Quantitative Analysis of Derivatized Proteins Prepared with Pyridyl

Feb 17, 1999 - Drug Targeting Laboratory, College of Pharmacy, SungKyunKwan University, 300 Chonchon-dong, Jangan-ku,. Suwon City 440-746, Korea...
0 downloads 0 Views 74KB Size
306

Bioconjugate Chem. 1999, 10, 306−310

TECHNICAL NOTES Quantitative Analysis of Derivatized Proteins Prepared with Pyridyl Disulfide-Containing Cross-Linkers by High-Performance Liquid Chromatography Dong Hee Na, Byung Ho Woo, and Kang Choon Lee* Drug Targeting Laboratory, College of Pharmacy, SungKyunKwan University, 300 Chonchon-dong, Jangan-ku, Suwon City 440-746, Korea. Received March 11, 1998; Revised Manuscript Received November 30, 1998

Determination of the introduced moieties into derivatized proteins is an essential step in the preparation and quality control of chemically defined immunoconjugates. For the derivatized proteins using pyridyl disulfide-containing cross-linkers such as N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) and 4-succinimidyloxycarbonyl-R-methyl-R-(2-pyridyldithio)toluene (SMPT), the derivatization degree (ratio of pyridyl disulfide moieties to protein) has been traditionally determined by measuring the absorbance of both the derivatized protein and 2-thiopyridone (2-TP) released from the dithiothreitol (DTT) treatment (spectrophotometric method). This method, however, causes several problems including false high and low determinations of the protein and 2-TP, respectively, low selectivity, poor reproducibility, and relatively large amounts of sample consumption. A quantitative determination method of the derivatization ratios using bovine serum albumin derivatized with SPDP and SMPT as the model system has been developed. The concentration of protein and 2-TP released from the DTT treatment of derivatized proteins was determined directly without consideration of different reagents used and their concentrations. The present HPLC method was proved to be better in terms of accuracy, selectivity, and reproducibility with micro sample consumption. Moreover, this HPLC method can be directly applied to all derivatized proteins prepared with pyridyl disulfide-containing cross-linkers.

INTRODUCTION

The majority of immunoconjugates including immunotoxins consisting of monoclonal antibody and toxin have been synthesized according to chemical methods with cleavable heterobifunctional cross-linkers (1-4). A major concern in immunoconjugate research is the synthesis and testing of the heterogeneous mixture without proper separation and characterization of the chemically defined immunoconjugates. Recently, there is increasing interest among pharmaceutical scientists in developing specific and reproducible chemical methods in the preparation of immunotoxins and appropriate controls and validatable assays (5). Determination of the introduced moieties into the protein derivatized with cross-linkers is therefore an essential step in the preparation and quality control of chemically defined immunoconjugates (6-7). The derivatization degree (ratio of pyridyl disulfide moieties to protein) of the derivatized proteins prepared with pyridyl disulfide-containing cross-linkers (8-20) has been traditionally determined by using the spectrophotometric method (SP method), which measures the absorbances of both protein derivatized and 2-thiopyridone (2-TP) released from the derivatized protein by reaction with dithiothreitol (DTT) at 280 and 343 nm, respectively (8). This method, although simple, raises several concerns including false high or low determination of the protein * Author to whom correspondence should be addressed (fax 82-331-290-7724; e-mail [email protected]).

and 2-TP concentration, respectively, low selectivity, and poor reproducibility due to the spectral interference of the introduced moieties (9, 10). In addition, it requires a large amount of sample consumption. Furthermore, this method has not been modified appropriately for all of the derivatized proteins prepared from different cross-linkers. Pyridyl disulfide moieties of the derivatized protein from different cross-linkers, for example, (2-pyridyldithio)propionate (PP) from N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) and R-methyl-R-(2-pyridyldithio)toluene (MPT) from 4-succinimidyloxycarbonyl-R-methylR-(2-pyridyldithio)toluene (SMPT), have their own absorbance at 280 nm (A280) in addition to protein. Therefore, these moieties may exert different effects on the A280 in protein assay (8, 9). In this study, we developed a quantitative determination method of the derivatization ratios of protein modified with pyridyl disulfide-containing heterobifunctional cross-linkers by measuring the protein and 2-TP using size exclusion and C-18 reversed phase HPLC, respectively. This new HPLC method was compared with the SP method using bovine serum albumin (BSA) derivatized with SPDP and SMPT as the model system. MATERIALS AND METHODS

Materials. BSA (fraction V) and DTT were obtained from Sigma (St. Louis, MO). BSA was further purified by using a Sephadex G-100 column, and the purified fraction (MW 66000) was used. SPDP and SMPT were purchased from Pierce (Rockford, IL). 2-Thiopyridone (2-

10.1021/bc980029g CCC: $18.00 © 1999 American Chemical Society Published on Web 02/17/1999

