relatively safe. Second, a catalytic antibody could remove cocaine selectively and not interfere with endogenous neurotransmitters or their receptors or transporters, thus reducing the ancillary effects associated with pharmacological agents directed against neurotransmitter-dependent sites. Finally, hydrolysis of cocaine, compound 1a, leads to pharmacologically inactive metabolites such as compounds 2a and 2b (Fig. 1; Landry and Yang, 1997). Although BuChE constitutively protects against ()-cocaine toxicity, it does so by hydrolyzing cocaine sluggishly (Inaba et al., 1978; Stewart et al., 1979), and there is a need to develop a catalyst that is much more efficient.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Top, proposed scheme for the hydrolysis of ()-cocaine, compound 1a, to methyl ecgonine, compound 2a, and ecgonine, compound 2b. TS refers to the putative transition state of the reaction. Bottom, structures of phosphonate transition-state analogs and haptens and conjugates used in the procurement of catalytic antibodies of cocaine.

In the present study, we report the procurement of anti-cocaine catalytic antibodies elicited by a stable, long-lived hapten, compound 5a (Fig. 2). The lack of degradation of the hapten used in this study may be in part responsible for the identification and procurement of efficient catalysts reported herein. Another novel feature of the work described was the use of rapid direct screening of hybridoma supernatants for antibody-mediated catalysis by using a continuous high-throughput photometric assay. Through the use of the direct screening of catalytic activity, we were able to rapidly evaluate large numbers of clones elicited against a phosphonate transition-state analog of cocaine. High-throughput screening of a large number of clones enabled the procurement of catalytic antibodies that approached the activity of human BuChE in the catalytic efficiency of ()-cocaine hydrolysis. The important contribution of this work is that by using a rapid high-throughput assay, numerous catalytic antibodies that would not have been identified on the basis of traditional screens using binding assays were obtained that possessed high hydrolytic activity against ()-cocaine.

View larger version (6K): [in this window] [in a new window]   Fig. 2.   Hydrolysis of cocaine benzoyl thioester, compound 6 (R = methyl), in the presence of DTNB provides a photometric assay for hydrolysis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Chemicals, reagents, and biological agents used in the present study were of the highest purity available from commercial sources. ()-Cocaine, [³H]N-8-CH₃-()-cocaine, and authentic synthetic metabolites of cocaine were provided by the National Institute on Drug Abuse Drug Supply Program (Rockville, MD). Cocaine benzoyl thioester was synthesized and characterized as previously described (Cashman et al., 1998). The phosphonate hapten and related derivatives and hydrolysis products were synthesized as previously described (Berkman et al., 1996) and conjugated to keyhole limpet hemocyanin (KLH) according to the method outlined later. The substrate, 2,2'-azino-di(3-ethyl-benzthiazoline sulfonate was obtained from Kirkegaard and Perry Labs, Inc. (Gaithersburg, MD). 5,5'-Dithiobis(2-nitrobenzoic acid) (DTNB) was purchased from Aldrich Chemical Co. (Milwaukee, WI). Peroxidase-linked goat anti-mouse IgG antibody was purchased from Bio-Rad (Hercules, CA). Fetal bovine serum was purchased from Gemini Bioproducts (Woodland, CA).

Instrument Analysis. ¹H NMR spectra were recorded on a Varian 300 spectrometer (Varian Analytical, Sunnyvale, CA) operating at a frequency of 300 MHz. Fast atom bombardment mass spectra were recorded on a VG 70SEQ instrument. Both instruments were housed at the Department of Medicinal Chemistry, University of Washington (Seattle, WA). An enzyme-linked immunosorbent assay (ELISA) was performed with a Bio-Tek ELISA Plate Reader (model 311; Winooski, VT).

