Evaluation of a Magnetic Particle-Based ELISA for the Determination

Jan 29, 1996 - A magnetic particle-based ELISA for chlorpyrifos-ethyl was investigated in estuarine water matrices, and the results obtained were corr...
20 downloads 9 Views 286KB Size
Environ. Sci. Technol. 1996, 30, 509-512

Evaluation of a Magnetic Particle-Based ELISA for the Determination of Chlorpyrifosethyl in Natural Waters and Soil Samples

The determination of chlorpyrifos-ethyl together with structurally related compounds in natural waters is usually performed by either liquid-liquid extraction (LLE) or solidphase extraction (SPE) followed by chromatographic techniques, mainly gas chromatography/nitrogen phosphorus detection or -mass spectrometric detection (GC/NPD or GC/MS, respectively) or by using liquid chromatography (LC) (6). Its determination in soil samples usually involves Soxhlet extraction, cleanup, and GC/NPD (4).

ANNA OUBIN ˜ A, JORDI GASCO Ä N, I M M A F E R R E R , A N D D A M I AÅ B A R C E L O Ä *

Enzyme immunoassays (EIAs) are sensitive, reliable, cost-effective, and rapid, being a viable alternative to traditional methods (7). Substances present in water and soil affect ELISA determinations. For example, false positive results may be caused by interferences from structurally similar compounds or natural substances, including halogens or dissolved organic carbon, that weakly interact with the antibody but are present at a sufficiently high concentration to give a false positive value with the assay, as has been shown for atrazine in estuarine waters (8). In these situations, the concentration of the target compounds is uncertain and in any case confirmation by alternative techniques is needed. These problems may be reduced by using a simple and rapid sample preparation technique before ELISA, such as like SPE or LLE followed by a cleanup step (LLE-Florisil) (9). Although applications of EIAs to environmental samples have grown considerably during the last few years, especially for herbicides like atrazine and alachlor, only very few EIAs were developed for chlorpyrifos-ethyl (10, 11). To our knowledge, only one study (from T. S. Lawruk, unpublished results) (12) has evaluated the use of the RaPID (Rapid Pesticide ImmunoDetection) ELISA for the determination of chlorpyrifosethyl. No applications were reported in ref 12 regarding various environmental parameters, e.g., the influence of water filtration, which is of importance in the determination of chlorinated organic compounds in surface waters (13) and/or soil samples. Application of ELISA to the determination of organic compounds in highly contaminated soils [e.g., at the milligram per kilogram level for polychlorinated biphenyls (14)] has grown in the last few years due to the need to develop rapid methods for the determination of contamination of soils following possible spills. In a recent review article (15), the scarcity of available information on the real environmental applications of EIAs was pointed out. The use of commercial EIAs for field testing and for demonstrating single-compound specificity was encouraged.

Department of Environmental Chemistry, CID-CSIC, c/ Jordi Girona, 18-26, 08034 Barcelona, Spain

A magnetic particle-based ELISA for chlorpyrifosethyl was investigated in estuarine water matrices, and the results obtained were correlated with automated on-line solid phase extraction (Prospekt) followed by liquid chromatography/diode array detection. The ELISA test was evaluated before and after filtration of the particulate matter present in estuarine water with and without prior acidification. The ELISA assay was used as a first screening method for determining chlorpyrifos-ethyl in a highly contaminated soil spiked at common termicidal application rate (∼1000 mg/ kg). The ELISA assay permitted us to distinguish soil contamination of chlorpyrifos-ethyl at 50 mg/kg in conjunction with Soxhlet extraction. Substantial saving time was achieved when compared to conventional methods, which require a cleanup step prior to chromatographic determinations.

