Environ. Sci. Technol. 2004, 38, 6795-6802
Development of a Class-Specific ELISA for Sulfonylurea Herbicides (Sulfuron Screen) PETRA DEGELMANN,† JEAN WENGER,‡ REINHARD NIESSNER,† AND D I E T M A R K N O P P * ,† Institute of Hydrochemistry and Chemical Balneology, Technical University of Munich, Marchioninistrasse 17, D-81377 Mu ¨ nchen, Germany, and Optimization Chemistry, Syngenta Crop Protection AG, CH-4002 Basel, Switzerland
The development of a direct competitive ELISA for the detection of a broad range of sulfonylurea herbicides (SUs) is described. Polyclonal antibodies were generated in rabbits using three different immunizing haptens. Antiserum with the broadest specificity was obtained with a mesosulfuronbenzylamine derivative which was coupled via a succinic acid spacer to keyhole limpet hemocyanine. A heterologous enzyme tracer which did not contain the succinic acid bridge was prepared using activated horseradish peroxidase. The direct competitive ELISA was optimized and applied for spiked tap and surface water samples. From 30 SUs, 8 compounds showed a molar cross-reactivity (CR) higher than 100% (this value was set for the hapten) and 11 compounds CRs between 10% and 100%. The ELISA can detect 16 SUs at a concentration of 0.1 µg/L or lower. Different surface and tap water samples were spiked with chlorimuron ethyl, metsulfuron methyl, or primisulfuron methyl at concentrations of 100, 200, or 500 ng/L and subsequently analyzed by both ELISA and HPLCUV. Correlation analysis revealed good agreement between both methods (r2 ) 0.983/0.948/0.982; n ) 21 for each analyte). Using ELISA, no sample pretreatment other than filtration was necessary.
Indroduction Sulfonylurea herbicides (SUs) were first introduced in 1982 by DuPont Crop Protection. At present more than 25 SUs are registered for agricultural uses worldwide (1, 2). Structurally common to all SU herbicides is the sulfonylurea group which is attached to two chemical structures. One structural element can constitute an aliphatic, aromatic, or heterocyclic group. The second structure mostly is a triazine or pyrimidine, but can also be a triazole. They are used for weed control in cereals such as wheat, rice, corn, and other crops such as potatoes, sugar beet, and turnip. The biological mode of action is based on the inhibition of acetolactate synthase (ALS), an enzyme which takes part in the protein synthesis of plants. The selectivity is caused by the ability respectively nonability of a plant to metabolize the SUs. The high herbicidal activity of SUs leads to low application rates of less than 100 g ha-1. These rates represent concentrations in environmental matrixes of about 10-100-fold less than * Corresponding author phone: +49 89 2180 78252; fax: +49 89 2180 78255; e-mail:
[email protected]. † Technical University of Munich. ‡ Syngenta Crop Protection AG. 10.1021/es0496266 CCC: $27.50 Published on Web 10/26/2004
2004 American Chemical Society
other herbicides, e.g., triazines (3). If used in accordance with label directions, these herbicides do not constitute an important environmental and mammalian hazard. However, if improperly applied, or after accidental events, higher amounts of herbicides can reach both surface water and groundwater. In conclusion, sensitive and reliable analytical methods are needed to evaluate their presence and persistence in water, soil, and plant material at these very low levels. Various methods have been described for the determination of SUs. While gas chromatography (GC) is preferred in pesticide residue analysis, the polar SUs are not directly amenable to GC, because of their thermal instability and very low volatility. However, after derivatization, for example, with diazomethane, GC analysis was reported (4). Other methods are capillary electrophoresis (CE) (5-8), high-performance liquid chromatography (HPLC) (3, 9, 10), and enzyme immunoassay (ELISA) (11-24). Most of the applications known are based on HPLC using reversed-phase columns followed either by ultraviolet (UV) or mass spectrometric (MS) detection (1, 25-29). Because of the low analyte concentrations, chromatographic techniques need preliminary enrichment and cleanup steps, such as solid-phase extraction (SPE) or liquidliquid extraction (LLE). Immunoassays are typically highly sensitive and comfortable to implement. However, because immunoassays are indirect assays relying on the antigen-antibody interaction, the results can be faulty due