Synthesis of Conjugated Hapten Compound 5. The hapten compound 5 (2.0 mg) was dissolved in H₂O (200 µl) and Na₂CO₃ (1.0 M, 100 µl). 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide was dissolved in H₂O (100 µl) and Na₂CO₃ (1.0 M, 100 µl) and added to the hapten-containing solution. This mixture was then added to a solution of KLH (10 mg in 500 µl of H₂O) (ICN Biomedicals, Inc., Aurora, OH) and was stirred overnight at room temperature. The reaction mixture was transferred to dialysis tubing and dialyzed against H₂O (500 ml) for 24 h, with the water changed once. The protein concentration of the hapten-conjugated KLH solution used was 4.22 mg/ml. Protein concentration was determined according to the bicinchoninic acid protein assay method (Pierce Chemical Co., Rockford, IL).

Quantification Conjugated Hapten Compound 5. The percentage of conjugation of hapten to carrier protein KLH for immunization was determined by titration of the number of free amines (lysine residues) using o-phthaldehyde. To a cuvette containing 321 µg of protein KLH in 200 µl of water, we added 800 µl of o-phthaldehyde reagent (10 mg of o-phthaldehyde, 10 µg of 2-mercaptoethanol, 125 µl of MeOH, and 1.5 ml of 0.4 M borate buffer, pH 10.4). The UV-visible (vis) absorbance (Abs_KLH) was measured at 336 nm after 5 min and subtracted from a matched cuvette containing only water and o-phthaldehyde reagent as a reference. To a cuvette containing hapten-conjugated KLH (0.32 mg), we added water and the o-phthaldehyde reagent. Absorbance (Abs_KLH-hap) was measured after 5 min at 336 nm using a matched cuvette containing an equivalent amount of water and o-phthaldehyde reagent as a reference. The percentage conjugation was determined using the following equation: conjugation (%) = (1  Abs_KLH/Abs_KLH-hap) × 100. Generally, the percentage conjugation was between 28 and 35%.

Relative Aqueous Stability of Conjugated Hapten. The relative rates of hydrolysis of compounds 3 and 4a were compared to investigate the stability in aqueous solution, pH 7.4. The phosphonate methyl ester, compound 4a, and the phosphonate caproamide, compound 3, were placed in 100 mM potassium phosphate buffer (pH 7.4). Aliquots were periodically withdrawn from the reaction, cold acetonitrile was added, and the mixture was lyophilized to dryness, dissolved in isopropanol, and injected onto the HPLC. The profile of products was determined and quantified after separation by HPLC with an Hitachi L6200A HPLC interfaced to an Hitachi D-2500 Chromato-Integrator with an Hitachi L-4000H UV detector set at 220 nm (Hitachi Scientific, Mt. View, CA). The end loop of the UV-Vis detector was routed into a Sedex evaporative light-scattering (ELS) detector (Richard Scientific, Novato, CA). The system was fitted with an AXXIOM silica gel column (25 × 0.4 cm; Richard Scientific, Novato, CA) and used a mobile phase of isopropanol/methanol/HClO₄ (60:40:0.4 v/v/v) at a flow rate of 1.5 ml/min. This system efficiently separated compounds 4a, 3, and 4b. Compound 4a had a retention time of 10.3 min, compound 3 had a retention time of 9.2 and 11.0 min (the molecule is composed of diastereomers), and the common product (i.e., compound 4b) had a retention time of 7.6 and 8.2 min (the product molecule is also composed of diastereomers). Quantification was based on calculating the percentage of conversion to product as determined by the peak heights of the chromatograms.

Synthesis of meta-Nitrococaine. To a stirred solution of ecgonine methyl ester (50 mg, 0.25 mmol) in CH₂Cl₂ at 0°C, we added meta-nitrobenzoyl chloride (93 mg, 0.5 mmol). Triethylamine (50 mg, 0.5 mmol) was added, and the reaction was allowed to warm to room temperature and stirred for 1 h, after which a white precipitate formed. The mixture was treated with cold Na₂CO₃ (10% w/v) and extracted twice with CH₂Cl₂. The organic fractions were combined twice and dried over Na₂SO₄. The crude mixture was purified by flash alumina oxide chromatography (MeOH/CH₂Cl₂/NH₄OH, 10:89:1, v/v/v; R_f = 0.83) to give the product as a yellow oil (67 mg, 77% yield). ¹H NMR (CDCl₃):  1.72-1.75 m, 2 H; 1.88-1.94 m, 1 H; 2.15-2.16 m, 1 H; 2.24 s, 3 H; 2.45 t, 1 H; 3.05 br s, 1 H, 3.32 br s, 1 H; 3.6 br s, 1 H; 3.75 s, 3 H; 5.27-5.32 m 1 H; 5.31 d, 1 H (J = 3.4); 7.61-7.67 m, 1 H; 8.34-8.42 m, 2 H, 8.84 br s, 1 H.

