An Improved Method for the Microscale Preparation and

The influence of the activated hapten/BSA molar ratio on the coupling efficiency was investigated by the direct method within the range 10:1 to 250:1...
0 downloads 0 Views 142KB Size
Download PDF
Bioconjugate Chem. 1999, 10, 1143−1149

1143

An Improved Method for the Microscale Preparation and Characterization of Hapten-Protein Conjugates: The Use of Cholesterol as a Model for Nonchromophore Hydroxylated Haptens Je´roˆme Naar,† Philippe Branaa,† Mireille Chinain,† and Serge Pauillac*,†,‡ Unite´ d’Oce´anographie Me´dicale, Institut de Recherches Me´dicales Louis Malarde´, P.O Box 30, Papeete, Tahiti, French Polynesia. Received April 12, 1999; Revised Manuscript Received August 3, 1999

A minute amount (0.446 µmol) of cholesterol (Chol) was converted into an hemisuccinate derivative (Chol HS) using an excess of succinic anhydride. The optimal conditions for synthesis of Chol HS were explored by checkerboard experiments in which various succinic anhydride/Chol molar ratios ranging from 5:1 to 30:1 were assayed over a wide temperature range (50-85 °C) and for various incubation times (3-8 h). Total conversion was obtained at the higher reagent ratios, temperatures, and incubation times. Subsequently, this carboxylic derivative was first covalently linked to bovine serum albumin (BSA) then to various proteins (casein, ovalbumin, and hemocyanins) or to a synthetic homopolymer (poly-DL-Lysine) via a modified version of the mixed anhydride method of Erlanger, performed in a reversed micellar medium. The assessment of the number of haptenic groups per mole of BSA (epitope density) was achieved chromatographically by two methods according to a Chol standard curve established at 207 nm with linearity in the range 0-50 µg. These procedures involving an alkaline hydrolysis of a sample of either the conjugate (direct method) or the unreacted Chol HS (indirect method) yielded an acceptable level of agreement and concordant results in all cases. The influence of the activated hapten/BSA molar ratio on the coupling efficiency was investigated by the direct method within the range 10:1 to 250:1. Using the optimal conditions determined for Chol HS synthesis (a molar reagent ratio of 30:1 with incubation at 65 °C for 6 h) and for BSA haptenation (a 100-fold molar excess of activated hapten, with a carrier stock concentration of 5 mg/mL), epitope density of the conjugates lied between 23 and 27. By reacting the same amount of activated hapten (∼216 µg) with identical amounts of various carriers (300 µg), conjugation efficiency was found similar on a microgram of Chol bound per milligram of carrier basis. This simple and reproducible conjugation and analysis procedures should provide a general method applicable to poorly available and weakly immunogenic haptens bearing hydroxyl groups such as polyether-type marine toxins.

INTRODUCTION

Despite occasional reports that conjugates of cholesterol (Chol)1 can be immunogenic (Bailey et al., 1964; Klopstock et al., 1964; Hara et al., 1979), Chol has long been considered as a poorly immunogenic hapten due to its ubiquitous distribution in the animal kingdom and important biological role. Nevertheless, specific antibodies have been generated using Chol-loaded liposomes containing lipid A as an adjuvant (Swartz et al., 1988). More recently, i.p. injections of silicone oil also have been found to induce specific antibodies through recruitment and adsorption of Chol at the injection site and potent immunostimulant properties (Alving et al., 1996). From all of these previous studies it can be concluded that * To whom correspondence should be addressed, Institut de Recherches Me´dicales Louis Malarde´. Phone: +689 41 64 69. Fax: +689 43 15 90. E-mail: [email protected]. † Unite ´ d’Oce´anographie Me´dicale. ‡ Unite ´ d’Immunocytochimie, CNRS URA 359, De´partement d'Immunologie, Institut Pasteur, Paris, France. 1Abbreviations: AOT, aerosol OT or sodium bis(2-ethylhexyl) sulfosuccinate; BSA, bovine serum albumin; CAS, casein; Chol, cholesterol; Chol HS, cholesterol hemisuccinate; HPLC, highperformance liquid chromatography; KLH, keyhole limpet hemocyanin; LH, limulus polyphemus hemocyanin; OVA, egg albumin; PA, peak area; PBS, phosphate buffered saline; TLC thin-layer chromatography on silica gel; PL, poly-DL-lysine (MW 30000-70000); RT, retention time.