Introduction Chlorpyrifos-ethyl is a broad-spectrum organophosphorus nonsystemic insecticide, and it is used to control a variety of soil insects. Chlorpyrifos-ethyl is one of the organophosphorus pesticides most used in the United States (1) and Europe (2), for either agricultural or nonagricultural applications, e.g., against household insects such as ants and cockroaches. Chlorpyrifos-ethyl is applied to the soil surrounding building structures as a barrier to termites with higher application doses (1000 mg/kg) as compared to typical agricultural uses (10 mg/kg) (3). As a result of crop and noncrop usage, chlorpyrifos-ethyl residues may be present in a variety of environmental matrices, e.g., crops, commodities, soil, and water. Its environmental relevance is known since it has been found in different sediment samples in tropical marine environments (4). There exists much concern about its stability in environmental samples, being one of the organophosphorus pesticides with a half-live varying from 16 to 72 days in water and soil under environmental conditions (5). As a consequence, chlorpyrifos-ethyl has been included in different priority lists of compounds to be determined in water samples (2). * To whom correspondence should be addressed; telephone: 343-4006118; fax: 34-3-2045904.

0013-936X/96/0930-0509$12.00/0

 1996 American Chemical Society

The present paper follows the research initiated by our group with the evaluation of the RaPID ELISA for atrazine and alachlor in different waters (8, 9, 16). The objectives of this paper were (i) to apply the RaPID ELISA chlorpyrifosethyl assay to evaluate the influence of the filtration step used to eliminate the particulate matter present in the estuarine waters at various pHs, (ii) to correlate the data obtained by ELISA with the use of the on-line SPE (Prospekt) followed by LC/DAD for the validation of the chlorpyrifosethyl assay in water, and (iii) to apply this ELISA test as a first screening device for the determination of chlorpyrifosethyl in a highly contaminated soil that is used as a candidate reference material for interlaboratory studies. Final determination and confirmation of the results is achieved by LC/DAD.

VOL. 30, NO. 2, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

509

Materials and Methods Chemicals. Methanol and acetonitrile were obtained from J. T. Baker (Deventer, The Netherlands). Chlorpyrifos-ethyl was purchased from Polyscience (Niles, IL), Dursban-48 was obtained from Ici-Zeltia, S. A. (Pontevedra, Spain). The RaPID magnetic particle-based ELISA assay from Ohmicron (Newtown, PA) was purchased through J. T. Baker. HPLCgrade water (pH ) 7.00), groundwater (pH ) 7.65), and Ebre estuarine water (pH ) 8.00; salinity 10 g/L) were also used. Water Analysis with HPLC. (a) Sample Preparation. Chlorpyrifos-ethyl spiked samples were prepared with HPLC-grade water and Ebre estuarine water at concentration values from 0.22 to 3 µg/L. Water samples were then split for ELISA and LC analyses to perform the correlation studies. (b) Prospekt-LC/DAD Analysis. Preconcentration of the water samples prior to LC analyses was performed by an on-line SPE (solid-phase extraction) system as reported in previous work from our group (6). A Prospekt (Spark, Emmen, The Netherlands) was used in combination with an LC/DAD system. The automated SPE device (Prospekt) used in this work consists of a cartridge exchange module, a pump (or solvent delivery unit, SDU), and an electrically operated low-pressure six-port valve, which is connected to the gradient pumps. Samples were preconcentrated on 10 mm × 2 mm i.d. disposable precolumns packed with 40 µm C18 (J. T. Baker). The precolumns were conditioned via a solvent delivery unit from Spark. The LC analyses were performed with a Waters 600-MS solvent delivery unit with a 20 µL injection loop and combined with a Waters 996 photodiode-array detector (Waters, Millipore, Bedford, MA). A 25 cm × 4.6 mm i.d. analytical column packed with 5 µm octyl (J. T. Baker) was used. Prior to LC analyses, HPLC water and Ebre estuarine water were filtered through 0.45 µm filters (Millipore) in order to remove suspended particles. A 60 mL water sample was preconcentrated at a flow rate of 3 mL/min and introduced into the chromatographic system. C18 cartridges were conditioned as in previous work (17). Gradient elution was performed as follows: from 40% A (acetonitrile) and 60% B (HPLC water) to 80% A-20% B in 11 min; to 100% A in 3 min; isocratic until 20 min, and then back to initial conditions in 5 min at a flow rate of 1 mL/min. Detection was realized at 290 nm and quantitation was done using external standard calibration. Calibration graphs were constructed for chlorpyrifos-ethyl by analyzing spiked aqueous samples over the concentration range of interest. Repeatability studies were performed preconcentrating 60 mL of HPLC water spiked at 1.5 µg/L through C18 precolumns (n ) 5). Soil Analysis with HPLC. (a) Sample Preparation. The collection of a soil sample (clay 31%, silt 57%, sand 12%, TOC 3.3%, pH 6.8, and cation exchange capacity of 7.1 mequiv/g) from the Ebre Delta (Tarragona, Spain) took place during July 1993. The total amount collected was ∼10 kg. After collection, the sample was frozen and stored at -20 °C, lyophilized, and sieved through a 120 µm sieve. Using an industrial mechanic mixer previously cleaned extensively, the sediment was mixed for 5 days using a kneader and a mixing machine of 100 L capacity. After kneading, the soil was spiked with 10 g of Dursban-48 [formulation of chlorpyrifos-ethyl at 48% p/v (480 g/L)] per kilogram using manual spraying. Homogenization of the soil took place during 2 days by mechanical mixer. Then, 1.0 g of soil sample was Soxhlet extracted for 18 h with 100 mL of methanol. These extracts were split in two portions,