Analysis of meta-Nitrococaine Formation from Ecgonine Methyl Ester. Catalytic antibody-mediated ()-cocaine hydrolysis activity was determined at room temperature. The amount of cocaine hydrolyzed to ecgonine methyl ester in the presence of antibody was determined by HPLC after derivatization with meta-nitrobenzoyl chloride and analysis of extracts as described later. Into each test tube approximately 200 µg of affinity-purified antibody was placed. The hybridoma supernatant was purified by affinity chromatography with protein A-Sepharose 4 Fast Flow (Pharmacia Biotech, Piscataway, NJ) as described later. The incubation was initiated by the addition of cocaine (250 nmol) in a total reaction volume of 0.5 ml at pH 7.0. The rate of hydrolysis was measured from the amount of product formed after correction for nonenzymatic background hydrolysis determined in parallel control incubations. Incubations were run for 2 h. Thereafter, 2 ml of cold CH₂Cl₂ was added, followed by 20 mg of solid Na₂CO₃ to stop the reaction. The incubation was mixed, and the organic layer was separated from the aqueous layer by a brief centrifugation and evaporated to dryness. To each sample, we added triethylamine (2.5 µmol) followed by meta-nitrobenzoyl chloride (1.25 µmol) in a total volume of 100 µl of tetrahydrofuran. After 1 h at 50°C, the derivatization reaction was stopped by the addition of 1 ml of diethyl ether, 250 µl of saturated Na₂CO₃ (50:50, v/v), and 10 mg of NaCl. After thorough mixing and separation of the organic layer by centrifugation, the solvent was evaporated with a stream of argon and placed in isopropanol (250 µl) for analysis by HPLC. Unhydrolyzed cocaine and meta-nitrococaine were separated and quantified with the Hitachi HPLC system described earlier with UV-vis detection set at 236 nm. The system was fitted with an AXXIOM silica gel analytical column (25 × 0.4 cm). The mobile phase consisted of an isocratic system set at methanol/isopropanol/60% perchloric acid (50:50:0.04, v/v) at a flow rate of 1.5 ml/min. This system efficiently separated cocaine and meta-nitrococaine, which had retention times of 8.55 and 9.99 min, respectively. The minor peak area corresponding to meta-nitrococaine in buffer controls (incubations with cocaine alone or derivatizing reagents alone) was subtracted from the antibody-mediated hydrolysis to determine the true rate of hydrolysis for each sample. For quantification, the HPLC peak area of cocaine and meta-nitrococaine was compared after accounting for the 6.5-fold increase in the extinction coefficient of meta-nitrococaine.

Measurement of Selectivity Constants. Inhibition of antibody-catalyzed cocaine benzoyl thioester were determined at room temperature spectrophotometrically according to the method described previously (Cashman et al., 1998). First, competitive inhibition of the hydrolysis of cocaine benzoyl thioester was established by conducting a Lineweaver-Burke study in the presence of a standard amount of competitor. To each cuvette, we added 880 µl of a 5.5 mM DTNB solution in 25 mM potassium phosphate buffer (pH 7.4) followed by the addition of antibody (348 µg) and the competitor (in a total reaction volume of 1 ml). The reactions were initiated by the addition of cocaine benzoyl thioester (3-200 µM), and the initial phase of the linear increase in absorbance was monitored for 2 to 10 min. The rate of hydrolysis was corrected at each substrate concentration for background hydrolysis. After establishing the competitive nature of the inhibition, Dixon analysis was performed to determine the K_i value for each competitor (range of competitor concentrations, 5-1000 µM for norcocaine and ecgonine and 100-5000 µM for ecgonine benzoyl ester and ecgonine methyl ester).

