Anal. Chem. 1999, 71, 1905-1909
Immunofiltration: A Methodology for Preconcentration and Determination of Organic Pollutants Sergi Morais, A Ä ngel Maquieira, and Rosa Puchades*
Departamento de Quı´mica, Universidad Polite´ cnica de Valencia, Camino de Vera s/n 46071 Valencia, Spain
A new concentration procedure using an immunofiltration-based method is described. The approach enables quantitative determination of organic pollutants by filtering large volumes of sample through a poly(vinylidene difluoride) membrane where antibodies have been immobilized by passive adsorption. The analysis is based on a sequential competitive enzyme immunoassay. A wide range of sample volumes have been tested (0.2-5.0 mL) for each type of antibody. The improvement on the assay sensitivity and specificity achieved by means of this concentration procedure is discussed. Using this technique and the insecticide carbaryl as a model analyte, a concentration factor of at least 13 and a limit of detection of 4.75 ng/L are accomplished. The suitability of this methodology is demonstrated by the quantification of the insecticide in several types of water samples (bottled, estuarine, and physiological-saline solutions) with recoveries ranging between 102 and 111%. This method has proved to concentrate carbaryl directly, in an accurate way, for residue analysis without using organic solvents or any extraction process. Furthermore, this procedure offers the advantages of carrying out in the same system both preconcentration and quantitative determination of the analyte. The determination of the most used environmental pollutants such as pesticides in the low ng/mL level by means of traditional analytical techniques is generally expensive and time-consuming since it requires extraction, sample cleanup, and often, analyte derivatization procedures. Nowadays, these processes are carried out by means of solid-phase cartridges or disks. To resolve these drawbacks, immunochemical techniques are gaining use as screening and quantitative methods because of their great specificity, sensitivity, and ability to process large numbers of samples.1 Heterogeneous assays are the most extensively used types of immunoassays in environmental-pollutants analysis.2 These immunoassays involve the separation of the unbound antibody or antigen in a liquid phase from a solid support to which the * Corresponding author: (fax) +34-96-3877349, (e-mail)
[email protected]. (1) Tijssen, P. Practice and Theory of Enzymeimmunoassay. In Laboratory Techniques in Biochemistry and Molecular Biology; Burdon, R. H., Van knippenberg, P. H., Eds.; Elsevier: Amsterdam, 1987; Vol. 15. (2) Meulenberg, E. P.; Mulder, W. H.; Stoks, P. G. Environ. Sci. Technol. 1995, 29 (3), 553-561. 10.1021/ac9811975 CCC: $18.00 Published on Web 03/24/1999
© 1999 American Chemical Society
antigen-antibody complex is cleaved. The solid phase differs in geometry (beads, membranes, or plates) and chemical composition (agarose, polystyrene, polyamide, nitrocellulose, silica, poly(vinylidene difluoride), etc.). In the antibody-immobilized format, the attachment of the antibody to the solid phase (immunosorbent) occurs by covalent binding or passive adsorption. The method for covalent immobilization is related to chemical composition of the support, while passive adsorption is a very general process.3 The most common applications using immunosorbents are immunoaffinity chromatography (IAC),4 immunosensors,5 ELISA,6 and immunofiltration.7 Immunoaffinity chromatography is mainly used as a clean-up procedure, for preconcentration purposes, or as a separative technique, but scarcely is used for quantitative analysis.8 Furthermore, the disruption of the analyte from the immunosorbent by means of desorbent agents, as well as evaporation stages and solvent changes, makes it a very tedious technique. Additionally, the quantification of the analyte must be carried out using chromatographic or immunochemical techniques which increases span time and cost of analysis. Immunofiltration is a