510

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 2, 1996

one for measuring directly by ELISA test and another for the cleanup procedure. Afterward, the extract was concentrated in a rotary evaporator to ∼1 mL, transferred, and carefully evaporated to dryness (nitrogen stream); it was dissolved in 500 µL of n-hexane. The cleanup step was carried out on glass columns (150 mm × 5 mm i.d.) filled in the laboratory with ∼2 g of Florisil. The Florisil was activated overnight at 300 °C, cooled, and deactivated with 2% water; after packing, the column was rinsed with n-hexane before use. The soil extracts dissolved in n-hexane (500 µL) were placed on top of the column and eluted with a mixture of 20 mL of n-hexane/ethyl ether (50:50). The extract was evaporated just to dryness and the residue dissolved in acetonitrile (1 mL) for injection of 20 µL into the LC/DAD and for ELISA determinations (18). To investigate the influence of the matrix at different concentration levels, aliquots of the soil reference candidate were spiked with Dursban-48 to final concentrations of 50, 100, and 500 mg/L chlorpyrifos-ethyl. We used an additional aliquot without Dursban-48 as blank. After Soxhlet extraction, the chlorpyrifos-ethyl was monitored by RaPID ELISA. (b) LC/DAD Analysis. The HPLC system from Gilson (Villiers-le-Bel, France) consisted of two Model 305 pumps, a 811c dynamics mixing chamber, an 805 manometric module, and a Model 1000S diode-array detector (Applied Biosystems). Samples were injected via a 20 µL loop with a Rheodyne valve (Cotati, CA). Columns (25.0 cm × 4.6 mm i.d.) packed with 5 µm LiChrosphere RP-18 from Sugelabor (Madrid, Spain) were used. The determinations of chlorpyrifos-ethyl took place using acetonitrile/water, 80:20 (isocratic) with a flow rate of 0.9 mL/min. The LC/DAD wavelength was set at 220 nm. RaPID-ELISA Analysis. (a) Sample Preparation. Water Samples To Investigate Acidification and Filtration Effects. Two nonfiltered estuarine water samples, one of them acidified with acetic acid glacial 99% to pH ) 4.0 and another one to pH ) 8.0 (natural pH), and two filtered (0.45 µm filters, Millipore) estuarine water samples, one of them acidified to pH ) 4.0 and another to pH ) 8.0 (natural pH), were spiked with 0.5 µg/L chlorpyrifos-ethyl. The samples were stored in the dark at room temperature for 24 h before direct analysis by RaPID ELISA assay. Soil Samples. The samples with and without a cleanup step (∼1000 mg/kg) were redissolved in HPLC-grade water until a final concentration of 1.0 µg/L. Various dilution steps with a final dilution factor of 1:106 were needed. The concentration of solvents was less than 1.0% in the final extract. (b) Competitive Immunoassay Procedure. All samples were assayed according to the RaPID Assay package insert. Briefly, the RaPID magnetic particle-based ELISA has monoclonal antibodies coated on paramagnetic beads. Samples (250 µL) to be analyzed were added to a disposable test tube, along with chlorpyrifos-ethyl-hapten-HRP enzyme tracer (250 µL), and a rabbit anti-chlorpyrifos-ethyl antibody covalently attached to magnetic particles (500 µL/ tube). The standard curve was prepared from calibrators containing known levels of chlorpyrifos-ethyl at 0, 0.22, 1.0, and 3.0 ng/mL. Tubes were vortexed and incubated for 20 min at room temperature. The reaction mixture was magnetically separated using a specially designed magnetic rack. After washing twice with 1 mL of distilled water, the presence of labeled chlorpyrifos-ethyl was detected by adding the substrate solution containing 3,3′,5,5′-tetramethylbenzidine (500 µL/tube). The tubes were vortexed and incubated for another 20 min at room temperature to