Hybridomas and Monoclonal Antibodies (mAbs). Four BALB/cBYJ mice were immunized each with 50 µg of compound 5a hapten in complete Freund's adjuvant. After 14 days, a boost was administered (25 µg per mouse in incomplete Freund's adjuvant), and two additional injections were given 5 and 3 days before harvesting the spleens. Fusion of spleen cells with SP2/O myeloma cells was described previously (Kohler and Milstein, 1975) with the exception that a 1:1 ratio of myeloma cells to spleen cells were used in the fusion and the fused cells were grown in RPMI-20 with hypoxanthine, aminopterin, and thymidine selection medium. Two to 3 weeks after fusions, the supernatants of the hybridomas were screened for cocaine benzoyl thioester hydrolase activity according to the procedure previously described (Cashman et al., 1998). To test for spontaneous hydrolysis, control reactions were performed in the presence of supernatant derived from hybridomas from sham-treated mice or in the presence of media alone. Each hybridoma cell line with significant cocaine benzoyl thioester hydrolase activity was cloned to homogeneity by limiting dilution at least twice. mAbs were propagated under standard conditions except the cells nearing confluency were gradually weaned off RPMI-1640 with EX-CELL 620 HSF (JRH Biosciences, Lenexa, KS), a serum-free medium that facilitated the isolation of the desired catalytic antibodies. The mAbs were purified with the use of tandem chromatography of Sephadex G-25 followed by an affinity column of protein A-Sepharose 4 Fast Flow and placed in PBS. The mAbs 3, 5, and 12 (Tables 1 and 2) were judged to be homogeneous on the basis of SDS-polyacrylamide gel electrophoresis under reducing conditions that showed only heavy and light chains. The isotype for mAbs 3, 5, and 12 was IgG₁ as determined with the IsoStrip Mouse mAb isotyping kit from Boehringer-Mannheim Biochemicals (Indianapolis, IN). mAb 3 grew slowly in cell culture, and the antibody ultimately used in ()-cocaine hydrolytic kinetic studies was derived from ascites fluid. Ascites fluid was produced at Sierra BioSource (Gilroy, CA) and purified by affinity chromatography as described earlier. The studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals as adopted by the National Institutes of Health.

[³H]Cocaine Hydrolysis. The profile of hydrolysis products of [³H]()-cocaine catalyzed by the highly purified mAbs 3, 5, and 12 were determined by quantifying radioactivity and peak height with HPLC by UV-vis and ELS detection. HPLC-purified [³H]()-cocaine stock solutions were freshly made up in isopropanol. A typical incubation mixture contained 100 to 300 µg of mAb in 25 mM potassium phosphate buffer (pH 8.4) in a final volume of 0.25 ml. The reaction was initiated by the addition of [³H]()-cocaine and incubated with shaking for 40 min at 37°C. As a control for spontaneous hydrolysis, a parallel nonenzymatic reaction was run under pseudo-first-order conditions, and the true rates of hydrolysis were calculated by subtracting the enzymatic rate from the spontaneous rate. The spontaneous or uncatalyzed rate of hydrolysis of cocaine (k_o) was determined by initial rate measurements (k_o = 1.4 × 10³ min¹) at pH 8.4. Each incubation was stopped by the addition of 500 µl of cold acetonitrile, mixed, and centrifuged to precipitate the protein. Approximately 60% of the mixture was decanted and evaporated to dryness under a stream of argon. The sample was taken up in methanol and immediately injected onto the HPLC. The hydrolysis products were separated using the Hitachi HPLC system described earlier set at 237 nm. The system was fitted with an AXXIOM silica gel analytical column (25 × 0.4 cm). The mobile phase consisted of an isocratic system set at methanol/isopropanol/ammonium hydroxide (85:15:0.03, v/v/v) at a flow rate of 1 ml/min. This system efficiently separated ecgonine methyl ester, cocaine, benzoyl ecgonine, and ecgonine that had retention times of 2.2, 4.7, 6.0, and 8.2 min, respectively. The eluant of the entire HPLC chromatogram was collected in scintillation vials, and each 0.5-min fraction was individually measured with a Beckman LS 5000CE scintillation counter (Sunnyvale, CA) and quantified on the basis of radioactivity.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Antibodies to a phosphonate transition-state analog of cocaine benzoate ester hydrolysis were elicited in mice. The phosphonate analog was used because of its known properties to mimic the negatively charged, tetrahedral transition-state intermediate of ester hydrolysis (Fig. 1). The approach taken was to determine the catalytic activity of culture fluid from hybridoma wells directly with rapid high-throughput photometric assays using cocaine benzoyl thioester. Cocaine benzoyl thioester was used because it provided a rapid means of identifying esterase activity in the face of a large number of catalytic antibody candidates. Results for cocaine benzoyl thioester hydrolysis in some cases were confirmed with studies of hydrolysis of cocaine as determined with an efficient and sensitive HPLC assay for meta-nitrococaine. The use of these assays allowed the detection of highly active catalysts that would have not been identified on the basis of screening for antibodies with high-affinity binding alone. In later studies, binding avidity of hybridoma supernatant to the phosphonate hapten was also determined. A significant number of highly active catalysts were present in hybridoma supernatants that did not have the highest binding affinity. Active antibody catalysts that had esterase activity significantly above uncatalyzed background rates were purified and characterized in detail for ()-cocaine hydrolysis.

