Direct Competitive Enzyme-Linked Immunosorbent Assay for the

Antibodies have been raised against a mixture of the corresponding horseshoe crab hemocyanin conjugates prepared by coupling the carboxylic acid to th...
4 downloads 0 Views 302KB Size
Download PDF
Anal. Chem. 2005, 77, 5283-5293

Direct Competitive Enzyme-Linked Immunosorbent Assay for the Determination of the Highly Polar Short-Chain Sulfophenyl Carboxylates M.-Carmen Este´vez, Roger Galve, Francisco Sa´nchez-Baeza, and M.-Pilar Marco*

Applied Molecular Receptors Group (AMRg), Department of Biological Organic Chemistry, IIQAB-CSIC, Jorge Girona, 18-26, 08034-Barcelona, Spain

A direct enzyme-linked immunosorbent assay for the detection of the short-chain sulfophenylcarboxylic acids (SPCs), the main metabolites of the linear alkylbenzenesulfonates, is reported. Six SPCs (2C3, 2C4, 3C4, 2C5, 3C5, 3C6), differing in the length of the alkyl chain (between C3 and C6) and in the position of the phenylsulfonic group versus the carboxylic group, have been synthesized. Antibodies have been raised against a mixture of the corresponding horseshoe crab hemocyanin conjugates prepared by coupling the carboxylic acid to the lysine amino acid residues. The immunoassay As115/ 3C4-HRP achieves an IC50 value of 23 nM (6.67 µg L-1) and a detection limit of 0.85 nM (0.24 µg L-1), using as standard analyte an equimolar mixture of the six SPCs. The immunoassay has found to work better in media with low or moderate ionic strength (4-30 mS cm-1). The decrease in the detectability produced by the potential formation of SPC salts with divalent cations such as Ca2+ can be prevented by lowering the pH of the assay medium below the pKa value of the SPC carboxylic group and using a buffer chelating with properties such as citrate buffer. The assay can be considered specific for short-chain SPCs since congeners with longer alkyl chains and other pollutants containing sulfonic groups in their structure do not interfere significantly in the assay. Preliminary experiments addressed to evaluate the potential application of this assay to environmental water samples demonstrate the usefulness of the assay. The use of surfactants has been widespread owing to their easy and fast biodegradation substituting natural soaps. Linear alkylbenzenesulfonates (LAS) are anionic surfactants representing 40% of the total surfactants consumed 1 as active agents of detergent formulations. The technical LAS consist of a complex mixture of alkylbenzenesulfonates resulting from the combination of linear alkyl chains of different size (usually ranging between 10 and 14 carbon atoms) and a varying number of positional isomers differing in the location of the aromatic ring (see Figure 1). Their high production volume (i.e., 434 × 103 metric tons in western * To whom correspondence should be addressed. Phone: 93 4006171. Fax: 93 2045904. E-mail: [email protected]. (1) Scott, M. J.; Jones, M. N. Biochim. Biophys. Acta 2000, 1508, 235-251. 10.1021/ac0502910 CCC: $30.25 Published on Web 07/15/2005

© 2005 American Chemical Society

Europe in 20002) and their broad use lead to important discharges from industrial and domestic wastewater sites. Although an efficient removal can take place in wastewater treatment plants (close to 99% in some cases3), their residues are often detected in surface waters, seawater, and sludge.1,4-6 LAS are considered biodegradable products due to their short half-life.7 The limiting step in the biodegradation pathway of LAS is the first oxidation of the terminal methyl group of the alkyl chain (ω-oxidation) to give long-chain sulfophenylcarboxylic acids (SPCs). Oxidations of both terminal methyl groups give sulfophenyldicarboxylic acids that have also been detected in the environment,8-10 although they are much less common. The following consecutive β-oxidations of the alkyl chain occurs much faster, leading to different shortchain SPCs11,12 although less common processes of R-oxidation have been also detected.13 Subsequent splitting of the aromatic ring and loss of the sulfonate group occurs much slower and, according to some authors,11 this would begin on SPCs with alkyl chains of four to five atoms. Considering the complexity of the LAS mixture and the above considerations, it can be assumed that complex mixtures of positional isomers of SPCs with different lengths of short alkyl chains are the final and more persistent metabolites of LAS. Whereas LAS can be found in aquatic emplacements and also adsorbed in particulate matter and sediments, SPCs present a higher polarity and are mainly found in solution. Thus, residues (2) European Committee of Surfactants and their Organic Intermediates (CESIO) 2001 CESIO News 5 http://www.cefic.be/files/Publications/ cesio•5.pdf. (3) Trehy, M. L.; Gledhill, W. E.; Mieure, J. P.; Adamove, J. E.; Nielsen, A. M.; Perkins, H. O.; Eckhoff, W. S. Environ. Toxicol. Chem. 1996, 15, 233-240. (4) Matthijs, E.; Holt, M. S.; Kiewiet, A.; Rijs, G. B. J. Environ. Toxicol. Chem. 1999, 18, 2634-2644. (5) Eichhorn, P.; Flavier, M. E.; Paje, M. L.; Knepper, T. P. Sci. Total Environ. 2001, 269, 75-85. (6) Eichhorn, P.; Rodrigues, S. V.; Baumann, W.; Knepper, T. P. Sci. Total Environ. 2002, 284, 123-134. (7) Perales, J. A.; Manzano, M. A.; Sales, D.; Quiroga, J. M. Bull. Environ. Contam. Toxicol. 1999, 63, 94-100. (8) Di Corcia, A.; Capuani, L.; Casassa, F.; Marcomini, A.; Samperi, R. Environ. Sci. Technol. 1999, 33, 4119-4125. (9) Di Corcia, A.; Crescenzi, C.; Marcomini, A.; Samperi, R. Environ. Sci. Technol. 1999, 33, 4112-4118. (10) Dong, W.; Eichhorn, P.; Radajewski, S.; Schleheck, D.; Denger, K.; Knepper, T. P.; Murrell, J. C.; Cook, A. M. J. Appl. Microbiol. 2004, 96, 630-640. (11) Scho ¨berl, P. Tenside, Surfactants, Deterg. 1989, 26, 86-94. (12) Divo, C.; Cardini, G. Tenside, Surfactants, Deterg. 1980, 17, 30-36. (13) Leon, V. M.; Gomez-Parra, A.; Gonzalez-Mazo, E. Environ. Sci. Technol. 2004, 38, 2359-2367.

