Reversible Surface Thiol Immobilization of Carboxyl Group Containing

Jason J. Davis , Jan Tkac , Rachel Humphreys , Anthony T. Buxton , Tracy A. Lee and Paul Ko Ferrigno ... Barry Byrne , Edwina Stack , Richard O'Kenned...
0 downloads 0 Views 68KB Size
Download PDF
188

Bioconjugate Chem. 2002, 13, 188−193

Reversible Surface Thiol Immobilization of Carboxyl Group Containing Haptens to a BIAcore Biosensor Chip Enabling Repeated Usage of a Single Sensor Surface Ulrich Schlecht,† Yoko Nomura, Till Bachmann,† and Isao Karube* Research Center for Advanced Science and Technology, University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8904, Japan. Received February 1, 2001

We describe a reversible immobilization method for carboxyl group containing haptens that makes the repeated usage of a BIAcore biosensor chip possible. Haptens which are immobilized according to the surface thiol method can be removed completely from the sensor surface again by a reducing step. In the first part of our study, analogues of the herbicides 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid were immobilized in succession to a biosensor surface of a BIAcore surface plasmon resonance instrument according to the thiol coupling method. Direct kinetic analysis of these ligands to a polyclonal anti-2,4-dichlorophenoxyacetic acid antibody were performed using these biosensor surfaces. In the second part of the study, different amounts of 2,4-dichlorophenoxyacetic acid were sequentially immobilized onto the same biosensor surface in order to generate a calibration plot for 2,4-dichlorophenoxyacetic acid. Using this plot, the quantitative detection of the herbicide down to a concentration of 0.1 µg/mL, the maximum admissible concentration of pesticides in drinking water, is possible.

INTRODUCTION

2,4-Dichlorophenoxyacetic acid (2,4-D) and 2,4,5trichlorophenoxyacetic acid (2,4,5-T) are common herbicides used for the control of weeds that can potentially contaminate groundwater and drinking water supplies. It has been reported that exposure of humans to 2,4-D produces acute congestion of all organs and degenerative changes in ganglionic cells of the central nervous system (1). 2,4-D is also a carcinogen, causing soft-tissue sarcoma in humans (2) and malignant lymphoma in dogs (3). Toxicity of 2,4,5-T is mainly based on its contamination by dioxins, such as TCDD and PCDD, during its production (4). In the European Community (EC) countries, the maximum admissible concentration for pesticides in drinking water is 0.1 µg/mL for any substance and 0.5 µg/mL for the combined total of pesticides (5). Rapid, easy, and economical methods for the detection of these pesticides, most of which are low molecular weight ligands (haptens), are of great importance. In the past few years, different approaches to the detection of pesticides have been described, for example, capillary gas chromatography with ion trap detection (6), piezoelectric biosensors (7), fiber-optic immunosensors (8), and amperometric cholinesterase biosensors (9). Beyond these techniques, the development of the surface plasmon resonance (SPR) technology has opened up a new and elegant way to detect haptens quantitatively. For instance, assays for the detection of the herbicides atrazine (10) and simazine (11) on the basis of the competition principle have been established. In the present study, we have detected 2,4-D for the first time using the BIAcore SPR technology and, in addition, we describe a new method of detection. * To whom correspondence should be sent. Tel: +81-354525220. Fax: +81-3-5452-5227. E-mail: [email protected]. u-tokyo.ac.jp. † Institute for Technical Biochemistry, University of Stuttgart.