production of antibodies to Chol may be readily achieved by utilizing a powerful adjuvant and/or a suitable Cholcarrier conjugate to help antigen presentation and recognition by the immune system. With the aim of developing new antibodies against lipophilic hydroxylated haptens such as polyether-type marine toxins [see, for review, Yasumoto and Murata (1993)], we have used Chol as a molecular model to investigate an improved method for the microscale preparation and characterization of hapten-protein conjugates. This choice was made considering Chol is a biologically relevant lipid hapten, is commercially available and inexpensive, possesses a unique secondary hydroxyl group, and has no useful absorption band suitable for detection, except the common absorption peak around 210 nm exhibited by most organic compounds. In a preceding report (Pauillac et al., 1998), we have described a miniaturized procedure for the conjugation of lipophilic carboxylic haptens to carrier molecules and the production of specific antibodies. The key processes in this synthesis were an effective adaptation of the mixed anhydride method (Erlanger et al., 1957) and some relevant modifications of the reversed micellar system (Kabanov et al. 1989; Yatsimirskaya et al., 1993) for miniaturization purpose and high haptenization yield. In this previous study conducted with chromophoric carboxylic haptens, conjugate analysis could be readily performed spectrophotometrically.

10.1021/bc990042g CCC: $18.00 © 1999 American Chemical Society Published on Web 10/08/1999

1144 Bioconjugate Chem., Vol. 10, No. 6, 1999

Naar et al.

Figure 1. Chemical coupling of cholesterol to BSA in the reversed micellar medium.

Because Chol is a steroid with unique secondary hydroxyl group, the present paper deals first with the microscale preparation of a cholesterol hemisuccinate derivative (Chol HS) and its subsequent conjugation to carrier molecules in the reversed micellar medium, with emphasis on the elucidation of optimal conditions for both synthetic steps. Second, considering previous demonstrations that both the nature of the carrier protein (Fasciglione et al., 1996) and the epitope density of the conjugates (Klaus and Cross, 1974; Erlanger, 1980; Cooper et al., 1993, 1994) can greatly influence the induction of specific antibodies (amount, class, and affinity), the conjugation efficiency to various carrier proteins and a synthetic poly-amino acid has also been chromatographically evaluated. Conjugates immunogenicity has not been tested in this study. MATERIALS AND METHODS

All chemical reagents, unless otherwise stated, were purchased from Sigma Chemicals Co (St. Louis, MO). High quality grade solvents from Prolabo (France) were dried according to standard procedures. Thin layer chromatograms (TLC) were run on silica gel 60 precoated aluminum foils from Merck (Darmstadt, Germany). Synthesis and Purification of Cholesterol Hemisuccinate. The optimal conditions for synthesis of Chol HS were explored by checkerboard experiments in which various succinic anhydride/cholesterol molar ratios ranging from 5:1 to 30:1 were assayed over a wide temperature range (50-85 °C) for various incubation times (3-8 h). All reactions proceeding in a sealed vial, the most efficient condition, are described hereafter. A 30-fold molar excess of succinic anhydride solubilized in 50 µL of anhydrous pyridine was added to 172 µg (0.446 µmol) of crystalline cholesterol in the internal cone