TABLE 1

Concentration Values of Chlorpyrifos-ethyl (in µg/L) Obtained with RaPID ELISA Test in Ebre Estuarine Waters (Tarragona, Spain) under Different Conditionsa nonfiltered acidified

nonfiltered nonacidified

filtered acidified

filtered nonacidified

0.50 (4.7)b

0.50 (5.1)

0.55 (1.8)

0.53 (2.8)

Natural acidified waters were buffered to pH ) 4.0. Natural nonacidified waters were at pH ) 8.00. b Coefficient of variation expressed in percentage (n ) 4). a

TABLE 2

Soil Sample Candidate Reference Material from Ebre Delta (Tarragona, Spain) Spiked with Dursban-48a FIGURE 1. Correlation of chlorpyrifos-ethyl between automated online solid-phase extraction (Prospekt) followed by LC/DAD and RaPID ELISA in two matrices: For HPLC-grade water, y ) -0.043 + 1.071x; r ) 0.958. For estuarine water, y ) 0.107 + 0.882x; r ) 0.991.

allow color development. The color reaction was stopped by addition of sulfuric acid (0.5%; 500 µL/tube). The spectrophotometric measurements were determined using the RPA-I analyzer (Ohmicron).

Results and Discussion Validation. Figure 1 shows the correlation between Prospekt on-line LC/DAD and RaPID ELISA data. High correlation with the two matrices, HPLC-grade water (r ) 0.958) and Ebre estuarine water (r ) 0.991), was obtained. Recoveries of 111% and 114% were achieved for HPLCgrade and Ebre estuarine waters using the Prospekt online LC/DAD or 110% and 105%, respectively, using RaPID ELISA. The use of a diode-array system has the advantage that good structural information, provided by the UV spectrum, is obtained. Chlorpyrifos-ethyl has a good chromophore with two maxima at 220-230 and 280-290 nm, as reported (6). For other compounds, which do not exhibit a good chromophore, other ways of confirmation to avoid false positive determinations are needed, e.g., a second HPLC column of different polarity as the cyano type or some type of mass spectrometric detection. The use of automated on-line SPE (Prospekt) enables the possibility of achieving a better, faster, and more accurate result as compared to conventional chromatographic methods (6). Influence of Filtration and pH. Regarding the experiments about particulate matter effect in estuarine waters, we did not observe any significant differences between filtrate and nonfiltrate estuarine waters with or without prior acidification (see Table 1). The results showed the absence of false negatives in these estuarine waters when the direct ELISA measurement was applied. Possibly, this was achieved because the water matrix used does not present an excess of particulate matter. Additionally, the results obtained could be attributed to the low level of chlorpyrifos-ethyl, which approaches current environmental pollution levels. In another experiment, an acid pH (4.0) was used to release any chlorpyrifos-ethyl coupled to the particulate matter. No difference between acidified and nonacidified samples was observed. This ELISA test can be applied in various environmental situations as it is not affected by the acidification of the samples at the

soil sample

LC/DAD

RaPID ELISA

A B

1467.9 (2.3)b

2750.9 (10.3) 3239.6 (3.1)

a Sample A is the result of a triplicate of the sample with a cleanup step measured by LC/DAD and RaPID ELISA test. Sample B is the result of a triplicate of the same sample but without a cleanup step analyzed only by ELISA. Results expressed in milligrams per kilogrram soil. b Coefficient of variation expressed in percentage.