The specific chemistry of 2-(2-pyridinyldithio)ethaneamine hydrochloride (PDEA) that is normally used to perform site-controlled coupling of proteins to biosensor surfaces in the BIAcore system (12, 13) inspired us to develop a reusable biosensor. With the described PDEAdependent thiol coupling method (14), haptens containing a carboxyl group (i.e., 2,4-D) can be immobilized to a biosensor surface and removed again by a reducing step with a dithiothreitol solution. Using one biosensor chip, a calibration plot for the model compound 2,4-D has been generated that makes the quantitative detection of this herbicide down to a concentration of 0.1 µg/mL, the maximum admissible concentration of this pesticide in drinking water, possible. The BIAcore SPR technology is also a powerful tool for binding affinity studies of haptens to related antibodies. The most common approach involves a competitive study in which an immobilized ligand and a soluble ligand free in solution compete for binding sites on a single analyte (solution competition) (15, 16). The advantage of the solution competition experiment is that a single biosensor surface can be utilized for the determination of equilibrium association and dissociation constants (KA and KD, respectively) for a series of related analogues. Unfortunately, the method does not provide apparent association and dissociation rate constants (ka and kd, respectively). These constants characterize the association and the dissociation phases of antibodies binding to ligands in particular. They are often more reliable than the equilibrium constants. One way to determine these parameters is by direct kinetic analysis. The analogues are immobilized to the sensor surface, normally according to the amino-coupling method, and the antibody is injected over the surface at different concentrations (17). The disadvantage of this method is the necessity of a different sensor surface for each analogue tested (18), as the amino-coupling method is irreversible. Sensor chips from BIAcore are of high-quality in terms of a well-defined

10.1021/bc0100399 CCC: $22.00 © 2002 American Chemical Society Published on Web 01/25/2002

Reversible Surface Thiol Immobilization

surface chemistry, but they are, at the same time, expensive. It is, therefore, of great interest to establish a fully reversible immobilization method which makes the repeated usage of a biosensor surface possible. Our novel approach allows for the economical direct kinetic binding studies of several ligands to related antibodies. Different ligands can be reversibly immobilized to the surface. We have carried out direct kinetic analysis of binding parameters of a polyclonal anti-2,4-D antibody to 2,4-D and 2,4,5-T analogues using one sensor chip.

Bioconjugate Chem., Vol. 13, No. 2, 2002 189 Scheme 1. PDEA/2,4-D and PDEA/2,4,5-T Conjugatesa

MATERIALS

2,4-D, 2,4,5-T, and all other reagents were obtained from Wako Pure Chemical Industries, Ltd. (Tokyo, Japan) unless otherwise noted. Sodium formate and formic acid were obtained from Kanto Chemicals Co., Inc. (Tokyo, Japan). Cystamine hydrochloride was purchased from Tokyo Kasei (Tokyo, Japan). Surface plasmon resonance measurements were carried out on a BIAcore 3000 (BIAcore AB, Uppsala, Sweden) automated system using C1 four-channel sensor chips (BIAcore AB, Uppsala, Sweden). Reagents for the BIAcore instrument consisted of a HBS-EP buffer [10 mM HEPES (pH 7.4), 150 mM NaCl, 3 mM EDTA, and 0.005% polysorbate 20 (v/v)], BIAnormalizing solution [70% (w/w) glycerol], a coupling kit (containing N-hydroxysuccinimide (NHS), N-ethyl-N-[3-(diethylamino)propyl] carbodiimide (EDC) and 1 M ethanolamine hydrochloride (pH 8.5)), and the thiol-coupling reagent 2-(2-pyridinyldithio)ethaneamine hydrochloride (PDEA), all from BIAcore. The polyclonal anti-2,4-D antibody was obtained from Guildhay, Ltd. (Guildford, Surrey, U.K.). METHODS