Analytical Chemistry, Vol. 77, No. 16, August 15, 2005 5283

Figure 1. LAS biodegradation pathway leading to the formation SPCs. Further degradation of the phenylsulfonic group occurs at lower rate.11,13,24 Attending to the composition of the LAS technical mixture regarding positional isomers, six short-chain SPC compounds have been proposed as representatives of the final products.

have been detected in seawater,13 in groundwater,14 and in drinking and surface waters6,15,16 usually at a low-microgram per liter level although they can reach higher levels. The high levels achieved in some areas, usually associated also with high levels of LAS, are clear indications of deficiently treated or untreated waters. The SPCs congeners found in higher concentrations are those with a medium alkyl length (C7-C9),5,6,10,17 but the shortest ones (eC5) have also been detected.3,17,18 SPCs can also be found in aquatic organisms exposed to LAS.19,20 The widespread use of LAS and their short half-life determine a continuous accumulation of short-chain SPCs in the environment. The toxicity of SPCs is not yet clear, but it seems to be lower than their parent compounds and no evidence of estrogenic effects has been demonstrated.21 SPCs are basically analyzed by LC-MS,5,6,15,16 and less frequently by GC/MS,3,18,22 after derivatization. The high polarity and the complexity of the SPC mixture together with the lack of the corresponding standards have become major drawbacks for their analysis. Moreover, the whole analytic procedure involves extraction/purification steps where an important amount of the shortchain SPC fraction can be lost due to their high polarity.15,16 These aspects may have determined that usually higher concentration values reported for the long-chain SPCs (usually between 7 and 11 atoms of carbon) while occurrence of short-chain SPCs (e5 µg L-1) is seldom reported. It also questions the fact that several authors claim that SPCs with alkyl chains between four and eight carbons could be the more persistent14 or the “key intermediates” of LAS degradation.23,24 (14) Field, J. A.; Leenheer, J. A.; Thorn, K. A.; Barber, L. B., II.; Rostad, C.; Macalady, D. L.; Daniel, S. R. J. Contam. Hydrol. 1992, 9, 55-78. (15) Gonzalez-Mazo, E.; Honing, M.; Barcelo, D.; Gomez-Parra, A. Environ. Sci. Technol. 1997, 31, 504-510. (16) Eichhorn, P.; Knepper, T. P.; Ventura, F.; Diaz, A. Water Res. 2002, 36, 2179-2186. (17) Leon, V. M.; Saez, M.; Gonzalez-Mazo, E.; Gomez-Parra, A. Sci. Total Environ. 2002, 288, 215-226. (18) Ding, W.-H.; Tzing, S.-H.; Lo, J.-H. Chemosphere 1999, 38, 2597-2606. (19) Tolls, J.; Haller, M.; Sijm, D. T. H. M. Anal. Chem. 1999, 71, 5242-5247. (20) Tolls, J.; Haller, M.; Seinen, W.; Sijm, D. T. H. M. Environ. Sci. Technol. 2000, 34, 304-310. (21) Navas, J. M.; Gonzalez-Mazo, E.; Wenzel, A.; Gomez-Parra, A.; Segner, H. Mar. Pollut. Bull. 1999, 38, 880-884. (22) Ding, W.-H.; Lo, J.-H.; Tzing, S.-H. J. Chromatogr., A 1998, 818, 270-279.

5284

Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

As an alternative, immunochemical techniques, running in aqueous media and usually achieving sufficient detectability to allow direct measurements of water samples, can be excellent screening analytical methods for highly water soluble anions such as SPCs. However, production of antibodies against an unknown mixture of substances is challenging and uncertain. With this objective, we have synthesized six positional and constitutional isomers of SPCs possessing alkyl chain lengths between three and six carbon atoms and used them as mixture to raise antibodies and establish an enzyme-linked immunosorbent assay. To our knowledge, this is the first time that an immunochemical method for the analysis of short-chain SPCs is reported. Moreover, the procedures to prepare short-chain SPC, described here also for the first time, can be extremely useful to understand fate and degradation of LAS. EXPERIMENTAL SECTION (A) Chemistry. Chemicals and Instruments. Thin-layer chromatography (TLC) was performed on 0.25-mm precoated silica gel 60 F254 aluminum sheets (Merck, Darmstadt, Germany). Unless otherwise indicated, purification of the reaction mixtures was accomplished by “flash” chromatography using silica gel as stationary phase. 1H and 13C NMR spectra were obtained with a Varian Unity-300 spectrometer (300 MHz for 1H and 75 MHz for 13C) or a Varian Inova 500 spectrometer (500 MHz for 1H and 125 MHz for 13C) (Varian Inc., Palo Alto, CA). Infrared (IR) spectra were measured on a Bomen MB 120 FT-IR spectrophotometer (Hartmann & Braun, Que´bec, Canada). Gas chromatography/ mass spectrometry (GC/MS) was performed on an MD-800 capillary gas chromatograph with an MS quadrupole detector (Fisons Instruments, VG, Manchester, U.K.), and the data are reported as m/z (relative intensity). The ion source temperature was set at 200 °C, a 15 m × 0.25 mm i.d × 0.15 µm (film thickness) DB-225 fused capillary column (J&W, Folsom, CA) was used; He was the carrier gas employed at 1 mL min-1.The injector temperature was set at 250 °C. Exact mass values of the compounds was determined by the service of the Unity of Mass (23) Taylor, P. W.; Nickless, G. J. Chromatogr. 1979, 178, 259-269. (24) Knepper, T. P.; Kruse, M. Tenside, Surfactants, Deterg. 2000, 37, 41-47.

Figure 2. Schemes of the synthetic pathways followed for the preparation of the six short-chain R- and β-SPCs (2C3, 2C4, 2C5, 3C4, 3C5, 3C6) selected as target analytes and of a ω-SPC (5C5).

Spectrometry of the Universidad de Santiago de Compostela (Spain) by electrospray ionization mass spectrometry with a timeof-flight analyzer (ESI-MS-TOF) using a Bruker Biotof II spectrometer. All the chemical reagents used for the synthesis of SPCs were purchased from Aldrich Chemical Co. (Milwaukee, WI). Following the usual nomenclature in the field, SPC are named as XCY-SPC, where X is the position where the phenylsulfonic group is located and Y indicates the length of the alkyl chain. Synthesis of SPCs. Seven different short-chain SPCs (2C3, 2C4, 2C5, 3C4, 3C5, 3C6, 5C5) were synthesized by sulfonation of the aromatic ring of the corresponding phenylcarboxylic acids 2 and 5-10 (see Figure 2). The acids 7-10 were obtained from commercial sources. The acids 2, 5, and 6 were prepared as described below. All spectroscopic and spectrometric data obtained in the characterization of the haptens and the intermediates are available as Supporting Information. The long linear 9C9-SPC used as competitor in the development of the enzyme-linked immunosorbent assay (ELISA) was previously synthesized.25 2-Phenylpentanoic 2. A mixture of phenylacetic acid (5 g, 36 mmol) in MeOH (70 mL) and few drops of concentrated H2SO4 were stirred at room temperature for 2 h until the total disappearance of the starting reagent by TLC. The solvent was evaporated, and the resulting crude was redissolved in Et2O and washed with saturated NaHCO3 and saturated NaCl. The organic (25) Ramon, J.; Galve, R.; Sanchez-Baeza, F.; Marco, M.-P. Submitted to Anal. Chem.