Catalytic Activity and Binding Avidity. From the initial cloning of spleen cells of a single mouse immunized and boosted with conjugated hapten compound 5a, approximately 450 cell culture supernatants were obtained. These parental cell line supernatants were screened for cocaine benzoyl thioester hydrolysis using a continuous DTNB-based photometric assay at pH 7.4 and 8.5 (Fig. 2). Kinetics of hydrolysis at two pH values were chosen because of numerous reports that esterolytic catalytic antibodies were more efficient at elevated pH values (Tramontano, 1994). Hydrolysis of cocaine benzoyl thioester by hybridoma supernatant was followed for 4 h, and initial rate constants for each supernatant were obtained. All of the supernatants examined were analyzed with plots of the relative change in A₄₀₅ versus time. From the linear portion of initial rate measurements of product versus time, observed rate constants (k_obsd) were obtained. Supernatants with efficient cocaine benzoyl thioester hydrolysis activity, moderate activity, and very low activity were observed. Each observed rate constant was replotted on a bar graph (e.g., shown in Figs. 3 and 4). Plots of the relative rate constants versus 96-well supernatant position for the same 96-well plates showed distinct profiles for hybridoma supernatant at pH 7.4 and 8.5 (Figs. 3 and 4, respectively). Control reactions showed that spontaneous hydrolysis of cocaine benzoyl thioester was detectable and increased with increasing pH. Thus, the background rate of hydrolysis is elevated at pH 8.5 compared with pH 7.4 (Fig. 3 versus Fig. 4). To examine the data directly, the spontaneous rate was not subtracted from the observed rate. For the 450 parental hybridoma supernatants examined, 45 candidates were chosen for subcloning on the basis of the catalytic activity. For example, for the plate shown in Fig. 3, hybridomas 4, 8, 10, 11, and 62 were chosen for expansion and further cloning. For the plate shown in Fig. 4, hybridomas 25, 32, 56, and 92 were chosen for expansion and further cloning.

View larger version (58K): [in this window] [in a new window]   Fig. 3.   Plot of the observed rate constants (kobsd minutes1) versus the position of a hybridoma supernatant on a 96-well microplate. Each initial rate constant is represented by a bar for the hydrolysis of cocaine benzoyl thioester, pH 7.4.

View larger version (65K): [in this window] [in a new window]   Fig. 4.   Plot of the observed rate constants (kobsd minutes1) versus the position of a hybridoma supernatant for the 96-well microplate shown in Fig. 3. Each initial rate constant is represented by a bar for the hydrolysis of cocaine benzoyl thioester, pH 8.5.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Scattergram plot of the affinity for phosphonate hapten 5b versus the nanomoles of ()-cocaine hydrolysis. Cocaine hydrolysis was determined for 70 purified antibodies by HPLC after derivatization of the ecgonine methyl ester formed with meta-nitrobenzoyl chloride.