Immobilization of 2,4-D Analogues to a C1-Sensor Chip. All of the experiments were carried out at 25 °C and with a flow rate of 5 µL/min. Immobilization of 2,4-D to a C1-sensor chip, which has a flat carboxymethylated surface without dextran matrix, was performed according to the surface thiol method described in the BIAapplications Handbook (14). Briefly, the biosensor surface was normalized with BIAnormalizing solution and primed with a HBS-EP buffer. It was then activated by a 2 min injection of a solution of 0.2 M EDC and 0.05 M NHS. Cystamine dihydrochloride (40 mM, solved in 100 mM boric acid (pH 8.5)) was then injected (3 min), followed by a 3 min injection of 100 mM dithiothreitol (DTT) reducing solution (solved in 100 mM boric acid (pH 8.5)) to reduce the dithio bridges of the immobilized cystamine. Two flow cells of the biosensor chip were prepared in the described way. In the following experiments, except for the linked reactions control experiment, one was used as the measuring cell and the other as the reference cell. A mixture consisting of 50 µL of 45 mM PDEA (solved in an MES buffer (pH 5.0)), 50 µL of NHS (0.1 M), 50 µL of EDC (0.4 M), and 50 µL of 1.81 mM 2,4-D (solved in a 10 mM PBS buffer (pH 7.4)) was incubated on ice for 1 h to obtain PDEA/2,4-D conjugates (Scheme 1a). The mixture was then injected over the modified biosensor surface of the measuring cell (7 min) in order to immobilize 2,4-D analogues; the thiopyridone leaving group from the PDEA/2,4-D conjugate was set free, and a new dithio bridge between the surface thiol and the thiol of the 2,4-D analogue was formed. A 4 min injection of PDEA/NaCl solution (20 mM PDEA and 1 M NaCl in a 100 mM formate buffer (pH 4.3)) followed to block remaining unreacted active thiol groups in both flow cells.

a (a) PDEA/2,4-D conjugate and (b) PDEA/2,4,5-T conjugate, which consist of a thiopyridone ring (leaving group) and the 2,4-D analogue or the 2,4,5-T analogue, respectively. The analogs are bound to the biosensor surface during the immobilization step.

Linked Reactions Control Experiments. For the linked reactions control experiment, 100 nM anti-2,4-D antibody solution was injected for 1, 3, and 20 min over the measuring cell of the prepared biosensor surface at a flow rate of 5 µL/min. The biosensor surface was regenerated after each run with a 100 mM hydrochloric acid wash (1 min). Direct Kinetic Binding Analysis. To determine binding affinity parameters of anti-2,4-D antibodies to immobilized 2,4-D analogues, we injected five concentrations of anti-2,4-D antibody solution (37.5, 75, 150, 300, and 600 nM) over the biosensor surface. The biosensor surface was regenerated after each run with a 100 mM hydrochloric acid wash (1 min). Sensorgrams for each solution were obtained in duplicate. The fitting of the curves to the bivalent analyte model (19) was carried out by the BIAevaluation software (version 3.1). Removal of 2,4-D Analogues from the C1-Sensor Chip. 2,4-D Analogues were removed from the biosensor surface by injection of a 100 mM DTT reducing solution for 4 min. To ensure that all analogues had been removed, a 600 nM anti-2,4-D antibody solution was injected for 2 min. Scheme 2 shows in an overview the preparation of the C1-sensor chip, the immobilization of carboxyl group containing hapten analogues, the blocking of unreacted thiol groups, and the removal of the analogues from the surface. Cross-Reactivity Study. A mixture consisting of 50 µL of PDEA (45 mM in an MES buffer (pH 5.0)), 50 µL of NHS (0.1 M), 50 µL of EDC (0.4 M), and 50 µL of 1.57 mM 2,4,5-T (solved in a 10 mM PBS buffer (pH 7.4)) was incubated on ice for 1 h to obtain PDEA/2,4,5-T conjugates (Scheme 1b). The mixture was then injected for 7 min over the biosensor surface of the measuring cell, from which the 2,4-D analogues had been removed before, to immobilize the 2,4,5-T analogues. To determine binding affinity parameters of anti-2,4-D antibodies to immobilized 2,4-D analogues, we injected five concentrations of anti-2,4-D antibody solution (37.5, 75, 150, 300, and 600 nM) over the biosensor surface. The biosensor surface was regenerated after each run with a 100 mM hydrochloric acid wash (1 min). Sensorgrams for each solution were obtained in duplicate. The fitting of the curves was carried out as described previously. Creation of a Calibration Plot for 2,4-D Using a C1-Sensor Surface. A second C1-sensor chip was prepared according to the surface thiol method as described previously. Eight mixtures consisting of 50 µL of 45 mM

190 Bioconjugate Chem., Vol. 13, No. 2, 2002

Schlecht et al.