phase was dried with MgSO4, filtered, and evaporated to obtain a yellow oil corresponding to the methyl phenylacetate ester 1 (5.07 g, 92% yield) (see Supporting Information). NaH 95% (760 mg, 28 mmol). The ester 1 (4 g, 27 mmol) in anhydrous THF (10 mL) was added to a suspension of NaH in anhydrous THF (15 mL) placed in a dry round-bottom flask equipped with a refrigerant and under argon atmosphere. Previously, the suspension had been exhaustively washed with anhydrous pentane (3 × 15 mL). Recently distilled 1-iodopropane was then added dropwise (2.9 g, 29.3 mmol), and the reaction mixture was heated to reflux for 15 h until the total disappearance of the starting material by TLC. The mixture was slowly added to a solution of 1 N HCl and extracted with Et2O. The organic phase was dried with MgSO4, filtered, and evaporated to obtain a brown oil that was purified by flash chromatography using a hexane/Et2O gradient as mobile phase to finally isolate 3.6 g of the methyl 2-phenylpentanoate ester (1a) (70% yield) (see Supporting Information). The methyl ester (3 g, 15.6 mmol) obtained was hydrolyzed in MeOH (40 mL) with 1M KOH (31 mL) for 2h until no more ester was observed on the TLC. The MeOH was then evaporated and the residue was redissolved in 1M KOH. The aqueous solution was washed with hexane, acidified with concentrated HCl to pH 1 and extracted with Et2O. The organic phase was dried with MgSO4, filtered and evaporated to dryness to finally obtain 2.41 g of the desired product 2-phenylpentanoic 2 (86.7% yield) 26 (see Supporting Information). Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

5285

3-Phenylpentanoic Acid 5. Methyl diethylphosphonoacetate (6.85 mL, 37 mmol) was added to a stirred suspension of 95% NaH (940 mg, 37, mmol) in THF (40 mL) placed in a round-bottom flask equipped with a refrigerant and under argon atmosphere. The NaH had previously been washed with anhydrous pentane (3 × 15 mL) and dried. Once the suspension was totally dissolved and no more hydrogen formation was observed, propiophenone (4.85 mL, 37 mmol) was added dropwise and the reaction was heated to reflux for 5 h until no evolution was observed by TLC. The mixture of the reaction was washed with saturated NaHCO3 and extracted with Et2O. The resultant organic phase was dried with MgSO4, filtered, and evaporated to dryness to give a yellow residue (6.7 g) consisting of a mixture of the propiophenone, diethyl phosphate, and the Z/E isomers of the 3-phenyl-2-pentenoic acid methyl ester. Hydrolysis of the ester was performed with 1 N KOH (65 mL) in MeOH (75 mL) at reflux for 20 h according to TLC analysis. The resulting acids were extracted as described above to obtain the desired Z/E 3-phenyl-2-pentenoic acid 3 (4.8 g, 67% yield) (see Supporting Information). Palladium on activated carbon (5%, 1.36 g of Pd/C 10%) was added to a solution of 3 (4 g, 22.8 mmol) in absolute EtOH (115 mL). The suspension was purged several times with vacuum/H2 cycles to remove the O2 present in the media and finally was kept under H2 at atmospheric pressure. The reaction was stirred at room temperature overnight, until the disappearance of the starting reagent followed by 1H NMR. The catalyst was removed by filtration and the EtOH was evaporated to dryness to obtain a solid corresponding to the desired product 3-phenylpentanoic acid 5 (3.76 g, 93% yield) (see Supporting Information). 3-Phenylhexanoic Acid 6. As described above for the phenylcarboxylic acid 5, butyrophenone (4.9 mL, 34 mmol) was reacted with methyl diethylphosphonoacetate (6.24 mL, 34 mmol) followed by the hydrolysis of the ester to obtain a mixture of Z/E isomers of 3-phenyl-2-hexenoic acid 4 (4.41 g, 64% yield) (see Supporting Information). The double bond of 4 (3.5 g, 18.5 mmol) was reduced using H2 with Pd as catalyst as described before to obtain the desired product 3-phenylhexanoic acid 6 (3.46 g, 98% yield) (see Supporting Information). 2C3, 2C4, 2C5, 3C4, 3C5, 3C6 and 5C5-SPC. Sulfonation of the phenyl carboxylic acids 2 and 5-10 was performed following a similar procedure as described by Sarrazin et al.27 The corresponding phenylcarboxylic acids (6 mmol, 1 equiv) were added to concentrated H2SO4 (2 mL, 36.7 mmol, 6.2 equiv) placed in a round-bottom flask equipped with a refrigerant and previously heated at 100 °C. The reaction mixture was stirred for 2 h and then slowly poured into H2O (80 mL). The aqueous solution was washed with Et2O (3 × 30 mL) and then neutralized with CaCO3 (6 g, 60 mmol). The solid formed was removed by filtration, and the aqueous solution was evaporated under reduced pressure to dryness to finally obtain the desired SPC as calcium salt accompanied by a certain amount of inorganic salts. Further purification attempts by crystallization in different solvent mixtures were unsuccessful, for which reason the purity was determined by quantitative 1H NMR using p-cresol as internal standard. The (26) Kinbara, K.; Kobayashi, Y.; Saigo, K. J. Chem. Soc., Perkin Trans. 2 1998, 8, 1767-1776. (27) Sarrazin, L.; Arnoux, A.; Rebouillon, P. J. Chromatogr., A 1997, 760, 285291.