of a 1 mL small reaction vial (Pierce, Rockford, IL). The vial was sealed and heated for 6 h at 65 °C, and the solvent evaporated under a stream of nitrogen. Finally, the residue was redissolved in 200 µL of CH2Cl2/MeOH (99:1) and succinylation was checked by TLC using CH2Cl2/MeOH (95:5) as developing solvent and 30% aqueous sulfuric acid as spray reagent. By heating, Chol and its derivative (Chol HS) appeared as pink spots with Rf values of 0.57 and 0.32, respectively. Chol HS was purified by rapid chromatography technique on a Sep-Pak Plus silica gel cartridge (WatersMillipore, Milford, MA) fitted with a 10 mL Luer-tipped syringe and conditioned with 5 mL of CH2Cl2. The 200 µL reaction mixture was loaded onto the cartridge, the vial was rinsed with 100 µL of CH2Cl2/MeOH (99:1) which was let to pass through the cartridge, and 5 mL of CH2Cl2 was added to induce sample stacking then elution was performed using a stepwise gradient beginning with 10 mL CH2Cl2-MeOH (90:10) and ending with 5 mL MeOH. The flow rate was maintained around 3 mL/min and 1 mL fractions were collected. Chol HS eluted in tubes 2-4 (10% MeOH) with a yield >95% whereas succinic anhydride was found in the pure MeOH fractions (tubes 1214). Therefore, subsequent calculations were made on the basis of full recovery. Preparation of Cholesterol-BSA Conjugates. Conjugates were prepared according to a modified version of the mixed anhydride coupling method of Erlanger et al. (1957) performed in a reversed micellar medium (Pauillac et al., 1998). The synthetic steps are outlined in Figure 1. Following the rapid chromatographic procedure, fractions containing Chol HS were combined and dried under nitrogen atmosphere in a 1 mL small reaction vial. To this dry derivative, a 10-fold molar excess of tributyl-

Technical Notes

amine (Vm ) 254.94 µL/mmol) and isobutyl chlorocarbonate (Vm ) 131.325 µL/mmol) were added as 1/12th dilutions in dry peroxide-free dioxane to provide a final volume of approximately 20 µL of reaction mixture. The tightly sealed vial was placed onto a mechanical stirrer, and the formation of the mixed anhydride was allowed to proceed for 30 min at room temperature. The reaction was monitored by TLC as described above. Sulfuric acid spraying and heating of the chromatogram showed conversion of Chol HS (Rf ) 0.32) to the mixed anhydride (Rf ) 0.96) to an extent greater than 95%. The conjugation of activated Chol HS to carriers, was carried out in the system of reversed micelles of aerosol OT [AOT or sodium bis(2-ethylhexyl) sulfosuccinate] in heptane (Kabanov et al. 1989; Yatsimirskaya et al., 1993) with relevant modifications for miniaturization purpose and high haptenization yield (Pauillac et al., 1998). Six carriers were used throughout: bovine serum albumin (BSA), ovalbumin (OVA), casein (CAS), poly-DL-Lysine (PL, average mol wt 77000), keyhole limpet, and Limulus polyphemus hemocyanins (KLH and LH, respectively). During these experiments various activated hapten/ carrier molar ratios were assayed in the range 10-250 using commercial Chol HS. A typical experiment performed at room temperature with two BSA concentrations is described. Four volumes of a 0.5 M solution of AOT in heptane were added to 1 vol of BSA at 5 or 7.5 mg/mL in 0.1 M borate buffer, pH 10.2, under vigorous shaking. The formation of reversed micelles was achieved when the solution turned completely clear. Immediately after, volumes of these BSA micellar solutions corresponding to 300 µg (0.004 41 µmol) were poured into the vials containing the activated hapten and the mixtures were again vigorously shaken, providing for each reaction a hapten/protein molar ratio around 100:1. After a slight and transient turbidity was observed, the reactions were allowed to proceed overnight at room temperature. Control experiments run in parallel, consisting in BSA alone and BSA mixed with nonactivated Chol HS, were incubated in the reversed micellar medium under identical conditions. All proteins in the vials (assay and control) were precipitated 20 min at -20 °C in acetone (3 vol), and the whole content was transferred into 5 mL of borosilicate glass test tubes for centrifugation (6000g at +4 °C for 15 min). The resultants pellets were redissolved in PBS (250 µL at +4 °C), precipitated again, then thoroughly washed three times with cold acetone (-20 °C) to remove unbound Chol HS. The precipitates were generally resuspended in 1.2 mL of distilled water, filter sterilized, dispensed into 6 aliquots of 200 µL, freeze-dried overnight and stored at -20 °C until use. However, depending upon expected conjugation efficiency, either the whole conjugate (300 µg in 300 µL of distilled water) or a 200 µL aliquot (50 µg) and the pooled acetone supernatants were saved for analysis (see below). Other fully characterized conjugates of Chol with OVA, CAS, PL, KLH, and LH were also obtained according to this procedure: 1 mg/ mL of micellar carrier solutions were prepared and incubated with activated or nonactivated hapten under identical conditions as for BSA. Characterization of Cholesterol-BSA Conjugates. Carrier haptenization was assessed chromatographically by two methods, both involving an alkaline hydrolysis of a sample of either the conjugate (direct method) or the unreacted Chol HS in the acetone phase (indirect method). Chol qualitative and quantitative analyses were performed using an HPLC system (Kon-