Scheme 2. Overview of the Preparation of a C1-sensor Chip, the Immobilization of Analogs of Carboxyl Group Containing Haptens, the Blocking of Unreacted Thiol Groups, and the Removal of the Analogs from the Surfacea

a (Step 1) Carboxyl groups on the biosensor surface are activated by the injection of a mixture of 0.2 M EDC and 0.05 M NHS (1:1) for 2 min. A 40 mM cystamine dihydrochloride solution is injected for 3 min. The dithio bridges of the immobilized cystamine are reduced by injection of a 100 mM DTT solution for 3 min. (Step 2) PDEA/x conjugates (x stands for carboxyl group containing haptens, like 2,4-D or 2,4,5-T) are injected over the sensor surface. The dithio bridges between thiopyridone, the x analogue is cleaved, and a new bridge between a thiol group on the sensor surface and the analogue is formed. (Step 3) Unreacted thiol groups on the sensor surface are blocked by injection of a PDEA/NaCl solution. (Step 4) Dithio bridges between the analogues and the thiol groups on the sensor surface are cleaved by a reducing step. After the reducing step, a 600 nM anti-2,4-D antibody solution is injected for 2 min in order to ensure that all 2,4-D analogues have been removed.

PDEA (solved in an MES buffer (pH 5.0)), 50 µL of NHS (0.1 M), 50 µL of EDC (0.4 M), and 50 µL of 2,4-D (0, 1.8, 4.5, 13.5, 27, 54, 180, and 450 µM solved in a 10 mM PBS buffer (pH 7.4)) were incubated for 1 h on ice. These mixtures were sequentially injected over the biosensor surface (7 min) of the measuring cell in order to immobilize 2,4-D analogues. A 4 min injection of PDEA/ NaCl solution (20 mM PDEA and 1 M NaCl in a 100 mM formate buffer (pH 4.3)) followed to block remaining unreacted active thiol groups in both flow cells. Immobilized levels of 2,4-D analogues were detected by injecting a 600 nM anti-2,4-D antibody solution (2 min) and measuring the shift of the relative response 2 min postinjection in the measuring and the reference cell. The antibody solution was injected 5 to 10 times, depending on how fast a stable relative response level was reached. The biosensor surfaces were regenerated after each run by washing three times with 100 mM hydrochloric acid (1 min). To remove all 2,4-D analogues from the biosensor surface before the injection of a new 2,4-D/PDEA/NHS/ EDC mixture, a DTT reducing solution was injected over the biosensor surface for 4 min. To ensure that all 2,4-D analogues were removed by this reducing step, anti-2,4-D antibodies (600 nM) were injected over the biosensor surface (2 min). RESULTS

Immobilization of 2,4-D Analogues on a C1-Sensor Chip and Linked Reactions Control Experiment. After conjugation of 2,4-D analogues to a C1sensor chip by the thiol coupling method, an immobiliza-

Figure 1. Linked reactions control experiment performed with anti-2,4-D antibodies injected over C1-sensor chip with immobilized 2,4-D analogues. A 100 nM anti-2,4-D antibody solution was injected for 1, 3, and 20 min over the prepared biosensor surface at a flow rate of 5 µL/min.

tion level of 10 RU was found. Sensorgrams obtained from the linked reactions control experiment are shown in Figure 1. The three sensorgrams show differently shaped association and dissociation profiles for increasing reaction times. This indicates a multistep reaction between the ligands on the biosensor surface and the soluble analyte. Because the analyte is a bivalent antibody, we have assumed that these results reflect its binding to one or two ligands on the surface. Dissociation of the antibody from the surface also occurs in two steps. Determination of Binding Affinities of Anti-2,4-D Antibodies to 2,4-D Analogues. Sensorgrams which were obtained during the binding studies and which were fitted according to the bivalent analyte model are depicted in Figure 2. Binding parameters are shown in Table 1. Binding of the first ligand molecule is described