5286

Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

spectroscopic data of each SPC and the reaction yields are detailed as Supporting Information. (B) Immunochemistry. General Methods and Instruments. The MALDI-TOF-MS used for analyzing the protein conjugates was a Perspective BioSpectrometry Workstation provided with the software Voyager-DE-RP (version 4.03) developed by Perspective Biosystems Inc. (Framingham, MA) and Grams/ 386 (for Microsoft Windows, version 3.04, level III) developed by Galactic Industries Corp. (Salem, NH). The pH and the conductivity of all buffers and solutions were measured with a pH meter pH 540 GLP and a conductimeter LF 340, respectively (WTW, Weilheim, Germany). Polystyrene microtiter plates were purchased from Nunc (Maxisorp, Roskilde, Denmark). Washing steps were performed on a SLY96 PW microplate washer (SLT Labinstruments GmbH, Salzburg, Austria). Absorbances were read on a SpectramaxPlus (Molecular Devices, Sunnyvale, CA). The competitive curves were analyzed with a four-parameter logistic equation using the software SoftmaxPro v2.6 (Molecular Devices) and GraphPad Prism (GraphPad Sofware Inc., San Diego, CA). Unless otherwise indicated, data presented correspond to the average of at least two well replicates. Chemicals and Immunochemicals. Proteins such as horseshoe crab hemocyanin (HCH), bovine serum albumin (BSA), and horseradish peroxidase (HRP) were obtained from Sigma Chemical Co. (St. Louis, MO). The specific immunoreagents (antibodies and protein and enzyme conjugates) were prepared as described below. Standards of SPCs and the phenylcarboxylic acids used in the cross-reactivity studies were synthesized as described before. LAS standards were a gift from Petresa (Ca´diz, Spain). The other sulfonated compounds were from Aldrich Chemical Co., Fluka (Buchs, Switzerland), and Merck. Buffers. Phosphate-buffered saline (PBS) is 0.01 M phosphate buffer and 0.8% saline solution, pH 7.5. PBST is PBS with 0.05% Tween 20. PBST (0.2T) is PBS with 0.2% Tween 20. Citrate (0.2T) is a solution of 0.04 M sodium citrate with 0.2% Tween 20 at pH 4.5. Borate buffer is 0.2 M boric acid-sodium borate, pH 8.7. Coating buffer is 0.05 M carbonate/bicarbonate buffer, pH 9.6. Citrate buffer is a 0.04 M solution of sodium citrate, pH 5.5. The substrate solution contains 0.01% 3,3′,5,5′-tetramethylbenzidine and 0.004% H2O2 in citrate buffer. Preparation of the Immunogens and the BSA Homologous Antigens. Mixed Anhydride (MA) Method. The conjugation of 2C3-, 2C4-, 2C5-, 3C4-, 3C5-, and 3C6-SPC through their carboxylic group to lysine residues of HCH and BSA was carried out by the MA method following similar conditions previously described28 by reacting tri-n-butylamine (23 µL, 96 µmol) and isobutyl chloroformate (11.4 µL, 88 µmol) with a stirred solution of the SPC (80 µmol) in anhydrous dimethylformamide (DMF, 400 µL) cooled to 4 °C in an ice bath. The solution was split in two parts, reacted with HCH (20 mg) and BSA (20 mg) in 1.8 mL of borate buffer, and left under stirring for 4 h at room temperature. After purification of the conjugate by dyalisis, stock solutions of 1 mg mL-1 were prepared in PBS buffer and stored in aliquots at -40 °C. Working aliquots were stored at 4 °C in 10 mM PBS at 1 mg mL-1. Hapten densities of the BSA conjugates were determined by MALDI-TOF-MS by comparing the molecular (28) Ballesteros, B.; Barcelo´, D.; Camps, F.; Marco, M.-P. Anal. Chim. Acta 1997, 347, 139-147.

weight of the standard BSA and that of the conjugates. Preparation of the Enzyme Tracers (ET). Active Ester Method. Haptens 2C3-, 2C4-, 2C5-, 3C4-, 3C5-, 3C6-, 5C5-, and 9C9-SPC were coupled to the lysine amino acid residues of HRP through their carboxylic group using the active ester method as previously described29 by reacting the haptens (10 µmol) dissolved in anhydrous DMF (140 µL) with dicyclohexylcarbodiimide (50 µmol in 30 µL of DMF) and N-hydroxysuccinimide (25 µmol in 30 µL of DMF) and adding the active ester to a solution of HRP (2 mg) in borate buffer (1.8 mL). Hapten Density Analysis. Hapten densities of the BSA conjugates were calculated by MALDI-TOF-MS by comparing the molecular weights of the native proteins to that of the conjugates. MALDI spectra were obtained by mixing 2 µL of the freshly prepared matrix (trans-3,5-dimethoxy-4-hydroxycinnamic acid, 10 mg mL-1 in CH3CN/H2O 70:30, 0.1% TFA) with 2 µL of a solution of the conjugates or proteins (5 mg mL-1 in MilliQ water). The hapten density (δ hapten) was calculated according to the following equation: {MW(conjugate) - MW(protein)}/MW(hapten). Polyclonal Antisera. Three female New Zealand white rabbits (115, 116, 117), weighting 1-2 kg were immunized with an equimolar mixture of the six immunogens prepared (2C3, 2C4, 2C5, 3C4, 3C5 and 3C6 coupled to HCH, SPCMIX-HCH) following the immunizing protocol described elsewhere.28 The evolution of the antibody titer was assessed by measuring the binding of serial dilutions of the antisera to a microtiter plate also coated with an equimolar mixture of the BSA conjugates (SPCMIX-BSA). After an acceptable antibody titer was observed, the animals were exsanguinated and the blood was collected on Vacutainer tubes provided with a serum separation gel. Antisera were obtained by centrifugation and stored at -40 °C in the presence of 0.02% NaN3. ELISA Development and Evaluation. Screening of the Antisera (As) and Enzyme Tracers. The avidity of the antibodies in the serum of each animal was determined using twodimensional checkerboard titration experiments by measuring the binding of serial dilutions (1/500-1/32 000, 2 to 0.031 25 µg mL-1 in PBS) of the enzyme tracers 2C3-, 2C4-, 2C5-, 3C4-, 3C5-, 3C6-, 5C5-, and 9C9-HRP to microtiter plates coated with 12 different dilutions of each sera (1/500-1/512 000 in coating buffer). Optimal concentrations for enzyme tracer and antisera dilution were chosen to produce absorbances around 0.7-1 units of absorbance after 30 min at room temperature. General Direct ELISA Protocol. The plates were coated with the As (100 µL/well in coating buffer) overnight at 4 °C covered with adhesive plate sealers. The day after, the plates were washed four times with PBST (300 µL/well) and the solutions of the analyte (50 µL/well in PBST; only PBST for zero analyte) and the ET (50 µL/well in PBST) were added and incubated for 30 min at room temperature. The plates were washed and the substrate solution was added (100 µL/well) and incubated 30 min protected from light at room temperature before the enzymatic reaction was stopped by adding 4 N H2SO4 (50 µL/well). The absorbances were read at 450 nm. The standard curve was fitted to a four-parameter logistic equation according to the following formula: Y ) {(A - B)/[1 - (C/x)D]} + B, where A is the (29) Gasco´n, J.; Oubin ˜a, A.; Ballesteros, B.; Barcelo´, D.; Camps, F.; Marco, M.P.; Gonza´lez-Martinez, M.-A.; Morais, S.; Puchades, R.; Maquiera, A. Anal. Chim. Acta 1997, 347, 149-162.