Bioconjugate Chem., Vol. 10, No. 6, 1999 1145

tron Instrument, France) equipped with a UV monitor set at 207 nm and a chromatointegrator for calculation of peak areas (PA). The column was a 4.6 mm id × 150 mm Develosil ODS-HG-5 reversed-phase (Normora Chemical, Japan), and separation was carried out under isocratic conditions with pure MeOH at a flow rate of 1 mL/ min. Chol content of each sample was determined according to a calibration curve with linear regression in the range 0-50 µg (150 µL/injection), obtained by serial dilutions of a stock solution of crystalline Chol in CH2Cl2/MeOH (1:99). After molar conversion, the epitope density of the conjugates (n) was expressed by the hapten/ carrier ratio of the samples. For these calculations, according to previous results (Pauillac et al., 1998) the recovery of the conjugates was considered total. Direct Method. A 200 µL aqueous sample containing 50 µg of Chol-BSA conjugate or the whole conjugate (300 µg in 300 µL distilled water) was mixed with 3 vol of 28% aqueous ammonia, and the mixture was allowed to stand overnight at room temperature. Afterward, the hydrolysis mixture was freeze-dried and the residue was extracted three times with 1 mL of CH2Cl2/anhydrous MeOH (20: 80). The pooled organic fractions were evaporated under a stream of nitrogen, and the residue was redissolved in 150 µL of CH2Cl2/MeOH (1:99) and analyzed by HPLC as described above. Indirect Method. The pooled acetone supernatant recovered from conjugate precipitation was evaporated to dryness, and the residue redissolved in 200 µL of CH2Cl2. Unreacted Chol HS was purified on a Sep-Pak Plus silica gel cartridge as described above. In this case, elution was performed using a stepwise gradient beginning with 10 mL of CH2Cl2/MeOH (97.5:2.5) and ending with 5 mL of MeOH. Again the flow rate was maintained around 3 mL/min and 1 mL fractions were collected. Chol HS eluted first (tubes 3-8) whereas AOT was found in the pure MeOH fractions (tubes 12-14). Chol HS containing fractions were combined, evaporated to dryness and redissolved in 200 µL of CH2Cl2/MeOH (5:95). Alkaline hydrolysis, freeze-drying, extraction, and solvent evaporation procedures were performed as described for the direct method. Finally, the residue was redissolved in 400 µL of CH2Cl2/MeOH (1:99), and a 100 µL sample was subjected to HPLC analysis as described above. RESULTS

Chol-carrier conjugates were prepared in a reversed micellar medium according to the procedure described in Materials and Methods (Figure 1). Optimal conditions for their preparation were established at each step of the two stages involved. Their recovery and substitution ratio were carefully checked by incorporating controls in which carriers have been incubated in the same conditions with nonactivated hapten. The epitope density of the conjugates (n) was expressed by their hapten/carrier molar ratio considering a total recovery of the conjugates. Chromatographic Quantitation of Cholesterol. Chol is a monohydroxylated, lipid-soluble compound with no characteristic peak absorption with respect to proteins, therefore a reversed-phase HPLC method employing UV detection at 207 nm has been set up for its quantitation (150 µL/sample run). Figure 2A illustrates a chromatogram obtained by injecting 25 µg of Chol, which eluted with a retention time (RT) varying from 9.6 to 10.13 min depending upon the ambient temperature. To assess the linearity of this method, a calibration curve with a working range of 0-50 µg was obtained by plotting peak areas (PA) expressed in millivolt minutes versus

1146 Bioconjugate Chem., Vol. 10, No. 6, 1999

Naar et al.

Figure 3. Determination of the optimal conditions for Chol HS synthesis. Effect of two temperatures (65 and 80 °C) and influence of the succinic anhydride/Chol molar ratio (5:1 to 30: 1).