Reversible Surface Thiol Immobilization

Bioconjugate Chem., Vol. 13, No. 2, 2002 191

Table 1. Overview of the Binding Parameters Obtained from Direct Kinetic Analysis hapten

ka1 (1/Ms)a

2,4-D 2,4,5-T

7.52 × 6.42 × 103 103

kd1 (1/s)b 0.024 0.215

KD (1/M) 10-6

3.19 × 3.35 × 10-5

ka2 (1/RUs)c 8.09 × 0.215

10-5

kd2 (1/s)d 2.00 × 0.149

10-2

KD2 (1/RU) 2.47 × 102 0.693

a Association rate constant for A + B ) AB. b Dissociation rate constant for AB ) A + B. c Association rate constant for AB + B ) AB2. Note that the association rate constants for the second binding are 1/RUs, not 1/Ms. This is because the reaction involves complex (in RU) binding to ligand (in RU). d Dissociation rate constant for AB2 ) AB + B. (A stands for antibody. B stands for immobilized ligand.)

Figure 2. Direct kinetic analysis of binding of anti-2,4-D antibodies to 2,4-D analogues immobilized to a C1-sensor chip. Five concentrations of anti-2,4-D antibody solution (37.5, 75, 150, 300, and 600 nM) were injected over the biosensor surface for 2 min at a flow rate of 5 µL/min. The biosensor was regenerated after each run with a 100 mM hydrochloric acid wash (1 min).

by a single set of rate constants so that the two sites on the analyte are equivalent in the first binding step. Binding of the second ligand molecule is described by a second set of rate constants. The first and the second equilibrium dissociation constant calculated for anti2,4-D antibodies binding to 2,4-D analogues is 3.19 × 10-6 M-1 and 2.47 × 10+2 RU-1, respectively. Removal of 2,4-D Analogues from the C1-Sensor Chip. All 2,4-D analogues could be removed from the biosensor surface after reduction with a 100 mM DTT solution. No remaining 2,4-D analogues conjugated to the biosensor surface could be detected by injecting a 600 nM anti-2,4-D antibody solution; the relative responses (signal in the measuring cell minus the signal in the reference cell) were always decreased to nearly 0 RU. Immobilization of 2,4,5-T Analogues on the C1Sensor Chip and Cross-Reactivity Study. After conjugation of 2,4,5-T analogues to the C1-sensor chip, an immobilization of 8 RU was found. Sensorgrams which were obtained during the binding studies and which were fitted according to the bivalent analyte model are shown in Figure 3. Binding parameters are shown in Table 1. The first and the second equilibrium dissociation constant for anti-2,4-D antibodies binding to 2,4,5-T analogues is 3.35 × 10-5 M-1 and 6.93 × 10-1 RU-1, respectively. The cross-reactivity of the anti-2,4-D antibody to 2,4,5-T is, therefore, 10%. Creation of a Calibration Curve for 2,4-D. As expected, the relative response level increased with the concentration of 2,4-D in the mixtures and, therefore, with the amount of immobilized 2,4-D analogues. To obtain a calibration curve, the antibody solution was injected 5 to 10 times, depending on how fast a stable relative response level was reached. We have found that the relative response decreases during the first few cycles of detection, as shown in Figure 4, and we think this is due to a removal of only loosely bound 2,4-D analogues from the biosensor surface during the washes with 100 mM hydrochloric acid. Therefore, we have repeated the detection steps until the average of five relative responses had a standard deviation of 10%. The calibration plot of

Figure 3. Direct kinetic analysis of binding of anti-2,4-D antibodies to 2,4,5-T analogues immobilized to a C1-sensor chip. Five concentrations of anti-2,4-D antibody solution (37.5, 75, 150, 300, and 600 nM) were injected over the biosensor surface for 2 min at a flow rate of 5 µL/min. The biosensor was regenerated after each run with a 100 mM hydrochloric acid wash (1 min).