maximal absorbance, B is the minimum absorbance, C is the concentration producing 50% of the maximal absorbance, and D is the slope at the inflection point of the sigmoid curve. Some physicochemical data such as pH, conductivity, and the amount of detergent of the media and also time-related parameters affecting the final features of the assay were evaluated in the chosen combination. Optimized Direct ELISA. Microtiter plates were coated with As115 in coating buffer (1/8000, 100 µL/well) covered with adhesive plate sealers overnight at 4 °C. The following day, the plates were washed (four times with PBST, 300 µL/well). An equimolar mixture of the six SPCs (SPCMIX) was considered as the standard analyte. The calibration curve (150 µM-2.56 pM prepared in citrate (0.2T) was added to the plate (50 µL/well) and was preincubated for 30 min. The enzyme tracer 3C4-HRP (0.5 µg mL-1 in citrate (0.2T)) was then added to the plates (50 µL/well). After 30 min of competition, the plate was washed and the plates were processed as described in the general protocol. Cross-Reactivity Determinations. Stock solutions of each SPC (6 mM in H2O or DMSO/MeOH) and several structurally related compounds were prepared and stored at 4 °C. Standard curves (150 µM-2.56 pM) were prepared in citrate (0.2T) by serial dilution. Each IC50 was determined in the competitive experiments following the optimized protocol described above. The crossreactivity (CR) values were calculated according to the equation {IC50(SPCs)/IC50(cross-reactant)} × 100. ELISA Matrix Effect Studies. Water samples from river, well, and sea were collected from Aigu¨esTortes (Pyrinees), Barcelona, and Pineda (Spain), respectively. Before measurements were made, their conductivity and pH were measured and adjusted to place the sample within the working conditions of the immunoassay by adding 5× citrate (0.2T). The matrixes were serially diluted with citrate (0.2T) and used to prepare standard curves. The parallelism of the sigmoidal curves was compared to that prepared in the assay buffer in order to evaluate the extent of the interferences caused by the matrix. Accuracy Studies with Blind Spiked Samples. Different blind spiked samples prepared in citrate buffer and in river and well water samples were measured by the optimized ELISA at different dilution factors (1/2, 1/4, and 1/8 in citrate 0.2T). Analyses were done in duplicate. The correlation was evaluated by establishing a linear regression between the spiked and the measured values in the case of buffer samples. RESULTS AND DISCUSSION The composition of LAS technical mixtures varies depending on the manufacturing process. Usually, mixtures of alkylbenzenesulfonates with alkyl chains ranging between 10 and 14 carbon units (i.e., Petresa S.A. C10, 12%; C11, 34%; C12, 30%; C13, 22%; C14, 0.4%), each of them composed by positional isomers regarding the position of the benzene ring, are obtained. The phenyl group can be located randomly in any position with the exception of the two terminal methyl groups, and the ratio of the different positional isomers can also vary depending on the conditions and reagents used in the production process.11 Attending to the reported degradation pathway of these compounds,11 an initial oxidation would usually occur at the more distal end of the longer alkyl chain (in respect to the carbon atoms to the benzylic position, even or odd number of carbon atoms) and the subsequent Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

5287

β-oxidation steps would require two linear methylene groups. Based on those assumptions, it has been proposed that a mixture of short-chain SPC congeners having a total alkyl chain length between three and six atoms of carbon (XC3-XC6) and possessing the benzene ring in R (R-SPC) or β (β-SPC) position to the phenylsulfonic group (2Cx and 3Cx, respectively) could be at the end of this degradation sequence, before further breaking up the aromatic ring and the sulfonate group. Therefore, 2C3-, 2C4-, 2C5-, 3C4-, 3C5-, and 3C6-SPCs have been considered as potential and representative final degradation products of LAS (see Figure 1). Determination of these short-chain SPCs by the usual analytical methods is difficult due to their high polarity. In addition, the unavailability of standards has also been one of the reasons contributing to the lack of information regarding the presence of these compounds in the environment. Thus, in this work, we first focused on the synthetic preparation and structural characterization of these analytes. Synthesis of Short-Chain SPCs. The six SPCs considered were synthesized by sulfonation following described procedures23,27,30 of the corresponding phenylcarboxylic acids following the schemes presented in Figure 2. The preparation of 2C3-, 2C4-, and 3C4-SPCs could easily be performed by direct sulfonation of the corresponding commercially available phenylalkylcarboxylic acids 7-9, using concentrated H2SO4 at 100 °C (see Figure 2, scheme 3). Maintaining the temperature was crucial in order to avoid polysubstituted byproducts and to favor the formation of the para isomer. The isolation of the compound from the reaction mixture was performed by preparing the corresponding Ca salts.27 The addition of CaCO3 produced the precipitation of CaSO4 salts in aqueous media that were removed by filtration. 2C5-SPC was obtained from phenylacetic acid, by alkylation of the benzylic position of the methyl ester 1 with 1-iodopropane, followed by hydrolysis of the ester and sulfonation of the aromatic ring (see Figure 2, scheme 1). Finally, 3C5-SPC and 3C6-SPC were prepared from propiophenon and butyrophenone, respectively, through a Horner-Wadsworth-Emmons reaction using methyl diethylphosphonoacetate leading to a mixture of the corresponding Z/E isomers of the alkenes (3 and 4, Z/E ratio of 40:60 determined by 1H NMR) (see Figure 2, scheme 2). The double bond was reduced with H2/Pd to obtain 5 and 6 and finally the sulfonic group introduced using concentrated H2SO4 (see Figure 2, scheme 2). Preparation of the Antibodies. Development of antibodies with the ability to recognize several congeners of the same family is challenging and risky. The strategy usually consists of using a hapten that maximizes the common epitopes in the molecule. The SPC structures possess two clearly different functional groups that can be considered as strong antigenic determinants, the carboxylic acid and the phenylsulfonate group. Under the physiological and usual assay media, the sulfonate group will always exist in its ionized form due to its extremely low pKa value. Previous experience of our group25 already pointed at this last group as one of the most important epitopes. Therefore, we took advantage of the carboxylic group already present in the SPC molecules to prepare the corresponding protein conjugates. The location of the carboxylic group, further from the selected epitope, would (30) Marcomini, A.; Di Corcia, A.; Samperi, R.; Capri, S. J. Chromatogr. 1993, 644, 59-71.