Figure 2. Chromatographic quantitation of cholesterol. chromatogram on a Develosil ODS-HG-5 reverse-phase column. Mobile phase 100% MeOH; flow rate 1 mL/min; detection UV at 207 nm; sample 25 µg of Chol applied in 150 µL of CH2Cl2/ MeOH (1:99); Retention time, RT ) 10.13 min; peak area, PA ) 47.6 mV × min. Calibration curve. Increasing amounts of cholesterol solubilized in 150 µL of CH2Cl2/MeOH (1:99) were subjected to HPLC analysis. Corresponding peak areas (PA) are plotted versus amounts of injected sample. Values are the mean of two independent experiments (error bars are omitted because of symbol overlapping).

corresponding Chol amounts in micrograms (Figure 2B). All reported values are the means of two series of data obtained separately. The adequacy and high accuracy of this dosage is deduced from the value of the correlation coefficient which was very close to 1. Optimization of the Succinylation Process. To optimize the succinylation process, the reaction was performed for 3-8 h at different temperatures (50-85 °C), using various succinic anhydride/Chol molar ratios (5:1 to 30:1). Regarding incubation time, in most cases it was found that 6 h was a good compromise because the reaction no longer progressed after this period. The reactions were monitored by TLC and accurately quantified by HPLC using an aliquot or the whole reaction mixture to evaluate remaining Chol amount. Appropriate factors were applied to restore the true value in the original reaction and a subtraction was made on a molar

Figure 4. Outline of the two methods for the analysis of CholBSA conjugates. x and y are the number of moles of reagents and products; n (normal letter or subscript) is the number of haptenic residues per mole of BSA (epitope density). Calculations are made assuming a total recovery of the protein. The direct method determines bound Chol in the conjugate whereas the indirect method evaluates unreacted Chol in the acetone supernatant (other details in the text).

basis to estimate the amount of Chol HS synthesized. Reaction yield expressed as percentage of the molar ratio of Chol HS per total Chol are reported in Figure 3 for two temperatures (65 and 80 °C). These results clearly indicate that the yield is dependent upon the reagent ratio at both temperatures. However, at 80 °C, maximal yield (100%) can be reached with a 2-fold lower reagent ratio (15:1) compared to 30:1 necessary at 65 °C, thus highlighting also the temperature dependence of the reaction. In addition, temperatures below 60 °C were not favorable to provide yields greater than 40%, and the

Technical Notes

Bioconjugate Chem., Vol. 10, No. 6, 1999 1147

Figure 5. Chromatographic determination of the epitope density of Chol-BSA conjugates. BSA (300 µg) was reacted with either activated or nonactivated Chol HS in a 100-fold molar excess of hapten. Bound and unreacted Chol were evaluated chromatographically from an aliquot of the conjugate and the acetone supernatant, respectively (see Figure 4). Direct method (one-sixth of the original sample): (A) Chol-BSA conjugate; (B) control. Indirect method (one-fourth of the original sample): (C) Chol-BSA conjugate; (D) control.

influence of the reagent ratio was minimized in this temperature range (data not shown). Consequently, optimal conditions for this reaction were set at 65 °C for 6 h with a 30:1 molar reagent ratio. Analysis of the Conjugates. Haptenation of the chemically activated Chol derivative (Chol HS) by the modified mixed anhydride method in the reversed micellar medium (Pauillac et al., 1998) yielded densely coated BSA carrier protein as revealed by chromatographic determination of bound Chol (direct method) or unreacted Chol HS (indirect method). These two methods are outlined in Figure 4 with the rationale for the determination of conjugates epitope density (n). Figure 5 illustrates these methods by chromatograms obtained with the same Chol-BSA conjugate and its control experiment (300 µg of BSA incubated under identical conditions

with either 216.48 µg of activated or nonactivated Chol HS, respectively). Using the direct method, the PA obtained for the conjugate sample (Figure 5A) corresponded to 7.26 µg of bound Chol (43.6 µg or 0.1127 µmol in the original reaction). Hence, n was found to be 26. It is noteworthy that no Chol was detected in control samples (Figure 5B) even up to a 500-fold molar excess of nonactivated Chol HS (data not shown). By the indirect method, the PA obtained from the conjugate sample (Figure 5C) corresponds to 29.27 µg of unreacted Chol (117.08 µg in the original reaction). Considering the PA obtained from control sample (Figure 5D) equivalent to 40.93 µg of unreacted Chol, the maximum amount that could be recovered in the original reaction is 163.72 µg. Therefore, a good estimation of