Kinetic Studies. As a prelude to detailed studies of the hydrolysis of ()-cocaine by selected mAbs, we examined the kinetics and selectivity of cocaine benzoyl thioester hydrolysis. This was done because of the limited amount of highly purified antibody available and relatively small amount of antibody required to study cocaine benzoyl thioester hydrolysis. From the 11 active mAbs shown in Table 1, the three most catalytically active antibodies (i.e., mAbs 3, 5, and 12) were chosen for detailed study. Hydrolysis of cocaine benzoyl thioester was linearly dependent on purified mAb up to 125 µg of antibody and linearly dependent on time for at least 10 min. Using standard incubation conditions, background studies of cocaine benzoyl thioester, compound 6, showed a minimum hydrolysis (Fig. 2). Previously, we observed that cocaine benzoyl thioester hydrolysis was strongly dependent on buffer concentration and pH (Cashman et al., 1998). By maintaining the buffer concentration at 25 mM, this significantly decreased the background rate of hydrolysis. Cocaine benzoyl thioester was not indefinitely stable at the pH values examined, and a detectable amount of spontaneous hydrolysis was observed. In practice, the minor amount of spontaneous hydrolysis was subtracted from the enzyme rate to yield the true rate for the data reported in Tables 1 to 3. When the concentration of cocaine benzoyl thioester was varied, the K_mapp and V_max values for thioester hydrolysis were obtained from linear Lineweaver-Burke double-reciprocal plots of 1/velocity versus 1/substrate. We observed clear enzyme saturation kinetic behavior that was consistent with Michaelis-Menton kinetics. The values are listed in Table 2. At very long incubation times, it is possible that some minor amounts of the methyl ester of cocaine benzoyl thioester hydrolysis did arise, but we did not examine this point.

View this table: [in this window] [in a new window]   TABLE 3 Inhibition of cocaine benzoyl thioester hydrolysis in the presence of mAb 5  Competitive inhibition was determined by the continuous photometric assay described in Materials and Methods at pH 7.4 (Cashman et al., 1998).

()-Cocaine Hydrolase Activity. On the basis of results of catalytic antibody studies with cocaine benzoyl thioester, we examined selected mAbs as catalysts for ()-cocaine hydrolysis. As a prelude to the study of cocaine esterolytic activity of the mAbs, the chemical stability of [³H]()-cocaine to spontaneous hydrolysis was studied. [³H]Cocaine was slightly but detectably hydrolyzed in the absence of the mAbs under the same conditions of the mAb incubations as analyzed by radiometric-HPLC analysis with ELS detection. Use of this detection system was necessary to quantify UV-active and non-UV-active products and substrate. The minor amount of spontaneous hydrolysis of ()-cocaine was subtracted from the antibody initial rates to obtain the true rate reported in Table 4. A Lineweaver-Burke plot of the reciprocal of the velocity versus the substrate concentration for mAb 3 is shown in Fig. 6.

View this table: [in this window] [in a new window]   TABLE 4 Kinetic constants for ()-cocaine hydrolysis by purified mAb and BuChE

View larger version (9K): [in this window] [in a new window]   Fig. 6.   Lineweaver-Burke plot of the reciprocal of nmol of ecgonine formation versus the reciprocal of substrate concentration for mAb 3 (r2 = 9909).

1

1

1

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

To elicit catalytic antibodies, we used the 2-carboxamido-3-phosphonate hapten 5a because of its ability to mimic the tetrahedral, negatively charged transition state of ester hydrolysis (Jencks, 1969) and because the carboxamido group provided enhanced stability over similar ester haptens. The stability of the amide moiety has been reported for another related cocaine hapten (Sakurai et al., 1996). Presumably, the anionic character of the phosphonate enlists functional groups of the antibodies to orient around the hapten for electrostatic complementarity, or "bait and switch" (Spivak et al., 1999). For most esterolytic antibodies, optimal activity has been observed at elevated pH values, and this is why hybridomas were screened at neutral and alkaline pH. In good agreement with previous work, catalytic antibody activity was greater at higher pH (Fig. 3 versus Fig. 4), and this suggests that hydroxide ion may be required for acyl-intermediate hydrolysis. Alternatively, antibody amino acid residues, including an active-site tyrosine, may be at work to increase the rate of hydrolysis (Martin et al., 1991; Guo et al., 1995).