Figure 4. Stability of the relative response levels during detection of 2,4-D analogues. A mixture of PDEA, NHS, EDC, and 450 µM 2,4-D was injected over a C1-sensor chip, which was prepared according to the surface thiol method. Immobilized levels of 2,4-D analogues were detected by injection of a 600 nM anti-2,4-D antibody solution. After each detection cycle, the antibodies were washed from the surface by injection of a 100 mM hydrochloride solution (3 times). During the first few cycles, the relative response levels decreased because of the removal of only loosely bound 2,4-D analogues until it became stable.

the average relative responses versus the concentration of 2,4-D in the immobilized mixtures is depicted in Figure 5. The relative responses after reduction with DTT reducing solution were always decreased to nearly 0 RU. A detection of 2,4-D analogues down to a concentration of 0.1 µg/mL was possible. In this case, the average relative response and standard deviation were 34.6 and 2.4 RU (7%), respectively. DISCUSSION

The analytical system, BIAcore (20), is based on a biosensor that uses surface plasmon resonance to monitor the adsorption of biomolecules on a sensor chip. The SPR phenomenon measures changes in the refractive index in the vicinity of the surface. Such changes are correlated to the change in adsorbed mass (21), which makes it

192 Bioconjugate Chem., Vol. 13, No. 2, 2002

Figure 5. Logarithmic calibration plot for 2,4-D. Mixtures consisting of PDEA, NHS, EDC, and eight different concentrations of 2,4-D (0, 1.8, 4.5, 13.5, 27, 54, 180, and 450 µM) were sequentially injected over a C1-sensor chip, which was prepared according to the surface thiol method. Immobilized 2,4-D analogues were detected by a 600 nM anti-2,4-D antibody solution. After each run, the biosensor surface was regenerated with a 100 mM hydrochloric acid wash (1 min). The antibody solution was injected until the average of five relative responses had an SD of 10%. 2,4-D analogues were removed from the biosensor surface by reduction with a 100 mM DTT solution.

suitable for the detection of biomolecules. The larger the mass of the analyte used, the larger the instrumental response; low molecular weight analytes, like herbicides and peptides, only generate a minimal instrument response. An experiment in which the low molecular weight species is conjugated to the biosensor surface is, therefore, advantageous with regard to the generation of an instrument response upon binding of the macromolecular analyte. In this case, the low molecular weight ligand normally is immobilized to the biosensor surface by the amino-coupling method (18, 22, 23). A big disadvantage of this approach is that it necessitates the production of a different biosensor for each analogue studied, as the amino-coupling method is irreversible. We have, therefore, developed a method by which a hapten can be immobilized to and removed from one single biosensor surface of a BIAcore sensor chip for at least seven times. Considering that a BIAcore biosensor allows for 50-100 analytical cycles, depending on regeneration conditions (20), our method makes an even higher frequency of analytical cycles possible to be carried out. We report, for the first time, a PDEA-dependent conjugation of haptens to a biosensor surface. Every hapten which contains a carboxyl group and which can, therefore, be conjugated to PDEA is suitable for this immobilization method. Because small molecules generally have fewer functional groups available for coupling to biosensor surfaces, in many cases they have to be modified for incorporation of a linker functionality (18, 23), like a carboxyl group, for example. Also, such modified haptens are suitable for the immobilization method described here. To obtain a calibration curve for the herbicide 2,4-D, seven different amounts of 2,4-D analogues were immobilized to the biosensor surface of a C1-sensor chip. Detection of 2,4-D analogues down to a concentration of 0.1 µg/mL, which is the maximum admissible concentration of pesticides in drinking water, was possible. A drawback with our method is that other haptens bearing carboxyl groups, such as amino acids or carboxyl acids that are present in environmental samples, might interfere with the amount of immobilized 2,4-D analogues.

Schlecht et al.