1148 Bioconjugate Chem., Vol. 10, No. 6, 1999

Naar et al.

direct method (Table 1). For comparison purpose, the same amount of carrier (300 µg) was reacted with either activated or nonactivated Chol HS (∼216 µg), each value represents the average determination from three independent reactions, which varied by no more than 8%. Although n lies between 9 and 25, these results clearly show that conjugation efficiency is identical on a microgram of Chol bound per milligram of carrier basis. DISCUSSION

Figure 6. Optimization of cholesterol coupling to BSA. Dependency on hapten/protein molar ratio and influence of the concentration of the BSA stock solution (5 and 7.5 mg/mL). BSA (300 µg) was reacted with activated Chol HS in various molar excess (0-250). After purification, conjugates were solubilized in distilled water then analyzed for their epitope density according to the direct method. Table 1. Determination of the Epitope Density of Various Chol-Carrier Conjugates conjugation efficiency carrier

mol wt

µg of Chol/ mg of carrier

BSA OVA KLH LH PL Casein

68000 45000 NDa NDa 70000 23000

142 125 112 139 119 151

nb 25 14 ND ND 21 9

a Not determined for the nonhomogeneous hemocyanins. Epitope density, each value represents the average determination from three independent reactions, which varied by no more than 8% (direct method). b

bound Chol in the conjugate is 46.64 µg (163.72-117.08), which gave n ) 27. This value is very similar to that determined by the direct method (26). The adequacy and concordance of the direct and indirect method was further assessed by dual analyses of three BSA conjugates obtained using various hapten/carrier molar ratios. In all cases, the indirect method exhibited no more than 10% increase in n compared to the direct method and increase of bound Chol rigorously paralleled the decrease of unreacted Chol (data not shown). Optimization of the Conjugates Preparation. Experiences aiming to investigate the influence of carrier concentration were designed using two BSA stock solutions (5 and 7.5 mg/mL) keeping the same hapten/carrier molar ratio. For convenience, these experiments were conducted using a commercial Chol HS activated at three concentrations (10, 20, and 40 mg/mL in dioxane), to keep approximately the same volume for the reaction mixtures. As shown in Figure 6, at the two stock BSA concentrations tested, the coupling rate is strongly dependent on the hapten/protein molar ratio up to a value of 100:1 but a plateau is eventually reached, indicating that conjugation no longer occurred for greater ratio values. Within the investigated reagent ratio range, the overall coupling efficiency of the assay using 5 mg/mL over the one using 7.5 mg/mL is clearly demonstrated. Maximal binding of Chol to BSA (n ≈ 25) occurs using a 100-fold molar excess activated Chol HS with a 5 mg/ mL protein stock concentration. Taking into account the above results, the optimized procedure was applied to the other carrier molecules and n was determined using the