Technical Notes

mercaptopyridine) was obtained from Aldrich (Milwaukee, WI), and HPLC grade methanol was purchased from J. T. Baker (Phillipsburg, NJ). All other materials were of reagent grade. BSA Conjugation with SPDP and SMPT. Each of 0-16 molar excess of SPDP or SMPT solutions (20 mM in dimethyl sulfoxide) was added to 1 mL of BSA (5 × 10-2 mM) in 10 mM phosphate buffer containing 0.15 M NaCl (PBS, pH 7.4). The reaction mixtures were incubated for 30 min at room temperature, and the excess of reagent was then removed by gel filtration on a Sephadex G-25 column (Sigma) equilibrated with PBS. Fractions with an absorbance >1.0 at 280 nm (∼8 mL) were collected and used in the assay. The derivatization ratio was determined by measuring the concentration of both BSA and 2-TP released by DTT treatment of the derivatized BSA using both the SP and HPLC methods described below. SP Method. This method was performed as described previously (8). In brief, BSA concentration was determined by measuring the A280 using a Spectronic 1201 UV-vis spectrophotometer (Milton Roy, Rochester, NY). The amount of pyridyl disulfide moiety (PP or MPT) in the derivatized BSA was determined by measuring the absorbance of 2-TP at 343 nm (A343), which was released by adding 20 µL of 1 M DTT in deionized water to the derivatized BSA solution (980 µL) after 1 h of incubation at room temperature. The molar concentration of 2-TP was calculated using its molar extinction coeffcient E343 ) 8.08 × 103 M-1 cm-1 (21). In the determination of the molar concentration of BSA, an erroneously high BSA concentration due to pyridyl disulfide moieties was corrected by subtracting the A280 contribution of the pyridyl disulfide moieties (2-pyridyl disulfide structure) in the derivatized BSA from the total A280 using the equation

pyridyl disulfide moieties/1 molecule of BSA ) A343 × 40.3/(A280 × 8.08 - A343 × 5.10) (a) The molar extinction coefficient of BSA at 280 nm was measured as E280 ) 4.03 × 104 M-1 cm-1 using purified BSA described under Materials. E280 of pyridyl disulfide moiety (PP or MPT) in the derivatized BSA was adopted as used previously without correction as E280 ) 5.10 × 103 M-1 cm-1 (8, 9, 11, 19, 21). HPLC Method. The concentration of BSA of the derivatized BSA was determined by a size exclusion HPLC using a BIOSEP SEC-S2000 column (300 × 7.8 mm, Phenomenex). PBS (pH 7.4) was used as a mobile phase, and the flow rate was 1 mL/min. Each 20 µL of standard and sample solutions was injected onto the HPLC, and the effluent was detected at 215 nm. The concentration of 2-TP released from the derivatized BSA solution by DTT treatment was determined by a reversed phase HPLC using a LiChrospher 100 RP-18 column (125 × 4.0 mm, 5 µm, Merck, Darmstadt, Germany) at 343 nm. The mobile phase was composed of distilled water with a linear gradient of 0-30% (v/v) methanol for 10 min. The flow rate of the mobile phase and the injection volume were 1 mL/min and 20 µL, respectively. A Gynkotek HPLC system with model 480 pumps, a model GINA50 autosampler, a model UVD340S photodiode array detector and Gynkosoft (Gynkotek, Munich, Germany) was adopted. RESULTS AND DISCUSSION

The important point of the required and ideal characteristics of immunotoxin conjugates is that the cross-

Bioconjugate Chem., Vol. 10, No. 2, 1999 307

linking process must minimize the deactivation and polymerization of both the antibody and toxin. In general, the chemistry of bioconjugation gives rise to various structures, resulting in a heterogeneous population of products that are difficult to separate from each other (5, 22, 23). These products usually consist of a major component containing equimolar amounts of antibody and toxin and several other conjugates with two or more toxin molecules per molecule of antibody. This heterogeneity results from the introduction of varying numbers of derivatized moieties per molecule of antibody (24). Due to the nature of the chemical reaction between the N-succinimidyl or hydrazide groups of the cross-linkers and the primary amino or carbohydrate groups of the antibody, the populations of antibody molecules substituted with none, one, two, three, and more than three pyridyl disulfide groups are obtained and the microenvironment of the protein attached linker will most likely affect the conjugation of antibody with toxin. Thus, it is very important to reproducibly determine the extent of derivatization exactly. To determine the derivatization degrees of modified proteins by cross-linkers, both protein and pyridyl disulfide moieties of derivatized proteins must be assayed precisely. Determination of protein by measuring A280 adopted in the SP method is not strictly quantitative and is fairly adequate for crude protein mixtures, because this assay is based on the strong absorbance of tyrosine, phenylalanine, and tryptophan residues. Different proteins may therefore have widely varying molar extinction coefficients, and the absorbance must be partially corrected for interfering substances in the range of 0.2-2 mg/mL (25). Because all of the pyridyl disulfide moieties of derivatized proteins have their own A280 values, an erroneously high protein concentration is obtained when calculation is done on the basis of absorbance at 280 nm. For the SPDP-derivatized BSA, the additional A280 by the pyridyl disulfide moieties has been calculated using the molar extinction coefficients of the PP moiety in derivatized BSA at 280 nm (E280 ) 5.10 × 103 M-1 cm-1). This E280 value of the PP moiety in the derivatized protein is not a measured value but a calculated value from the absorbance of SPDP or 2,2′-dipyridyl disulfide (2-Py-SS-Py) (8, 21). This means that the SP method must be validated and confirmed by the experimental method. Also, the SP method has not been modified adequately for different derivatized proteins prepared from different cross-linkers. Different pyridyl disulfide moieties of the derivatized proteins are obtained depending upon the cross-linkers used, for example, (2-pyridyldithio)propionate (PP) from SPDP and R-methyl-R-(2-pyridyldithio)toluene (MPT) from SMPT. Therefore, if the concentration of BSA of the SMPT-derivatized BSA is determined in the same manner as in SPDP derivatization, false high protein concentrations may result due to differences in the contribution of A280 between PP from SPDP and MPT from SMPT (9, 11, 19). To assess the effect of the introduced moiety on the A280 value of derivatized BSA, the absorbance of each derivatized BSA prepared from different ratios of crosslinkers to BSA and fractionated by size exclusion HPLC was measured using the SP method. Higher ratios of the cross-linker to BSA resulted in higher A280 (data not shown). In the case of SMPT derivatization, the additional absorbance from pyridyl disulfide moieties introduced was higher than that of SPDP derivatization, resulting from the methyl toluene moiety in addition to the PP moiety. Therefore, specific equations for different