technique based on the use of membranes as the adsorptive surface which lacks development of its analytical potential. This technique involves the filtration of rinsing solutions through the membrane where the antibody has previously been immobilized. Most of the applications using the immunofiltration technique are intended for semiquantitative and/ or quantitative determination of chemical compounds, including small molecules such as 2,4,6-trinitrotoluene,9 pesticides,10 or PCB.11 However, in this paper a new application of the immunofiltration method for simultaneous concentration and pesticide quantitative determination at the ng/mL level is described. The (3) Butler, J. E. Solid Phases in Immunoassays. In Immunoassay; Diamandis, E. P., Christopoulos, T. K., Eds.; Academic Press: San Diego, CA, 1996; Chapter 9. (4) Ballesteros, B.; Marco, M. P. Food Technol. Biotechnol. 1998, 36 (2), 145155. (5) Hock, B.; Dankwardt, A.; Kramer, K.; Marx, A. Anal. Chim. Acta 1995, 311, 393-405. (6) Nunes, G. S.; Toscano, I. A.; Barcelo´, D. Trends Anal. Chem. 1998, 17 (2), 79-87. (7) Ijsselmuiden, O. E.; Herbrink, P.; Meddens, M. J. M.; Tank, B., Stolz, E.; Van Eijk, R. V. W. J. Immunol. Methods 1989, 119, 35-43. (8) Fischer-Durand, N.; Daniel, R.; Pichon, V.; Chen, L.; Hennion, M.-C.; Le Goffic, F. Agro-Food-Ind. Hi-Tech. 1997, 8 (1), 21-24. (9) Keuchel, C.; Niessner, R. Fresenius’ J. Anal. Chem. 1994, 350 (7-9), 538543. (10) Hock, A. Biosens. Bioelectron. 1993, 8 (7-8), XX-XXI.
Analytical Chemistry, Vol. 71, No. 9, May 1, 1999 1905
proposed method gathers the advantages shown by IAC such as cleanup and preconcentration processes with those of immunoassays which are ease and rapidness. The main step in this method involves the filtration of large volumes of sample through a membrane where the antibody is immobilized, allowing the concentration of the analyte. In this procedure, the use of desorbent agents is avoided, and 96 simultaneous analyses can be carried out in approximately 2 h. The aim of this work is to set up an immunofiltration procedure that allows both the preconcentration and determination of the analyte using immobilized antibodies. For this purpose, both monoclonal (mAb) and polyclonal (pAb) anti-carbaryl antibodies, immobilized on a poly(vinylidene difluoride) membrane serving as the filtration surface, are used, and the antigen-antibody interaction is enzymatically detected. EXPERIMENTAL SECTION Chemicals. Analytical grade standard carbaryl in methanol was purchased from Dr. Ehrenstorfer (Augsburg, Germany). Working standards (0.001-200 ng/mL) were prepared by appropriate serial dilutions of carbaryl in phosphate-buffered saline (PBS; 10 mM phosphate, 137 mM NaCl, 2.7 mM KCl, pH 7.4). Monoclonal antibody (mAb) and the hapten 6-[[(1-naphthyloxy)carbonyl] amino] hexanoic acid (CNH) used to prepare the enzymatic tracer were obtained from Integrated Laboratory of Bioengineering (LIB) and previously characterized by ELISA.12 Polyclonal antibody (pAb) and the hapten N-(1-naphthoyl)-6aminohexanoic acid (4) were kindly provided and characterized by Marco et al.13 Bovine serum albumin (BSA), o-phenylenediamine (OPD), 1-naphthol, and Tween 20 were purchased from The Sigma Chemical Co. (St. Louis, MO). Horseradish peroxidase (HRP) was obtained from Boehringer (Mannheim, Germany). A poly(vinylidene difluoride) (0.45-µm pore-sized) membrane (Immobilon-P) was purchased from Millipore Corp. (Bedford, MA). Before the immunofiltration assay, the membrane was preconditioned as suggested by the manufacturer. All other reagents were of analytical-grade. Instrumentation. The immunofiltration device was acquired from Pierce Chemical Co. (Rockford, IL). The device consists of a 96-well sample application plate, a 96-transfer cannula set, and a collection chamber. These three pieces are sealed with silicone gaskets to provide constant flow rates in all wells. The membrane is placed between the gaskets and the sample application plate in the typical 8 × 12 microtiter plate format. Under the membrane, 96 