5288 Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

Table 1. Characteristics of the Best Immunoassays Obtained in the Screening of As115, As116, and As117 with the Enzyme Tracersa ET IC50 IC50 As dilb ETc (µg mL-1) Amax Amin (nM) (µg L-1) slope 115 1/12 1/16 1/8 1/12 1/12 116 1/16 117 1/12

2C3 2C4 2C5 3C4 9C9 3C4 3C4

1 0.5 0.5 0.25 0.5 0.25 1

0.719 1.016 1.162 1.328 1.383 1.410 0.727

0.112 758 0.026 632 0.004 754 0.001 205 0.165 738 0.059 1195 0.015 910

189 158 189 51 185 299 228

-0.83 -0.73 -0.64 -0.52 -1.67 -0.96 -0.794

r2 0.9546 0.9839 0.9944 0.9803 0.8533 0.9700 0.9592

a Competitive assays using an equimolar mixture of SPCs as standard analytes. The parameters were extracted from the fourparameter equation used to fit the standard curves. Those assays with IC50 values below ∼350 µg L-1 have been included. The concentration (expressed in µg L-1) has been calculated by considering as molecular weight the average of the six SPCs (MW ≈ 250). b Dilution factor of the antiserum (As) (× 1000). c The enzyme tracers (ET) are HRP conjugates.

maximize recognition of the phenylsulfonate moiety. Each of the short-chain SPCs (2C3, 2C4, 2C5, 3C4, 3C5, 3C6) was covalently attached to lysine residues of the HCH by the mixed anhydride method, and the conjugation reaction was verified by MALDITOF-MS. Hapten densities between 20 and 25 were recorded for the BSA conjugates, indicating a high degree of conjugation. The immunogen consisted on an equimolar mixture of six SPC-HCH conjugates (SPCMIX-HCH). Three rabbits where inoculated following immunization protocols established in our group. The antibody titer of the serum of each animal was tested against SPCMIX-BSA on an indirect ELISA monthly until no significant increase of the titer was observed. The final antisera obtained from each rabbit were named As115, As116, and As117. The avidity of the antibodies for a battery of enzimatic tracers (the short-chain SPCs, and two ω-SPCs 5C5- and 9C9-SPC coupled to HRP conjugates) was evaluated by using noncompetitive direct ELISAs. Hapten 5C5-SPC was prepared in one step by sulfonation of 5-phenylpentanoic acid 10 (see Figure 2, scheme 3) following the same procedure described before. The preparation of 9C9SPC is described elsewere.25 In general, As115 and As116 showed broader recognition (all HRP tracers were recognized except 5C5-HRP) and higher antibody titers than As117 (only 2C3-, 2C4-, and 3C4-HRP tracers were recognized). Competitive Direct ELISA. Those antisera/enzyme tracer combinations showing acceptable titers were used to determine the ability of the antibodies to recognize the analyte in competitive assays. For this purpose, and keeping the idea of obtaining an assay capable of recognizing not only one, but a broad family of short alkyl chain SPCs, the mixture of the six SPCs used in the immunization was also considered as a standard analyte by preparing an equimolar mixture in MilliQ water. Following the general protocol described in the Experimental Section, several competitive assays were obtained, although in many cases the IC50 values were higher than 500 µg L-1 (2000 nM), which were considered not usable for environmental monitoring. Thus, Table 1 shows only the features for the competitive assays with IC50 below 350 µg L-1 (1400 nM). As it can be observed, As115 gave the higher number of usable assays. In all cases, 3C4-HRP was found to be the best enzyme tracer. At the light of the immu-

Figure 4. Standard curves prepared using different buffer systems. Except for the PBST at pH 7.5, the pH values of all the buffers were adjusted to 4.5. Tween 20 was added to the solutions to have a total content of detergent of 0.2%. β,β-DMG: β,β-dimethylglutaric acid. Standard curves were prepared using two well replicates.

Figure 3. Influence of several physicochemical parameters on the SPCs immunoassay (As115/3C4-SPC-HRP) features. (A) Effect of the concentration of Tween 20 in the competition buffer. (B) Effect of the conductivity of the media. (C). Effect of the pH of the media. Left axes show the maximum absorbance (Amax) and the quotient of the Amax and the IC50 (Amax/IC50). The right axes show the values of IC50 expressed in µg L-1. The data shown have been extracted from the four-parameter equation used to fit the standard curves. Standard curves were prepared using two well replicates.

noassay parameters and of the reproducibility observed, immunoassay As115/3C4-HRP was chosen for further evaluation. ELISA Evaluation. After evaluating different times (0-60 min), preincubating the analyte in the antibody-coated plate for 30 min, followed by a competitive step with the ET for 30 min, was found to improve significantly performance of the assay in terms of sensitivity without sacrificing the maximum signal of the assay and without altering significantly other important features such as the slope and the signal-to-noise (S/N) ratio (data not shown). The concentration of detergent (Tween 20) produced a clear influence on the assay parameters (see Figure 3A). Thus, high concentrations of Tween 20 in the assay buffer diminished the values of the IC50 (higher detectability) and increased the maximum signal, whereas the slope and the signal-to-noise (S/ N) ratio remained unchanged. Thus, a concentration of 0.2% was

selected for further experiments. An improvement at higher concentrations of detergent has also been observed in other immunoassays for other high polar analytes such as the dealkylated hydroxytriazines.31 In contrast, this behavior is opposed to the effect generally observed for ELISAs of nonpolar analytes,32,33 in which case the sensitivity improves at lower concentrations of detergent. The increase of the maximum signal allowed diluting further the immunoreagents (As115, 1/16 000 and 3C4-HRP, 0.125 µg mL-1). An increase of the ionic strength (from 4.3 to 115 mS cm-1, 2.5-100 mM in terms of PBS concentration and MilliQ water, 0 mS cm-1) produced a clear negative effect on the immunoassay detectability and in the maximum signal (see Figure 3B). In contrast, the assay performed very well in media with conductivity values between 0 and 30 mS cm-1 although precision was poorer when conductivity was at or below 4 mS cm-1. Attending to these results, further experiments were performed in 10 mM PBS, which produces a conductivity of 16 mS cm-1 within the interval where the immunoassay parameters remain constant. A broad degree of tolerance to media was found regarding pH values of the assay buffer. The assay performed very well within 4.5-10.5 units of pH, although the best detectability was accomplished at acidic pH values. At pH 4.5, the maximum absorbance of the assay was high and the IC50 was much lower (see Figure 3C). Thus, whereas in PBST at pH 7.5 the IC50 was ∼145 nM (36 µg L-1), this value dropped down to 60 nM (15 µg L-1) setting the pH value at 4.5. Since this pH is outside of the buffering interval of the PBS, other buffer systems, such as citrate, acetate/acetic, β,β-dimethylglutaric acid/NaOH, and succinic acid/NaOH, were evaluated, adjusting the pH at 4.5 and maintaining the concentration of Tween 20 at 0.2%. The results shown in Figure 4 demonstrate that, under these assay conditions, only the citrate buffer and the PBS afforded acceptable calibration curves. In fact, the detectability achieved using citrate buffer (pH 4.5, 0.2% Tween 20, conductivity 10 mS cm-1) was 23 nM (5.7 µg L-1), significantly better than in PBS. The significant improvement in the SPC recognition produced after lowering the pH and changing the buffer system can be explained by considering the type of immunogen used, the pKa (31) Sanvicens, N.; Pichon, V.; Hennion, M. C.; Marco, M. P. J. Agric. Food Chem. 2003, 51, 156-164. (32) Sanvicens, N.; Varela, B.; Marco, M.-P. J. Agric. Food Chem. 2003, 51, 39323939. (33) Manclus, J. J.; Montoya, A. J. Agric. Food Chem. 1996, 44, 4063-4070.

Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

5289

Figure 5. Scheme showing the equilibria between the different SPC species potentially present in the aqueous media depending on the pH (7.5 or 4.5) and on the buffer system used (PBS or citrate) and the effect of these on the IC50. Those conditions favoring (i.e, pH e pKaCOOH, chelating agents such as citrate) the presence of the SPCs as single free species can improve the immunoassay detectability. Table 2. Working Conditions and Most Important Features of the Optimized Immunoassay As115/3C4-HRPa As115 dilution [3C4-HRP], µg mL-1 buffer preincubation step competition step Amax Amin IC50, µg L-1 dynamic range, µg L-1 LOD, µg L-1 slope R2

1/8000 0.5 citrate 40 mM, pH 4.5, 0.2% Tween 30 min 30 min 0.940 ( 0.132 0.055 ( 0.018 4.32 ( 1.65b 0.48 ( 0.32 to 43 ( 14 0.15 ( 0.24 -0.619 ( 0.111 0.9793 ( 0.0115

a

The parameters are extracted from the four-parameter equation used to fit the standard curve. The data presented correspond to the average and the standard deviation of 20 calibration curves run on different days. Each curve was built using two-well replicates. The concentrations (expressed as µg L-1) has been calculated by considering as molecular weight the average of the six SPCs (MW ≈ 250). b Expressed in nM, the IC is 17.3 ( 6.6 and the LOD is 0.62 ( 0.49. 50

values, and the fact the SPC standards are calcium salts. In general, antibody binding could be affected by the presence of these ions in the media; however, for the particular case of the SPCs, another explanation can be considered by contemplating all these factors. According to the pKa values of the carboxylic groups (∼4.5, for 3Cx SPCs and 4.15 for 2Cx SPCs), the carboxylate is the predominant form of these compounds at neutral pH, while at pH 4.5 the carboxylic acid form would be more abundant. This neutralized form of the carboxylic group mimics better the immunogen structure used to raise antibodies, since the carboxylic group was blocked by the amide bonds formed while coupling this molecule to the lysine amino acid residues of the proteins. Moreover, as SPCs are calcium salts, at pH 7.5, SPC symmetric (for both carboxylic or sulfonate groups of two molecules) or asymmetric (with one carboxylic and one sulfonate group) dimers (interactions with both anionic charges 5290 Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

Figure 6. Standard curve for the SPCs immunoassay As115/3C4HRP obtained under the optimized conditions established in the protocol described in the Experimental Section. The data presented correspond to the average and standard deviation of 20 assays, run on different days. The curves were prepared using two-well duplicates. The features of the assay are summarized in Table 2.

of a same molecule could also be formed, but seem more unlikely) can be stabilized through the Ca2+ ions. A different situation might be expected at pH 4.5, where the calcium cations could only form dimers through the sulfonate groups (see Figure 5). The SPC molecules are captured by the calcium ion more efficiently at pH 7.5 than at pH 4.5, hampered from interacting with the antibodies as single free molecules, which could explain the improvement of immunoassay detectability. The use of citrate buffer, with recognized strong chelating properties for calcium ions, shifts the equilibrium to the free form, allowing interaction with the antibody and therefore improving even more the immunoassay detectability. A final working protocol was established consisting of a standard coating step of the antibodies on the microplate overnight at 4 °C, a preincubation of the analyte with the antibody for 30 min, followed by 30 min of competition with the enzyme tracer. The assay buffer used was 40 mM citrate buffer (0.2% Tween 20, conductivity 10 mS cm-1, pH 4.5). These conditions as well as the immunoassay features are summarized in Table 2. Figure 6

Table 3. Specificity of the Immunoassay As115/ 3C4-HRP for Some Related Compoundsb

a 2C , 2C , 2C , 3C , 3C , and 3C are the SPCs used as immunogens. 3 4 5 4 5 6 LAS, linear alkylbenzenesulfonates; p-TS, p-toluensulfonic acid; EBS, ethylbenzenesulfonic acid; p-XS, p-xylenesulfonic acid; BDS, benzenedisulfonic acid; 1-Napht, naphthalene-1-sulfonic acid; 2-Napht, naphthalene-1,5-disulfonic acid; 3-Napht, naphthalene-1,3-6-trisulfonic acid. b The percentage of recognition has been expressed as cross-reactivity (CR%) according to the expression [IC50(SPCs)/IC50(cross-reactant)] × 100.