308 Bioconjugate Chem., Vol. 10, No. 2, 1999

Na et al.

Figure 2. Reversed phase HPLC chromatogram of SPDPderivatized BSA treated with DTT at 343 nm and the identification of the peak by photodiode array detection: (Insets) absorbance curves of 2-thiopyridone and DTT.

Figure 1. Size exclusion HPLC chromatograms of SPDPderivatized BSA prepared from 0 to 8.7 molar excess of SPDP: (A) 280 nm detection; (B) 215 nm detection; (a) underivatized BSA; (b-e) SPDP-derivatized BSA (SPDP/BSA molar ratio: b, 1:1; c, 2.6:1; d, 5.0:1; e, 8.7:1).

derivatized proteins must be confirmed to determine the accurate derivatization ratios. In the HPLC method, the detection wavelength was set at 215 nm, which was sensitive to the peptide bonds. To determine the effect of the pyridyl disulfide moiety on the detection wavelengths, the HPLC chromatograms obtained at both wavelengths of 215 and 280 nm were compared (Figure 1). In contrast to the increased absorbance as a result of increased concentration of crosslinkers at 280 nm, no additional absorbance of the introduced moieties into the derivatized protein was observed at 215 nm. This means that no correction is needed to measure the protein concentration regardless of the cross-linkers and reaction conditions used. Therefore, this HPLC method may be applied to all of the derivatized proteins prepared with pyridyl disulfidecontaining cross-linkers. In addition, the detection sensitivity of protein at 215 nm was 13 times higher than that at 280 nm. From the HPLC method, the correlation coefficient of the detector linearity was found to be >0.999 over the concentration range 1.5-15.2 µM in 10 mM PBS (pH 7.4) The introduced moieties, PP from SPDP and MPT from SMPT, have been routinely quantitated by measuring the difference in the absorbance before and after DTT treatment of derivatized proteins at 343 nm using the molar extinction coefficient E343 ) 8.08 × 103 M-1 cm-1 without separation of 2-TP from the reaction solution. However, it is laborious to obtain reproducible data because the value of A343 of 2-TP is variable depending upon the reaction conditions, for example, pH, reagents used and

their concentrations, and proteins derivatized (7, 8, 21, 26). Moreover, this method was found to be unreliable because the DTT treatment changes the absorbance at 280 nm, resulting in small baseline changes at 343 nm that are of an order of magnitude similar to those expected for the released 2-TP (10). In fact, the HPLC chromatogram of the reaction mixture shows the presence of the DTT peak detected at 343 nm (Figure 2), indicating that DTT used as a reducing agent might cause falsely high concentrations of 2-TP, especially for proteins with a low degree of derivatization and high concentrations of DTT under the SP method. As an alternative method, the measurement of sulfhydryl groups in the DTT-treated derivatized protein using Ellman’s reagent has its own limitation, due to possible damage of endogenous disulfide bonds and unfavorable effects of the reaction conditions (8, 10). In the present HPLC method, the concentration of 2-TP released from the DTT treatment of SPDP- or SMPTderivatized BSA was determined with photodiode array detection at 343 nm. 2-TP was well resolved as shown in Figure 2. The linearity was observed over the range of 2-80 µM in 10 mM PBS (pH 7.4), with the correlation coefficient being >0.999. Using this HPLC method, 2-TP released from other proteins prepared with various pyridyl disulfide-containing cross-linkers may be determined quantitatively without any interference from the reagents or reaction conditions used. The derivatization ratios determined according to the SP and HPLC methods were compared (Table 1). The ratios of the SPDP-derivatized BSA were comparable between these two methods. However, there was a significant difference in the derivatization ratios for the SMPT-derivatized BSA. At low derivatization ratios