to interference of matrix constituents in complex samples such as soil or wastewater. This had led to concern about data reliability in immunoassay, and this is one of the main reasons why many analytical chemists still prefer methods which yield a more direct analyte-related signal, e.g., a peak in a chromatogram. Another feature, sometimes invoked as a limitation of immunoassays compared to multiresidue chromatographic techniques, is the high specificity of the antibodies which generally allows the determination of only one single analyte. However, antibodies can be prepared which can be used for a rapid and inexpensive multianalyte screening analysis of structurally related compounds that might be a class of pesticides, for example. Alternatively, corresponding antibodies can be used to prepare immunoaffinity supports as selective sorbents for analyte enrichment from complex samples followed by chromatographic separation and detection. Up to now several immunoassays for SUs such as chlorsulfuron, triasulfuron, metsulfuron methyl, bensulfuron methyl, and chlorimuron ethyl have been reported which make use of either polyclonal or monoclonal antibodies (1124). Most of the assays can be considered rather specific for one single analyte. Only two papers described the preparation of antibodies which could recognize 4-8 different SUs with a percentage of cross-reactivity higher than 10% (14, 21). Another approach to establish a multiple-analyte method for commercial product monitoring was chosen by Strahan et al. (19). Nine polyclonal antibodies of different specificities were prepared and applied in parallel in single wells on the same polyantigen microtitration plate in corresponding ELISA. The obtained LOD was 5 µg/kg for a single analyte in the formulated product. It was the aim of the present investigation to prepare antibodies of broad specificity for SUs which can be used to establish a class-specific ELISA for rapid and cost-effective sample monitoring. At a later stage the use of these antibodies for selective analyte enrichment (immunoextraction) coupled with HPLC or CE is envisaged. VOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6795
FIGURE 1. Hapten-KLH conjugates that were used as immunogens: hapten 1 (1-[2-(methyl ester)phenylsulfonyl]monoamidosuccinic acid); hapten 2 (2-[[[[(4,6-dimethoxy-2-pyrimidinyl)amino]carbonyl]amino]sulfonyl]-2-(carboxymethylbenzene)-4-[methyl]amino]succinic acid); hapten 3 (6-(4-methoxy-6-methyl-1,3,5-triazinyl-2-amino)hexanoic acid).
Experimental Section Reagents. All reagents were of analytical grade unless specified otherwise. The analytical standards of amidosulfuron, cinosulfuron, nicosulfuron, primisulfuron methyl, prosulfuron, the primary compound for hapten 1 (2-(methyl ester)phenylsulfonamide), and hapten 2 (2-[[[[(4,6-dimethoxy-2-pyrimidinyl)amino]carbonyl]amino]sulfonyl]-2-(carboxymethylbenzene)-4-[[methyl]amino]succinic acid) were from Syngenta (Basel, Switzerland). Hapten 3 (6-(4-methoxy6-methyl-1,3,5-triazinyl-2-amino)hexanoic acid) was kindly provided by Dr. M. Franek (VRI, Brno, CR) (Figure 1). The standards flucarbazon and propoxycarbazon were a gift from Dr. R. Fritz (Bayer AG, Leverkusen, Germany). All other SU standards were purchased from Dr. Ehrenstorfer (Augsburg, Germany), except for flazasulfuron, which was obtained from Riedel-de Ha¨en (Seelze, Germany). Amine-reactive horseradish peroxidase (EZ-Link Plus activated peroxidase), unconjugated goat anti-rabbit IgG, and dialysis cassettes were purchased from Pierce (Rockford, IL). Bovine serum albumin (BSA) was obtained from Serva (Heidelberg, Germany). Keyhole limpet hemocyanine (KLH), 3,3′,5,5′-tetramethylbenzidine (TMB), dimethyl sulfoxide (DMSO), dimethylformamide (DMF), N-hydroxysuccinimide (NHS), and N-ethylN′-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) were purchased from Sigma-Aldrich (Taufkirchen, Germany). Freund’s complete and incomplete adjuvants were obtained from Difco Labs (Detroit, MI). Highly pure deionized water (Milli-Q) was used for preparation of all aqueous solutions. All other reagents were obtained from VWR (Ismaning, Germany). Preparation of Immunogens and Tracers. The synthesis of hapten 1 (1-[(2-methyl ester)phenylsulfonyl]monoamidosuccinic acid) was performed as described earlier (14). Briefly, equimolar amounts (6.0 mmol) of the phenylsul6796
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004