shows an example of the calibration curve obtained under these conditions. A LOD of 0.82 nM (0.20 µg L-1, considering an average molecular weight of 250) was reached, which is suitable to perform studies of the levels of these products in real samples. The specificity of the assay was first evaluated by regarding individual recognition of the six SPCs used as immunogens (see Table 3). Contrary to the expected similar recognition of the six SPCs, the following sequence was observed: 3C6 > 3C5 . 2C5 ≈ 3C4 > 2C4 > 2C3, which shows that the longer the alkyl chain of the SPC, the better was its recognition, especially considering the size of the alkyl group at the benzylic position. Moreover, SPCs with the aromatic ring in the β position (3Cx) seem to be better recognized that those with the phenyl group in R position. (2Cx), which could be explained by a slightly greater steric hindrance, caused by the protein with the distance to the phenylsulfonic group smaller. Thus, except for 3C4-SPC, which only cross-reacted 10% with respect to the equimolecular mixture of the six SPCs, 3C5- and 3C6-SPCs, with the phenylsulfonic in the β position were recognized at 198 and 84%. Thus, despite the superior immunogenic properties of the phenylsulfonic group, the antigenic value of the alkyl chain should not be underestimated. The recognition decreased drastically for the less abundant ω-SPCs, with the phenylsulfonic group in the terminal position. Thus, 5C5-SPC was not recognized at all. The same is applicable for long alkyl chain SPCs such as 9C9- and 12C12-SPCs (see Table 3). These data demonstrate that the ELISA developed can be considered specific for short alkyl chain SPCs. The technical mixture of LAS and p-ethylbenzenesulfonic acid were recognized with cross-reactivity values around 6 and 5%, respectively; ptoluensulfonic acid (p-TS) with only a methyl group in para position was even less recognized. Other environmental pollutants with sulfonic groups such as naphthalenesulfonates were tested, and no significant recognition was observed in any case. Finally, the corresponding alkylphenylcarboxylic acids, lacking the sulfonic group, were also evaluated but did not interfere in the assay, indicating that the presence of the sulfonic group in the para position is necessary for the recognition. Immunoassay Matrix Effects. The application of this technique to measure real aqueous samples was preliminarily studied regarding potential interferences caused by the different types of water matrixes. For this purpose sea, river, and well water samples were collected. The pH of these waters was in all the cases ∼7.5, and the conductivity was very low for the river and well water samples (0 and 2.1 mS cm-1, respectively) whereas the seawater showed a higher value (55 mS cm-1). Standard curves were prepared by spiking these samples with known concentrations of the SPC standard mixture. As it can be observed in Figure 7, direct analysis of those samples is not possible since a very different behavior is observed when those curves were compared with the ones prepared in the assay buffer. For the case of the seawater, a diminution of the maximal absorbance and a decrease of the assay detectability was observed. Buffering the samples by adding 20% 200 mM citrate buffer to adjust pH and conductivity readily improved significantly immunoassay performance. The remarkable effect of the buffer observed for the mountain spring clean river water has been interpreted by a potential high content of calcium or other bivalent cations in this water sample (sample was collected from a karstic area). Further dilutions of these solutions Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

5291

Figure 7. Matrix effect produced by sea, river, and well water samples on the SPC ELISA. The pH and the conductivity were adjusted to the assay conditions by buffering with 200 mM citrate and maintaining the Tween 20 to a final concentration of 0.2%. Serial dilutions were done with citrate (0.2T) buffer. Buffering the samples significantly improves immunoassay performance. Table 4. Results from the Accuracy Studies spiked concn (nM)

measd concn (nM)

% recovery

30 60 120 300

Buffer Samples 32.8 ( 2.7 69.5 ( 3.4 147.0 ( 18.7 233.6 ( 19.7

109 115 122 78

8 40 200

Well Samples 7.72 ( 1.7 41.1 ( 1.7 153.6 ( 25.7

96 103 77

River Samples 1.85 ( 0.27 8.1 ( 0.9 43.6 ( 9.2

115 101 109

1.6 8 40

a The accuracy was evaluated by preparing blind spiked samples in buffer and in natural water samples and measuring them with the As115/3C4-HRP direct ELISA. Each value corresponds to the mean of two replicates.

were carried out in 40 mM citrate (0.2T). For the case of the seawater sample, it was necessary to further dilute more of the sample with citrate (0.2T) in order to diminish the conductivity of the media. However, after a 32 times dilution of the sample (11.3 mS cm-1), the values of the IC50 (55 nM) were still slightly higher than those obtained in buffer (20 nM), which indicates the necessity to perform some type of sample treatment in order to perform measurements without compromising the detectability of the method. We must also consider the possibility that SPCs from domestic effluents could be present in this sample since it was collected close to an urban area. This fact has to be studied in more detail on further experiments. In contrast, dilution of the river and well water samples between 2 and 4 times with citrate (0.2T) after buffering the samples was sufficient for direct analysis. Immunoassay accuracy was preliminarily assessed by measuring buffer and natural water samples spiked with the SPC mixture. Results shown in Table 4 demonstrate that the analyses provided good recovery values, between 96 and 115%, pointing to an acceptable accuracy level. Only for those samples spiked at higher concentration levels (>200 nM), the recoveries were lower but still acceptable (>75%). This lower recovery could be related to a higher number of dilution steps performed to place the samples within the dynamic range of the ELISA; however, further experiments have to be performed to prove this fact. More exhaustive validation studies using chromatographic methods and increasing 5292 Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

the number of environmental samples analyzed are being performed in order to provide more objective data. CONCLUSIONS The analysis of highly polar substances such as the short-chain (3-6 carbon units) SPCs by conventional techniques is troublesome due to the difficulties and low yields encountered during the extraction and preconcentration steps. Moreover, there has been a lack of SPC standards of these alkyl sizes and placement of the phenylsulfonate group at the R and β positions in respect of the carboxylic group. Consequently, environmental monitoring of the shortest SPCs has been rarely reported. Immunochemical techniques are not subject to these previous procedures since they are performed in aqueous media and usually the detectability achieved is sufficient to avoid preconcentration. For the first time, we have reported here an immunochemical method for the rapid detection and quantification of short-chain SPCs. Six congeners (2C3-, 2C4-, 2C5-, 3C4-, 3C5-, and 3C6-SPC) have been synthesized and used as standards and also prepared as an immunogen that consists of an equimolar mixture of them. The direct ELISA developed is specific for short-chain SPCs and can detect that mixture with a limit of detection of 0.15 µg L-1 (0.6 nM). However, due to the fact that not all SPC congeners synthesized have identical immunogenic properties, these are not equally detected by the antibodies raised. Thus, 3C6-SPC and 3C5-SPC are much more recognized (198 and 84% cross-reactivity in respect to the equimolar mixture) while 2C3-SPC is almost not detected (2% cross-reactivity). Moreover, recognition is higher for the more abundant short-chain R- and β-SPC, while the less probable ω-SPCs are not recognized. The assay reported here is highly tolerant to media with different pH values (pH 4-10), although the best sensitivity levels are achieved in acidic media (pH 4.5). It performs well in buffers with a moderate ionic strength (0-30 mS cm-1). It has been found that bivalent cations may interfere with the assay and that this effect can be minimized by using a buffer with chelating properties, such as the citrate buffer. Finally, the preliminary matrix effect studies performed with natural water samples indicates that, except for the seawater sample, small dilution factors can be sufficient to avoid the interference. Preliminary experiments that have been performed spiking natural water samples indicate the potential of the method presented here as a screening tool in environmental monitoring programs. Future work will be directed to the validation of the assay with chromatographic methods.

ACKNOWLEDGMENT This work has been supported by the Ministry of Science and Technology (Contract TEC2004-0121-E) and by the European Community (IST-2003-508774). We thank PETRESA for providing samples of the technical LAS. SUPPORTING INFORMATION AVAILABLE Spectroscopic data (1H and 13C NMR, IR) and HRMS of the

compounds described in the paper. This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review February 17, 2005. Accepted June 16, 2005. AC0502910

Analytical Chemistry, Vol. 77, No. 16, August 15, 2005

5293