Hybridoma supernatants were directly assayed for their ability to catalyze the hydrolysis of cocaine benzoyl thioester. We chose to first screen for catalysts as opposed to initially screening for high-affinity hapten-binding activity to enhance our ability to select cocaine esterase mAbs (Tawfik et al., 1993). The use of a continuous photometric assay for catalytic activity with a thioester analog of cocaine affords several advantages, including 1) the catalytic antibody selected possesses increased recognition of substrate, 2) the relatively low substrate concentrations used increases the percentage conversion to product, and 3) the high-throughput nature of the process conserves limited amount of antibody to allow the focus on catalyst candidates with the greatest activity in the face of a large number of hybridoma clones. As shown in Fig. 5, binding affinity of mAbs to the phosphonate hapten did not exactly correlate with the cocaine hydrolytic catalytic activity of a large number of antibodies examined. This point was further verified when selected mAbs were examined (Table 1). It is possible that the hapten chemical features that produce efficient binding but nonproductive recognition for catalysis were obviated. From an enzyme energetics point of view, the screening procedure may have identified catalytic antibodies with greater free energy dedicated to improving catalytic activity rather than devoting most of its free energy to binding phenomena.

Assay with cocaine benzoyl thioester also allowed additional information about the mechanism of mAb catalysis. The kinetic constants obtained for mAb were comparable with those of mammalian BuChEs (Table 2). Metabolites of ()-cocaine had very little affinity for mAb 5 on the basis of competitive inhibition studies. With the exception of ecgonine, the mAb 5 was relatively selective for the cocaine nucleus. The results are consistent with ()-cocaine metabolites not serving as product inhibitors of mAb-5 (Table 3). Although we did not expressly test other mAbs for competitive inhibition, it is likely that based on the structural identity of catalytic antibodies elicited against a common hapten, mAbs 3 and 12 would also show considerable lack of affinity for ()-cocaine metabolites. Product inhibition is commonly observed for mammalian cholinesterases, and the lack of product inhibition for mAb 5 would improve the usefulness of a catalytic antibody as a therapeutic agent.

The feasibility of using cocaine benzoyl thioester as a surrogate substrate for screening anti-cocaine catalytic antibodies was borne out by examining the three most active mAbs for ()-cocaine hydrolysis by HPLC. It was necessary to use a radiometric assay with UV-vis and ELS detection to quantify the UV-active and non-UV-active products and substrates. An intriguing feature of the cocaine hydrolytic activity is the nature of the products formed. For mAbs 3 and 5, the sole product observed was ecgonine. This means that hydrolysis of both the C2 and C3 esters was observed. Once the methyl ester is hydrolyzed, the benzoyl ester is rapidly hydrolyzed in an apparent A  B  C reaction. That no detectable ecgonine methyl ester is formed suggests that the preferred site of hydrolysis is at C3, but C2 methyl ester hydrolysis may be a prerequisite for efficient C3 benzoyl ester hydrolysis. Steric inhibition to C3 hydrolysis is also possible, and after relief of steric strain, hydrolysis of the benzoyl ester is much more efficient. Previously, such multistep enzyme reactions have been observed where initial reaction products have not desorbed from the surface of an enzyme (Du et al., 1995), and this phenomenon has been called "catalytic facilitation." Alternatively, it is possible that the kinetic parameters of the first step in the A  B  C reaction determine the overall kinetic outcome. This point was not specifically investigated.