fonamide and succinic anhydride were dissolved in 15 mL of dioxane. Then, 12 mmol of 1,8-diazabicyclo[5.4.0]undec7-ene (DBU) dissolved in 6 mL of dioxane was added drop by drop under cooling. After a reaction time of 2 h, the mixture was acidified and dioxane was evaporated to dryness. The product was redissolved in ethyl acetate and dried over Na2SO4. Solvent was evaporated and the product recrystallized from ethanol. The structures of haptens 1-3 were confirmed by 1H NMR (300 MHz, Bruker ARX 300): (hapten 1) (CD3CN) δ [ppm] ) 2.17 (s, 1H, NH), 2.47 (m, 2H, CH2), 2.53 (m, 2H, CH2), 3.91 (s, 3 H, OCH3), 7.73 (m, 3H, Ar), 8.18 (m, 1H, Ar), 9.40 (s, 1 H, COOH); (hapten 2) (CDCl3) δ [ppm] ) 2.54 (m, 2H, (CH2)2), 2.7 (m, 2H, (CH2)2), 3.89 (s, 3H, OCH3), 4.0 (s, 6H, 2 OCH3), 4.52 (m, 2H, CH2), 5.79 (s, 1H, Ar), 6.95 (m, 1H, NH), 7.68 (m, 2H, Ar), 8.15 (s, 1H, NH), 8.27 (s, 1 H, Ar), 10.64 (s, 1H, NH), 12.60 (s, 1H, COOH); (hapten 3) (CDCl3) δ [ppm] ) 1.52 (m, 2H, (CH2)3), 1.74 (m, 4H, (CH2)3), 2.33 (s, 3H, CH3), 2.41 (m, 2H, CH2(COOH)), 3.49 (m, 3 H, CH2(NH)), 4.0 (s, 1H, OCH3), 7.88 (s, 1H, NH). The haptens were covalently linked via their carboxyl function to KLH by the activated ester method (30). For the preparation of the immunogens, 0.2 mmol of each hapten was dissolved in 2 mL of DMF and 0.8 mmol of NHS and an equimolar amount of EDC was added. The reaction mixture was stirred overnight at room temperature and then added to the protein solution which contained 40 mg of KLH in 6 mL of Borax buffer (0.1 M, pH 9.5). The mixture was further incubated at room temperature for 4 h. The conjugates were purified by extensive dialysis with 10-fold-diluted PBS buffer (0.08 M phosphate buffer containing 0.15 M NaCl, pH 7.6) and with water, finally. The obtained conjugate solutions were freeze-dried and stored at 4 °C until they were used for immunization. Homologous Tracers. For homologous tracer preparation, the active ester was formed as described for the immunogens. A volume of 0.2 mL of active ester solution was added to a solution of 5 mg of horseradish peroxidase (HRP) in 0.5 mL of Borax buffer, and then the resulting solution was incubated for 3 h at room temperature. The tracers were purified by size exclusion chromatography (Sephadex PD-10, Amersham Pharmacia, Freiburg, Germany) and stored at 4 °C. Heterologous Tracer of Hapten 2. A heterologous tracer was prepared with a derivative of hapten 2 (i.e., without the succinic acid spacer). According to the protocol, 0.5 mL of hapten solution (1.0 mg/mL) in DMSO/PBS (1:10, v/v) was added to 1 mg of activated HRP (contains 1-3 mol of aminereactive groups/mol of HRP) in 0.1 mL of water, followed by the addition of 10 µL of reduction buffer (5 M NaCNBH3 in 1 M NaOH), and the resulting solution was incubated for 1 h at room temperature. Then, 20 µL of quenching buffer (3.0 M ethanolamine, pH 9.0) was added, and the resulting solution was incubated for an additional 15 min. The tracer was purified by size exclusion chromatography as described above. Determination of the Coupling Ratio. The coupling density of the immunogens was determined by reaction of free amine groups at the protein with trinitrobenzenesulfonic acid (TNBS) according to Habeeb (31). MALDI-TOF-MS was employed for the determination of the coupling rates of HRP conjugates (tracers). In the case of immunogens, the coupling density was estimated by comparing the absorbance with the corresponding values of analyte-free protein. From about 2000 lysine amino groups at KLH, which are available for binding, 16% were conjugated with hapten 1 (immunogen 1), 14% with hapten 2 (immunogen 2), or 20% with hapten 3 (immunogen 3) (30). MALDI-TOF-MS analysis of the tracers was done by AnagnosTec (Luckenwalde, Germany). The average coupling ratio was 2-3 mol of hapten/mol of enzyme. Immunization. For raising polyclonal antibodies, randombred rabbits were immunized in pairs with one of the three