individual cannulae emptied the filtrated solutions into a waste chamber where a typical ELISA plate can be placed. A one-channel peristaltic pump, Minipuls-3 (Gilson, Villiers LeBel, France), controlled the flow rate and provided the vacuum for pulling the reagents through the membrane. Photometric measurements at 490 and 650 nm were recorded using a Victor 1420 multilabel counter reader (Wallac, Turku, Finland). Polystyrene 96-well microtiter plates were from Costar Corp. (Cambridge, MA). (11) Del Carlo, M.; Cagnini, A.; Palchetti, I.; Hernandez, S.; Mascini, M. In Biosensors and Environmental diagnostics; Hock, B., Barcelo´, D., Cammann, K., Hansen, P.-D., Turner, A. P. F., Eds.; B. G. Teubner Stuttgart: Leipzig, Germany, 1998; Chapter 2. (12) Abad, A.; Primo, J.; Montoya, A. J. Agric. Food Chem. 1997, 45, 14861494. (13) Marco, M. P.; Gee, S. J.; Cheng, H. M.; Liang, Z. Y.; Hammock, B. D. J. Agric. Food Chem. 1994, 41, 423-430.
1906 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
Figure 1. Scheme of the immunofiltration protocol based on sequential competitive enzyme immunoassay.
Immunofiltration Protocol. The basic immunofiltration protocol steps are shown in Figure 1. The assay is based on a sequential competitive enzyme immunoassay procedure. In the first step, the optimum antibody solution (1 µg/mL or 1:2000 for mAb and pAb, respectively) is added and filtered through the membrane and the remaining binding sites blocked with BSA 1% (wt/v). Afterward, the buffer solution in absence of analyte (maximal signal) or with analyte at different concentrations (signal inhibited) was continuously filtered. The unbound antibodies are detected using enzymatic tracers (CNH-HRP or 4-HRP for mAb and pAb, respectively) prepared by the anhydride mixed method.14 Volumes of 0.2 mL/well were used in each step and filtered at a flow rate of 0.2 mL/well/5 min. In all steps, PBS was used as buffer. Finally, five washing steps, with PBS containing 0.05% Tween 20, were carried out. (14) Rajskowsky, K. M.; Cittanova, N.; Desfosses, B.; Jayle, M. F. Steroids 1977, 29, 701-713.
Signal ) {(A - D)/1 + (x/C) } + D B
at the inflection point, C is the x value at the inflection point (corresponding to the analyte concentration giving 50% inhibition of the maximal signal, I50), and D is the asymptotic minimum (background signal). The limit of detection (LOD) was estimated as the concentration of the analyte giving 10% inhibition of the maximal signal. RESULTS AND DISCUSSION Attending to a previous selection and characterization study of several membranes regarding sensitivity and high binding capacity, Immobilon-P was the selected membrane and was used throughout this work.16 Immunofiltration Study Using Monoclonal Antibodies. Prior to immunoconcentration, because of the flow of large volumes through the membrane, the immobilization stability of the antibody was studied. For this purpose, the different volumes of PBS (0.2-5.0 mL) were assayed, and the absorbance signals obtained after tracer addition (see Figure 1) were compared. Absorbance signals were constant even after running 5.0 mL through the membrane, indicating a stable immobilization. For immunoconcentration purposes, a comparison of inhibition curves for carbaryl was done by filtering different volumes (from 0.4 to 5.0 mL) of carbaryl working standards (0.001-200 ng/mL). As volume increased, a displacement of the curves toward lower concentrations of analyte occurred. This fact indicates that the larger the volume filtered the more sensitive was the assay because of the concentration of the analyte by the antibody. The evolution of the sensitivity, estimated as the I50 value, as a function of the solution volume used, is shown in Figure 2. It was observed that under conventional conditions (0.2 mL), I50 reached 0.63 ng/mL, while after running 3.0 mL through the membrane, the I50 value decreased to 0.048 ng/mL, obtaining a (15) Raab, G. M. Clin. Chem. 1983, 29, 1757-1761. (16) Morais, S.; Maquieira, A.; Puchades, R. J. Immunol. Methods 1998, in press.