In comparison to environmental samples, solutions for binding affinitiy studies are well-defined and, therefore, more suitable for our method. In this study, 2,4-D analogues were immobilized on a C1-sensor chip, and direct kinetic analysis was performed. The sensorgrams obtained were fitted according to the bivalent analyte model, which delivered detailed information about the first and the second binding steps of the bivalent analytes to the immobilized haptens. We have used the same biosensor surface to immobilize 2,4,5-T analogues for direct kinetic analysis. This shows that our method allows for easy and economical direct kinetic analysis of binding of several ligands to related antibodies. Crossreactivity studies can be performed as shown in this paper. The equilibrium dissociation constants (KD) calculated for the first (and the second) site binding were 3 × 10-6 M-1 (2.5 × 10+2 RU-1) and 3 × 10-5 M-1 (7 × 10-1 RU-1) for 2,4-D and 2,4,5-T analogues, respectively, showing a reasonable selectivity of the antibody. The previously published equilibrium dissociation constants for the binding of anti-2,4-D antibodies to 2,4-D were 7 × 10-8 M-1 and 4 × 10-8 M-1 (6, 24). These values showed a tighter binding than those we obtained. This might be due to the fact that we used polyclonal antibodies which were not purified and not due to the immobilization method itself. CONCLUSION

We describe a reversible immobilization method for low molecular weight ligands containing carboxyl groups to biosensor surfaces of a BIAcore surface plasmon resonance instrument. Our method allowed us to obtain a calibration curve for the detection of the commonly used herbicide 2,4-D and to perform direct kinetic analysis of the binding affinity of a polyclonal anti-2,4-D antibody to the analogues of the haptens 2,4-D and 2,4,5-T using only one biosensor surface of a BIAcore sensor chip. There may be drawbacks in the described method of detection, namely, that other substances present in the environmental sample which also carry a carboxyl group might interfere with the level of immobilized 2,4-D. We have proved, however, that our method of immobilization, quantitative detection, and removal of the haptens works in principle. Regarding the kinetic analysis, we have obtained very detailed information. Apparent association and dissociation constants for the first and the second binding step of the bivalent analyte to the immobilized ligands were determined. Calculated equilibrium dissociation constants were smaller as compared to those shown in other publications, but purified monoclonal antibodies or Fab fragments should deliver much better results. ACKNOWLEDGMENT

This research was supported by the German Academic Exchange Service (DAAD). LITERATURE CITED (1) Matsumura, F. (1994) Differential toxicities of insecticides and halogenated aromatics. International Encyclopedia of Pharmacology and Therapeutics, Section 113, p 453, Pergamon Press, New York. (2) Ibrahim, M. A., Bond, G. G., Burke, T. A., Cole, P., Dost, F. N., Enterline, P. E., Gough, M., Greenberg, R. S., Halperin, W. E., McConnel, E., Munro, I. C., Swenberg, J. A., Zahm, S. H., and Graham, J. D. (1991) Weight of the evidence on the human carcinogenicity of 2,4-D. Environ. Health Perspect. 96, 213

Reversible Surface Thiol Immobilization (3) Carlo, G. L., Cole, P., Miller, A. B., Munro, A. I. C., Solomon, K. R., and Squire, R. A. (1992) Review of study reporting an association between 2,4-dichlorophenoxyacetic acid and canine malignant lymphoma. Regul. Toxicol. Pharmacol. 16, 245. (4) Neuberger, M., Rappe, C., Begek, S. Cai, H., Hansson, M., Jaeger, R., Kundi, M., Lim, C. K., Wingfors, H., and Smith, A. G. (1999) Persistent health effect of dioxin contamination in herbicide production. Environ. Res., Sect. A 81, 206-214. (5) Wittmann. C., and Hock, B. (1991) Development of an ELISA for the analysis of atrazine metabolites deethylatrazine and deisopropylatrazine. J. Agric. Food Chem. 39, 11941200. (6) Pena, A. L. S., and Silveira, I. N. (1997) Determination of chlorophenoxy herbicides in waters by capillary gas chromatography with ion trap detection. J. Chromatogr., A 758, 324331. (7) Horacek, J., and Skla´dal, P., (1997) Improved direct piezoelectric biosensors operating in liquid solution for the competitive label-free immunoassay of 2,4-dichlorophenoxyacetic acid. Anal. Chim. Acta 347, 43-50. (8) Mosiello, L., Nencini, L., Serge, L., and Spano, M. (1997) A fibre-optic immunosensor for 2,4-dichlorophenoxyacetic acid detection. Sens. Actuators, B 38-39, 353-359. (9) Medyantseva, E. P., Vertlib, M. G., Kutyreva, M. P., Khaldeeva, E. I., Budnikov, G. K., and Eremin, S. A. (1997) The specific immunochemical detection of 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid pesticides by amperometric cholinesterase biosensors. Anal. Chim. Acta 347, 71-78. (10) Minunni, M., and Mascini, M. (1993) Detection of pesticide in drinking water using real-time biospecific interaction analysis (BIA). Anal. Lett. 26, 1441-1460. (11) Harris, R. D., Luff, B. J., Wilkinson, J. S., Piehler, J., Brecht, A., Gauglitz, G., and Abuknesha, R. A. (1999) Integrated optical surface plasmon resonance immunoprobe for simazine detection. Biosens. Bioelectron. 14, 377-386. (12) Catimel, B., Nerrie, M., Lee, F. T., Scott, A. M., Ritter, G., Welt, S., Old, L. J., Burgess, A. W., and Nice, E. C. (1997) Kinetic analysis of the interaction between the monoclonal antibody A33 and its colonic epithelial antigen by the use of an optical biosensor. A comparison of immobilization strategies. J. Chromatogr., A 776, 15-30. (13) Lo¨fas, S., Johnsson, B., Edstro¨m, A., Hansson, A., Lindquist, G., Mu¨ller Hillgren, R., and Stigh, L. (1995) Methods for site controlled coupling to carboxymethyldextran surfaces in surface plasmon resonance sensors. Biosens. Bioelectron. 10, 813-822.