concentration levels studied. In a previous paper (6), the half-life of chlorpyrifos-methyl in water was not affected by the particulate matter when the same type of estuarine water samples were used. Chlorpyrifos-ethyl had an intermediate behavior with a log Koc of 3.4, lower than typical compounds which absorb into the particulate or suspended sediments, such as PAHs and PCBs with log Koc of 5-6 (19), making the present ELISA test very effective. Also, the concentration of the pesticide is of importance. Probably, when increasing the pesticide concentrations above the microgram per liter level used here, e.g., in the case of a spill, or the contact time (4-5 days), the effects of the particulate matter can be of importance. Soil Samples. The soil samples were prepared as candidate reference material for an interlaboratory study which is currently being performed by the Central and South America Network RAQAL (Red para Ana´lisis Quı´micos Ambientales en Ame´rica Latina) with the collaboration of Spanish, Swiss, and German laboratories. Results from soil samples analyzed by LC/DAD and ELISA assay are reported in Table 2. Overestimation by a factor of 2 in the results can be observed, but it must be kept in mind that soils contained a high concentration of chlorpyrifos-ethyl (∼1000 mg/kg). For the ELISA assay, a dilution factor of 1:106 was needed, 1000 times more than for LC/DAD. These dilution factors were so high that an analytical error was present with important variations when measured by two different analytical techniques. On the other hand, looking at the results without a cleanup step, differences in relation to the LC/DAD results (see Table 2) can be observed. This overestimation of the ELISA results was also observed for atrazine in soil (20) and was attributed to potential interferences from coextracted soil compounds with the enzyme conjugate antibody reaction. Overall, the method presented here is useful as a prescreening test and offers similarities with a recently described PCB ELISA assay (14). False positive errors and overestimation for samples containing 2 mg/kg were observed and classified as >5 mg/kg in 39% of the cases. Recently Harrison et al. (21) applied an ELISA test for PCB

VOL. 30, NO. 2, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

511

TABLE 3

Soil Sample Candidate Reference Material from Ebre Delta (Tarragona, Spain) Spiked with the Final Concentrations of Chlorpyrifos-ethyla chlorpyrifos-ethyl 0 50 100 500 average

RaPID ELISA

recovery (%)

0 49 (8.2)b 101 (5.2) 510 (6.6)

98 101 102 100

a

Dursban-48 was used for soil spiking and values were corrected for chlorpyrifos-ethyl; analyzed by RaPID ELISA test. Results expressed in milligrams per kilogram soil (n ) 6). b Coefficient of variation expressed in percentage.

detection at levels of >50 mg/kg, indicating the usefulness of such an approach for rapid detection of the pollutant. This would be the case, for instance, when using this chlorpyrifos-ethyl ELISA assay for investigation of the urban termiticide residues of 1000 mg/kg, commonly applied (3). To check the probable matrix effect, an additional experiment reported at Table 3 was performed. No interferences from the matrix were noted, and this RaPID ELISA test can distinguish between levels of chlorpyrifosethyl in soil varying from 50 to 500 mg/kg. These results imply that this ELISA test can be immediately used after the chlorpyrifos-ethyl is extracted from soil samples without any type of cleanup. No false negative values were observed at the milligram per kilogram levels in soil matrices, which is also of importance for implementation of these EIA tests for environmental screening purposes. The amount of solvent present in the soil sample did not exceed 1% of solvent, though Lawruk et al. (12) found that amounts less than 2.0% acetonitrile and 10.0% methanol did not affect this antibody. Finally, the percentage of acetonitrile in soil samples without the cleanup step was 0.01% and 1500 mg/kg, a higher overestimation of the RaPID ELISA was detected (the concentration varied between 2750 and 3240 mg/kg). Dilution require-

512

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 2, 1996

ments for the samples to be analyzed by chromatographic methods can be defined in advance by the ELISA assay. Validation of the ELISA system using automated online SPE (Prospekt) followed by LC/DAD was achieved. However, many laboratories cannot afford such automated equipment. The cost is about $80 000 (U.S.) as opposed to $5000 (U.S.) for the total cost of the ELISA. So, it is always feasible to estimate the need of each laboratory, and the Prospekt LC/DAD can be placed in a central laboratory for a final positive confirmation, whereas the ELISA tests are used for routine laboratories involved in monitoring of chlorpyrifos-ethyl in water and soil samples.

Acknowledgments This work has been supported by the European Commission (Contract TS3*CT94-0315) and by CICYT (AMB 95-1694CE). E. Martinez is thanked for the preparation of soil samples containing Dursban-48. Research Agreement 7978/CF from the International Atomic Energy Agency (IAEA) is also acknowledged.