The k_cat values for hydrolysis of ()-cocaine by mAb 3 (k_cat = 6.6 min¹) compare favorably with those of BuChE from human (k_cat = 3.9 min¹) or horse (k_cat = 7.5 min¹; Xie et al., 1999). However, due to the relatively high K_m value for ()-cocaine, the catalytic efficiency of mAb 3 (k_cat/K_m = 6921 M¹ min¹) is significantly less than that of BuChE from human (k_cat/K_m = 280,000 M¹ min¹) or horse (k_cat/K_m = 170,000 M¹ min¹). The catalytic efficiency of mAb 5 (k_cat/K_m = 160,000 M¹ min¹; Table 4) approaches the above noted wild-type enzyme values. mAb 12 was considerably less active than mAbs 3 and 5. In contrast to mAbs 3 and 5, detectable amounts of benzoyl ecgonine were formed in the presence of mAb 12. Benzoyl ecgonine product formation obeyed Lineweaver-Burke kinetics (K_m = 97.7 µM and V_max = 0.04 nmol/min/mg of protein). Apparently, as ester hydrolysis at C3 becomes less favorable, the first step (i.e., C2 methyl ester hydrolysis) in the A  B  C reaction may become kinetically more apparent, and this may translate into detectable levels of intermediates being formed and detected along the ultimate product pathway.

Previous studies have shown that BuChE enzymes bind (+)- and ()-cocaine with similar efficiencies, but their rate of hydrolysis differ by approximately 2000-fold (Gately, 1991). It was proposed that the methyl ester on the tropane ring interferes with the binding of ()-cocaine but not (+)-cocaine, and this explains why (+)-cocaine is rapidly hydrolyzed by BuChE (Xie et al., 1999). This observation was extended to include phosphonothiolate transition-state analog inhibitors of ()- and (+)-cocaine hydrolysis (Berkman et al., 1997). The data suggest that the binding domain and the catalytic domain of BuChE are distinct regions. The enzyme models proposed are consistent with the presence of three enzyme-substrate complexes leading to the critical acyl-enzyme intermediate (Radic et al., 1993). It is unknown whether the catalytic antibodies described herein possess distinct binding and catalytic regions in analogy with the cholinesterases, but the development of catalytic antibodies that hydrolyze both esters has an advantage in producing detoxication products that have no potential for central nervous system stimulation.

In conclusion, direct screening for catalytic activity of hybridoma supernatants of spleen cells fused with myeloma cells from a mouse immunized with a phosphonate transition-state analog for cocaine hydrolysis resulted in the procurement of clones that bound the hapten and catalyzed cocaine ester hydrolysis. Three clones were taken forward for detailed kinetic analysis as catalysts of ()-cocaine hydrolysis. Because ecgonine was the prominent product formed, a stable acyl intermediate, if one exists, does not appear to accumulate. The production of ecgonine also suggests that mAb-catalyzed turnover of the substrate occurs and that mAb does not simply provide stoichiometric binding and/or hydrolysis. The K_m values observed are similar to those previously observed for anti-cocaine catalytic antibodies (Landry et al., 1993), but because of the relatively high K_m values described herein, the catalytic efficiencies for mAb 5 and 12 were low. The k_cat values shown herein are significantly greater than those observed previously. However, the rate acceleration (k_cat/k_o) is similar, suggesting that the rate of spontaneous hydrolysis of cocaine measured in our incubation system is greater than the rate previously reported (Landry et al., 1993).

Animal studies have shown that immunization with antibodies selective for cocaine (Fox et al., 1996) or active immunization with a cocaine immunogen (Carrera et al., 1995) can suppress the psychoactive effects of cocaine. In addition, anti-cocaine catalytic antibodies have recently shown promise as a useful cocaine detoxication catalyst in vivo. The administration of BuChE has also been used effectively in animals to prevent toxicity associated with the administration of ()-cocaine. The procurement of catalytic antibodies described herein that completely hydrolyze cocaine and do so with hydrolytic activity approaching that of human BuChE suggests that it may be possible to elaborate catalytic antibodies that could be useful therapeutically. The administration of a humanized immunotherapeutic agent in conjunction with appropriate relapse-prevention intervention by a trained physician may promote the cessation of cocaine abuse. Thus, a maintenance medication to facilitate abstinence may be a clinical application for use of an anti-cocaine catalytic antibody.