immunogens: hapten 1, KLH/rabbits R01 and R02; hapten 2, KLH/rabbits R03 and R04; hapten 3, KLH/rabbits R05 and R06. Routinely, 1 mg of immunogen was dissolved in 1 mL of sterile physiological saline and emulsified with an equal volume of Freund’s adjuvant. For priming, 1 mL of the emulsion which was prepared with Freund’s complete adjuvant was intradermally injected at 10 different sites on the neck and the back of a rabbit in 0.1 mL aliquots. Up to eight booster injections with the use of Freund’s incomplete adjuvant were administered in the same manner over a period of time of 50 weeks. Antibody production was monitored during immunization by ELISA. The final bleeding was done through the ear vein on three consecutive days. Sera were separated and pooled. Conveniently sized portions were stored in liquid nitrogen. Direct Competitive ELISA. For the assay, microtiter plates (96 flat-bottom wells with high binding capacity; Greiner, Frickenhausen, Germany) were coated with goat anti-rabbit IgG (0.5 µg/mL, 250 µL/well) in coating buffer (0.05 M sodium carbonate buffer, pH 9.6) overnight at room temperature. The plates were covered with adhesive plate sealing film (ThermaSil, Sigma) to prevent evaporation. On the following day, the plates were washed three times with PBS-Tween (PBS buffer containing 0.05% Tween 20, pH 7.6) using an automatic plate washer (1296026 Delfia Platewash, Wallac ADL, Freiburg, Germany) and thoroughly tapped dry. Then, 200 µL of diluted rabbit antiserum (1:10000 in PBS) was added to each well and incubated for 1 h with shaking (EAS 2/4, SLT Labinstruments, Crailsheim, Germany). After a further washing step, 200 µL of standard or sample was added to each well and preincubated for 30 min at room temperature. For construction of the calibration curves sulfonylurea stock solutions (0.1 g/L) were prepared with methanol and then further diluted with pure water to obtain individual standard solutions. Next, tracer dilution (1:10000 in PBS, 50 µL/well) was pipetted, and the plates were incubated for another 15 min. The plates were washed again, and 200 µL of substrate solution (TMB/H2O2) was added to each well. It was prepared by addition of 500 µL of TMB (375 mg of TMB in a mixture of 5 mL of DMSO, 20 mL of methanol) and 100 µL of 1% H2O2 to 25 mL of citrate buffer (0.04 M sodium citrate, pH 3.8). The plates were shaken for about 15 min for color development. Finally, the enzyme reaction was stopped with sulfuric acid (5%, containing 0.01% SDS; 100 µL/well), and the absorbance was read at 450 nm with a plate reader (Easy Reader 340 ATC, SLT Labinstruments). All determinations were made at least in triplicate. The sigmoidal curves were calculated by mathematically fitting experimental points using Rodbard’s four-parameter function with the software Origin (Microcal, Northampton, MA). Graphs were plotted in the form of absorbance against the logarithm of analyte concentration. The inflection point of the calibration curve (IC50), the limit of detection (LOD), and the limit of quantification (LOQ; was defined as analyte concentration at 20% inhibition) are the main parameters received directly from this calculation. Cross-Reactivity Determination. The relative sensitivity of the immunoassay toward the SUs listed in Table 1 was determined by assaying a dilution series of each compound in water. All chemicals were tested in the concentration range from 0.001 to 1000 µg/L. The IC50 values (molar concentration of inhibitor that produces a 50% decrease of the maximum normalized response) were compared and expressed as a percent IC50 on the basis of a 100% response of the corresponding hapten. Water Samples. Surface water samples were collected in agricultural areas near Wolfratshausen and Landsberg, two district towns located south and west of Munich in South Bavaria. Surface water was collected in brown glass bottles (1 L) near the banks of lakes Ammersee, Starnberger See,
and Wo¨rthsee as well as the rivers Windach, Loisach, and Isar. A short sensory analysis of fresh samples including appearance and odor as well as pH measurement and determination of the dissolved organic carbon (DOC) was made. Tap water was taken from the municipal water supply of the laboratory. Samples were analyzed on the same day or stored in the fridge and analyzed in the following week, at the latest. While tap water was used without any preparation, surface water samples were filtered over a glass microfiber filter (GF/C, Whatman, Maidstone, England) to remove particles larger than 1.2 µm. No further sample treatment was done for ELISA. HPLC-UV Procedure. Analysis was performed on a Shimadzu LC system equipped with a SCL-6A controller, two LC-6A pumps, a photodiode array UV-vis detector SPDM6A, and a CTO-10A column oven (Shimadzu, Duisburg, Germany). Chromatographic separations were carried out by means of an NC (250 × 4.6 mm i.d.) (3.0 µm particle size) ProntoSIL 120-3-C18-AQ column (Bischoff, Leonberg, Germany). Injection was performed with a model 7125 injector from Rheodyne (Bensheim, Germany) equipped with a 500 µL sample loop. Analysis was carried out using a mobile phase composition of acetonitrile (ACN)/water (40:60, v/v; containing 3.0 mM TFA) for the separation of metsulfuron methyl and ACN/water (50:50, v/v; containing 3.0 mM TFA) for separation of chlorimuron ethyl and primisulfuron methyl. The flow rate was set at 1.0 mL/min. The analytical column temperature was kept at 25 °C. The UV detector was set at a wavelength of 227 nm. Data were acquired and evaluated by using the CLASS-M10A package from Shimadzu. Peak areas were used for quantification. The calibration curve of each of the SUs was used to calculate the recoveries of the analytes. For the HPLC analysis of water samples a precleaning and enrichment step was necessary. Solid-phase extraction (SPE) was performed using HLB-OASIS supports (Waters, Eschborn, Germany). Conditioning, loading, and elution of the cartridge were performed according to the instructions of the supplier. The supports were preconditioned with 1 mL of methanol and the same amount of water. Then, the acidified sample (obtained by addition of 0.5 mL of 1% phosphoric acid and 2.5 mL of phosphate buffer, 1 M, pH 3.0, to a 100 mL water sample) was passed through the column (flow rate 5-10 mL/min). It was washed with 1 mL of methanol/water (5:95, v/v), and finally the analytes were eluted with methanol. The eluates were dried for 20 min using a gentle stream of nitrogen. The residues were reconstituted with 1.0 mL of ACN/water (40:60, v/v; acidified with 3.0 mM TFA) and analyzed by HPLC-UV.
Results and Discussion Synthesis of Haptens and Immunogens. To obtain polyclonal antibodies with a broad recognition of SUs, three different haptens, which represent typical structural elements of these chemicals, were synthesized. Hapten 1, a phenylsulfonamide residue, represents the left part of the molecule, hapten 2, a benzylamine of mesosulfuron, represents a whole SU compound, and hapten 3, a triazine residue, is typical for the right part of the molecule. All haptens were linked via a spacer of four or six carbon atoms to the carrier protein (Figure 1). One reason for the selection of these derivatives was that SUs themselves do not have appropriate functional groups, such as amino or carboxyl groups, for direct covalent binding to a protein carrier. Second, immunization with only partial molecules can lead to antibodies with broad specificity, as was shown recently (14, 21). However, the same approach led to antibodies of high specificity as well, as was reported for triasulfuron and bensulfuron methyl (12, 16). Progress of Immunization. The response of rabbits during immunization was measured by the direct ELISA. HaptenVOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6797
TABLE 1. CR of Sulfonylurea Herbicides with Mesosulfuron Antiserum (R03) and Heterologous Tracer (Hapten 2 Derivative)
a Inhibitor concentration for 50% inhibition in the competitive ELISA. the analogue) × 100.
specific antibodies were detectable in each of the six sera already after the first booster injection. The highest antibody 6798
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004
b
Percentage of molar cross-reactivity defined as (IC50 of hapten 2/IC50 of
titers were measured after the sixth boost and then remained almost constant. As was found in the final stage of the
immunization, antiserum R03, which was obtained with a complete SU molecule as a hapten, displayed an extraordinary broad cross-reactivity with several SUs. Therefore, this antiserum, which was obtained from final bleeding after the eighth booster injection, was selected for further use. Optimization of Direct Competitive ELISA. To develop sensitive ELISAs, the test conditions, such as the appropriate concentration of immunoreagents used in the assay, the suitable blocking reagent, the effect of temperature, etc., should be carefully optimized. Criteria used to evaluate the optimization were maximum absorbance, dynamic range, IC50, and limit of detection (LOD). Especially in the final phase of immunization an increasing baseline signal of the sigmoidal ELISA calibration curve was experienced (about 40% of maximal absorbance). The problem could be eliminated with the preparation of a new heterologous enzyme tracer which did not contain the C4spacer arm; i.e., the free amine function of the mesosulfuron derivative was directly linked to a carboxyl function of the horseradish peroxidase (base signal