Bioconjugate Chem., Vol. 13, No. 2, 2002 193 (14) (1998) BIAapplications Handbook, pp 4/6-4/15, BIAcore AB, Uppsala, Sweden. (15) Karlsson, R. (1994) Real-time competitive kinetic analysis of interactions between low-molecular-weight ligands in solution and surface-immobilized receptors. Anal. Biochem. 221, 142-151. (16) Sternesjo, A., Mellgren, C., and Bjorck, L. (1995) Determination of sulfamethazine residues in milk by surface plasmon resonance-based biosensor assay. Anal. Biochem. 226, 175-181. (17) Alfthan, K. (1998) Surface plasmon resonance biosensors as a tool in antibody engineering. Biosens. Bioelectron. 13, 653-663. (18) Adamczyk, M., Gebler, J. C., Gunasekera, A. H., Mattingly, P. G., and Pan, Y. (1997) Immunoassay reagents for thyroid testing. 2. Binding properties and energetic parameters of a T4 monoclonal antibody and its Fab fragment with a library of thyroxine analogue biosensors using surface plasmon resonance. Bioconjugate Chem. 8, 133-145. (19) (1997) BIAevaluation Software Handbook (version 3.0), pp C9-C12, BIAcore AB, Uppsala Sweden. (20) Malmquist, M. (1993) Biospecific interaction analysis using biosensor technology. Nature 361, 186-187. (21) Karlsson, R. Mo, J. A., and Holmdahl, R. (1995) Binding of autoreactive mouse anti type II collagen antibodies derived from the primary and the secondary immune response investigated with the biosensor technique. J. Immunol. Methods 188, 63-71. (22) Weller, M. G., Schuetz, A. J., Winklmair, M., and Niesser, R., (1999) Highly parallel affinity sensor for the detection of environmental contaminants in water. Anal. Chim. Acta 393, 29-41. (23) Adamczyk, M., Johnson, D. D., Mattingly, P. G., Moore, J. A., and Pan, Y. (1998) Immunoassay Reagents for thyroid testing. 3. Determination of the solution binding affinities of a T4 monoclonal antibody Fab fragment for a library of thyroxine analogues using surface plasmon resonance. Bioconjugate Chem. 9, 23-32. (24) Skla´dal, P., Minunni, M., Mascini, M., Kola´ø, V., and Fra´nek, M. (1994) Characterization of monoclonal antibodies to 2,4-dichlorophenoxyacetic acid using a piezoelectric quartz crystal microbalance in solution. J. Immunol. Methods 176, 117-125.

BC0100399