Literature Cited (1) Aspelin, A. L. Pesticide Industry Sales and Usage. 1992 and 1993 Market Estimates; 733-K-94-001; U.S. EPA: Washington DC, 1994; pp 1-34. (2) Barcelo´, D. In Environmental Analysis. Techniques, Applications and Quality Assurance; Barcelo´, D., Ed.; Techniques and Instrumentation in Analytical Chemistry 13; Elsevier: Amsterdam, NL, 1993; pp 149-180. (3) Racke, K. D.; Lubinski, R. N.; Fontaine, D. D.; Miller, J. R.; McCall, P. J.; Oliver, G. R. In Pesticides in Urban Environments. Fate and Significance; Racke, K. D., Leslie, A. R., Eds.; ACS Symposium Series 522; American Chemical Society: Washington, DC, 1993, pp 70-85. (4) Readman, J. W.; Weekwong, L. L.; Mee, L. D.; Bartocci, J.; Nilve, G.; Rodriguez-Solano, J. A.; Gonza´lez-Farias, F. Mar. Pollut. Bull. 1992, 24, 398-402. (5) Racke, K. D. Degradation of Organophosphorus Insecticides. In Environmental Matrices in Organophosphates: Chemistry, Fate, and Effects; Academic Press, Inc.: New York, 1992; pp 47-78. (6) Lacorte, S.; Lartiges, S. B.; Garrigues, P.; Barcelo´, D. Environ. Sci. Technol. 1995, 29, 431-438. (7) Van Emon, J. M.; Lo´pez-Avila, V. Anal. Chem. 1992, 64, 79-89. (8) Gasco´n, J.; Durand, G.; Barcelo´, D. Environ. Sci. Technol. 1995, 29, 1551-1556. (9) Gasco´n, J.; Barcelo´, D. Chromatographia 1994, 38, 633-636. (10) Hill, A. S.; Skerrit, J. H.; Bushway, R. J.; Pask, W.; Larkin, K. A.; Thomas, M.;Korth, W.; Bowmerm, K. J. Agric. Food Chem. 1994, 42, 2051-2058. (11) Manclu ´ s, J.; Primo, J.; Montoya, A. J. Agric. Food Chem. 1994, 42, 1257-1260. (12) Lawruk, T. S.; Gueco, A. M.; Jourdan, S. W.; Herzog, D. P.; Rubio, F. M. 108th AOAC International Meeting, Portland, OR. 1994. (13) Herman, J. H.; Smedes, F.; Hofstraat, J. W.; Cofino, W. P. Environ. Sci. Technol. 1992, 26, 2028-2035. (14) Waters, L. C.; Smith, R. R.; Stewart, J. H.; Jenkins, R. A. J. Assoc. Off. Anal. Chem. Int. 1994, 77, 1664-1671. (15) Meulenberg, E. P.; Mulder, W. H.; Stoks, P. G. Environ. Sci. Technol. 1995, 29, 553-561. (16) Gasco´n, J.; Martı´nez, E.; Barcelo´, D. Anal. Chim. Acta 1995, 311, 357-364. (17) Pichon, V.; Hennion, M-C. J. Chromatogr. A 1994, 665, 269-281. (18) Durand, G.; Barcelo´, D. Anal. Chim. Acta 1991, 243, 259-271. (19) Domagalski, J. L.; Kuivila, K. M. Estuaries 1993, 16, 416-426. (20) Goh, K. S.; Hernandez, J.; Powell, S. J.; Greene, C. D. Bull. Environ. Contam. Toxicol. 1990, 45, 208-214. (21) Harrison, R. O.; Melnychuk, N. Int. J. Environ. Anal. Chem. 1995, 59, 179-185. (22) Getzin, L. W. J. Econ. Entomol. 1981, 74, 158-162. (23) Aga, D. S.; Thurman, E. M. Anal. Chem. 1993, 65, 2894-2898.

Received for review April 6, 1995. Revised manuscript received August 27, 1995. Accepted September 9, 1995.X ES950245P X

Abstract published in Advance ACS Abstracts, December 1, 1995.