To establish immunological reagents for the quantitation of haptens, there is a strict need for conjugation to appropriate carriers for preparing satisfactory material for immunization and antibody screening. Dealing with mice immunization, various immunogenic carriers have been successfully used; among them, BSA was the most widely assayed for many reasons. BSA is inexpensive and readily available, contains 60 amino groups of which ∼35 are exposed to the surface of the molecule under physiological conditions, in addition, this protein proved to be heterologous enough to provoke adequate mice immune response in most cases. A simple conjugation reaction consists of a mixture of the protein, the hapten, and a coupling reagent in such a way that the yield of conjugate is maximal. Nevertheless, conjugation must not alter the antigenic properties of the hapten. To this goal, other practical aspects appear uppermost relevant, as the purity and concentration of the components, products yield, and finally conjugate recovery. Since all these critical factors can individually have a profound effect on the success of the procedure, optimal coupling conditions must be carefully checked. Haptens bearing nonreadily reactive functional groups must be converted into more reactive derivatives using adequate chemical reagents (Erlanger, 1973, 1980; Wong, 1993). When hydroxyl groups are available, hemisuccinate derivatives are generally produced in the presence of an excess of succinic anhydride in pyridine, linkage to the amino groups of the carrier is subsequently accomplished through a reagent-mediated carboxyl activation. This process has the advantage of introducing a four-carbon spacer arm which exposes the hapten from the surface of the carrier, thus favoring its presentation and recognition by the immune system. However, it has been previously demonstrated that increasing the length of the chemical spacer can severely affect the production of hapten-specific antibodies (Beckett et al., 1978) since, as an artifact, it can also be part of the epitope recognized by the induced antibodies (Weyrer et al., 1990). In this study, minute amount (0.446 µmol) of Chol was first entirely converted into an hemisuccinate (Chol HS), then the recently described conjugation technique proceeding in a reversed micellar medium (Pauillac et al., 1998) was successfully adapted to this carboxylic derivative as a one-pot synthesis method. Because the degree of conjugation or epitope density is an essential attribute of a conjugate either as an immunogen (Klaus and Cross, 1974; Erlanger, 1980; Cooper et al., 1993, 1994) or a target antigen (Vyjayanthi et al., 1995), its assessment must be correctly addressed. For these reasons investigations were previously aimed (i) at the study of complete conjugate recovery (Pauillac et al., 1998) and (ii) at accurate determination of Chol on a reversed-phase HPLC column according to a calibration curve. The assessment of the Chol content of the conjugates either by the direct or indirect method gave consistent results. In the direct method, where covalently linked Chol is released by alkaline hydrolysis, corre-

Technical Notes

sponding PA was linearly related to sample contents and a simple multiplication by 6 gave the Chol content in the whole conjugates. In the indirect method which measures unbound Chol, a 4-fold factor was applied to restore the true value in the whole supernatant and a subtraction from the total unreacted Chol recovered from control experiments was made to estimate bound Chol in the whole conjugates. By reacting the same amount of activated hapten (∼216 µg) with identical amounts of various carriers (300 µg), epitope density of the conjugates lied between 9 and 25, but on a microgram οf Chol bound per milligram of carrier basis, conjugation efficiency was identical. These values were found to fall within the optimal range to ensure hapten-specific antibody production (Erlanger, 1980). As a matter of fact, using this procedure, PbTx-3 (400 µg; 0.446 µmol)sa member of the brevetoxin polyethertype toxins (Baden, 1989)swas totally converted into an hemisuccinate derivative (PbTx-3 HS) and covalently linked to BSA and OVA. Following mice immunization with (PbTx-3)18-BSA conjugate, highly specific antibodies have been produced as revealed by ELISA on (PbTx3)10-OVA-coated plates (Naar et al., unpublished results). In this study dealing with Chol, a hydroxylated lipidsoluble hapten, four major improvements have been achieved in the process of conjugates preparation and analysis: (1) Chol HS was produced with a 100% yield with as little as 172 µg of starting material; (2) the mixed anhydride method of Erlanger was successfully adapted to such minute hapten quantities with a moderate hapten/carrier molar ratio of 100:1; (3) quantitation of both bound Chol and unreacted Chol HS was accurately performed chromatographically; (4) no radioactive probe was required to monitor antigen preparation. In conclusion, this simple conjugation procedure appears most valuable for poorly available and weakly immunogen haptens bearing hydroxyl groups such as polyether-type marine toxins through the specific induction of high-affinity antibodies. ACKNOWLEDGMENT

The authors thank Philippe Cruchet for excellent technical assistancewith HPLC analyses. This work was supported by grants from the Government of French Polynesia (Tahiti) and Pasteur Institute (Paris, France). LITERATURE CITED (1) Alving, C. R., Wassef, N. M., and Potter, M. (1996) Antibodies to cholesterol: biological implications of antibodies to lipids. Curr. Top. Microbiol. Immunol. 210, 85-92. (2) Baden, D. G. (1989) Brevetoxins: unique polyether dinoflagellate toxins. FASEB J. 3, 1807-1817. (3) Bailey, J. M., Bright, R., and Tomar, R. (1964) Immunization with a synthetic cholesterol-ester antigen and induced atherosclerosis in rabbits. Nature 201, 407-408. (4) Beckett, G. J., Hunter, W. M., and Percy-Robb, I. W. (1978) Investigation into the choice of immunogens, ligand, antiserum and assay conditions for radioimmunoassay of conjugated cholic acid. J. Clin. Acta 88, 257-266. (5) Cooper, L. J., Shikhman, A. R., Glass, D. D., Kangisser, D., Cunningham, M. W., and Greenspan, N. S. (1993) Role of