L
L
Before substrate addition, a microtiter plate was held in the waste chamber in order to collect precisely and quantitatively the product of enzymatic reaction. The substrate solution (0.1 mL/ well) was 20 mg/mL OPD and 0.012% H2O2 in 25 mM citrate and 62 mM phosphate buffer, pH 5.4. After 10 min, the enzyme reaction was stopped by adding 0.1 mL of 2.5 M sulfuric acid, and the resulting colored solution was filtered through the membrane and collected in the wells of the microtiter plate. After taking away the plate from the device, the absorbancies at 490 and 650 nm were read in a microtiter plate reader. For immunoconcentration purposes, the protocol was the same as described above for immunofiltration, but a large range of working volume (0.4-5.0 mL/well) was used. The larger volumes tested represented a huge increase in sample volume when compared to the surface area of the membrane that is exposed. To apply the larger sample volumes, a homemade system composed of 96 individual 10-mL cannulae made from polypropylene was tightened in the immunofiltration system. Inhibition curves were mathematically analyzed15 by fitting experimental results to the following four-parameter logistic equation where A is the asymptotic maximum, B is the curve slope
L Figure 2. Evolution of sensitivity (I50) using mAb (B) and pAb (9) as a function of working volume.
concentration factor of 13. The concentration factor is defined as the relationship between the I50 values achieved by conventional conditions and those obtained after running large volumes of sample through the membrane. The filtration of 5.0 mL through the immunosorbent did not reduce the I50 value (0.037 ng/mL) as expected besides extending the assay. The inhibition curve accomplished by using 5.0 mL theoretically should give a concentration factor of 25 (I50, 0.025 ng/mL); however, a concentration factor of 17 was reached. The largest volumes did not significantly improve the sensitivity, probably as a result of saturation of the antibodies’ active sites and adsorptive losses during the longer contact times. Therefore, a volume of 3.0 mL was chosen because of high sensitivity (I50, 0.048 ng/mL), high concentration factor, and lower span time revealed. This concentration factor (13) will permit the analysis of carbaryl in drinking water below the maximum level allowed in the European Community (0.1 ng/mL). The limit of detection (LOD) was 4.75 ng/ L, and 96 simultaneous measurements are available in approximately 1h and 45 min. Immunofiltration Study Using Polyclonal Antibodies. Once the immunofiltration system using mAb was optimized, the procedure was applied to carbaryl determination using rabbit anticarbaryl sera (pAb). The aim of this study was to demonstrate the flexibility of the immunofiltration method for other antibodyantigen systems as far as immunoconcentration was concerned. As occurred for mAb, when working volume increased, a displacement of the inhibition curves toward lower concentrations of analyte was noted. In fact, the sensitivity (I50) for the 0.2 mL inhibited curve was 30 ng/mL, while for larger volumes (0.4, 1.0, 2.0, and 3.0 mL) it improved (smaller I50). However, small differences in sensitivity were found when comparing the inhibition curves obtained with 1.0-, 2.0- or 3.0-mL flow. In Figure 2, more detailed I50 evolution is drawn for the investigated volumes. The sensitivity did not meaningfully improve when using volumes greater than 1.0 mL. Thus, it was concluded that the optimum working volume to carry out the assay was 1.0 mL, as far as sensitivity (I50, 6.8 ng/mL), concentration factor (4.4), and span time is concerned. Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1907
Table 1. Recovery of Carbaryl from Spiked Water Samples carbaryl added (ng/L) 0 5 20 80 Mean
estuarine water mean ( SDa recovery (ng/L) (%)
bottled water mean ( SD recovery (ng/L) (%)