Bioconjugate Chem., Vol. 10, No. 6, 1999 1149 heavy chain constant domains in antibody-antigen interaction. Apparent specificity differences among streptococcal IgG antibodies expressing identical variable domains. J. Immunol. 150 (6), 2231-2242. (6) Cooper, L. J., Robertson, D., Granzow, R., and Greenspan, N. S. (1994) Variable domain-identical antibodies exhibit IgG subclass-related differences in affinity and kinetic constants as determined by surface plasmon resonance. Mol. Immunol. 31 (8), 577-584. (7) Erlanger, B. F. (1973) Principles and methods for the preparation of drug protein conjugates for immunological studies. Pharmacol. Rev. 25, 271-280. (8) Erlanger, B. F. (1980) The preparation of antigenic haptencarrier conjugates: a survey. Methods Enzymol. 70, 85-104. (9) Erlanger, B. F., Borek, F., Beiser, S. M., and Lieberman, S. (1957) Steroid-protein conjugates. I. Preparation and characterization of conjugates of bovine serum albumin with testosterone and with cortisone. J. Biol. Chem. 228, 713727. (10) Fasciglione, G. F., Marini, S., Bannister, J. V., and Giardina, B. (1996) Hapten-carrier interactions and their role in the production of monoclonal antibodies against hydrophobic haptens. Hybridoma 15 (1), 1-9. (11) Hara, I., Muramatsu, T. Fukuda, T., and Sato, J. (1979) Immunochemical properties of phosphatidyl cholesterol and its homologue. Chem. Phys. Lipids 23, 2-12. (12) Kabanov, A. V., Ovcharenko, A. V., Melik-Hubarov, N. S., Bannikov, A. I., Alakhov, V. Yu., Kiselev, V. I., Sveshnikov, P. G., Kiselev, O. I., Levashov, A. V., and Severin, E. S. (1989) Fatty acid acylated antibodies against virus suppress its reproduction in cells. FEBS Lett. 250 (2), 238-240. (13) Klaus, G. G. B., and Cross, A. M. (1974) The influence of epitope density on the immunological properties of haptenprotein conjugates. I. Characteristics of the immune response to hapten-coupled albumen with varying epitope density. Cell. Immunol. 14, 226-241. (14) Klopstock, A., Pinto, M., and Rimon, A. (1964) Antibodies reacting versus steroid haptens. J. Immunol. 92, 515-519. (15) Pauillac, S, Naar, J., Branaa, P., and Chinain, M. (1998) An improved method for the production of antibodies to lipophilic carboxylic hapten using small amount of haptencarrier conjugates. J. Immunol. Methods 220, 105-114. (16) Swartz, G. M., Jr., Gentry, M. K., Amende, L. M., Blanchette-Mackie, E. J., and Alving, C. R. (1988) Antibodies to cholesterol. Proc Natl. Acad. Sci. U.S.A. 85, 1902-1906. (17) Vyjayanthi, V., Capoor, A. K., and Sashidhar, R. B. (1995) Binding characteristics of bovine serum albumin-aflatoxin B1 to polystyrene microtiter plates: importance of hapten to carrier protein molar ratio. Indian J. Exp. Biol. 33 (5), 329332. (18) Weyrer, W., Most, J., Solder, B., Fuchs, D., Wachter, H., and Dierich, M. P. (1990) Production and characterization of a monoclonal antibody against neoptrin. Hybridoma 9 (1), 71-79. (19) Wong, S. S. 1993. Preparation of immunotoxins and other drug conjugates for targeting therapeutics. Chemistry of protein conjugation and cross-linking, Chapter 11, pp 267290, CRC Press, Inc., Boston. (20) Yasumoto, T., and Murata, M. (1993) Marine toxins. Chem. Rev. 93, 1897-1909. (21) Yatsimirskaya, E. A., Gavrilova, E. M., Egorov, A. M., and Levashov, A. V. (1993) Preparation of conjugates of progesterone with bovine serum albumin in the reversed micellar medium. Steroids 58 (11), 547-550.

BC990042G