Development of monoclonal antibody-based

Journal of Environmental Science and Health, Part B (2011) 46, 509–517 C Taylor & Francis Group, LLC Copyright ISSN: 0360-1234 (Print); 1532-4109 (Online) DOI: 10.1080/03601234.2011.583871

Development of monoclonal antibody-based immunoassays for the analysis of bisphenol A in canned vegetables ´ 1 and ANGEL ´ MARIÁ J. MORENO1, PASQUALE D’ARIENZO2, JUAN J. MANCLUS MONTOYA1

Downloaded by [UPVA Universidad Politecnica de], [Angel Montoya] at 01:07 05 July 2011

1

Instituto Interuniversitario de Investigación en Biongenier´ıa y Tecnolog´ıa Orientada al Ser Humano, Universidad Politécnica de Valencia, Spain 2 Dipto. di Scienza dei Metalli, Università di Bologna, Italy

The aim of this work was the development of monoclonal antibodies (MAbs) and highly sensitive immunoassays (ELISAs) to bisphenol A (BPA), a well–known endocrine disruptor able to migrate from the internal coating of cans to food contained inside, particularly vegetables. To produce MAbs to BPA, four synthetic compounds were conjugated to proteins and used as immunizing haptens in mice. By applying hybridoma technology, several MAbs were produced and selected. These antibodies were characterized in the conjugate-coated and in the antibody-coated formats, using both homologous and heterologous conjugates. Three indirect ELISA based on the MAbs showing the highest affinity to BPA were selected. The limit of detection of the most sensitive ELISA was 0.22 nM (0.05 ng/mL), with an I50 value of around 1 nM (0.23 ng/mL). An homologous ELISA based on the MAb BPAB-11 was applied to the simple, direct determination of BPA in the liquid portion of canned artichoke, peas, and sweet corn. Only sample dilution in an appropriate saline buffer was required to minimize matrix effects and to enter the ELISA working range. Recovery and precision of the method were evaluated by spiking the liquid portion of these cans with BPA at 20, 50, and 100 ng/mL. Coefficients of variation were below 20% in most cases. With regard to recovery, the analytical data obtained were also acceptable. This immunoassay has therefore proved its potential as a new tool for the rapid, sensitive and accurate determination of BPA in canned food. Keywords: ELISA, monoclonal antibodies, bisphenol A, polycarbonate plastics, epoxy resins, canned food.

Introduction Bisphenol A (2,2-bis(4-hydroxyphenyl)propane, BPA) is an estrogenic, hormonally active monomer widely employed in the production of epoxy resins and polycarbonate plastics. These plastics are used in many food and drink packaging applications, while the resins are commonly used as internal coating in canned food to prevent undesirable interactions between the metal from the can and the food. Due to an increase in the use of products based on epoxy resins and polycarbonate plastics, exposure of humans to BPA has increased. Possible routes for the exposure of humans to BPA have been reviewed.[1] The environment (aquatic, air and soil) can be one source of BPA contamination for humans, but the primary route is through food contaminated with BPA. In this sense, recent studies have shown that the residual non-polymerized BPA is easily released from plastic wastes[2] and can linings[3], with the ´ Address correspondence to Dr. Angel Montoya, Instituto In´ en Biongenier´ıa y Tecnolog´ıa teruniversitario de Investigacion Orientada al Ser Humano (I3BH), Universidad Politécnica de Valencia, Camino de Vera, s/n, Edificio 8B, access N. 46022Valencia, Spain; E-mail: [email protected]

amount of BPA leached depending on temperature processing, liquid composition or pH.[4] Low levels of BPA have already been detected in canned foods from the aqueous portion, from the solid portion and from the whole content of the can.[5–8] Yoshida and co-workers,[9] who separately analysed the aqueous and the solid portions of canned food, reported that BPA was mainly detected in the solid portion of canned food and found at the maximum level of 95 ng/g (11 µg per can). These results suggested that BPA leached out of can surfaces might migrate from the aqueous portion to the solid portion during storage in some types of foodstuffs. All of these considerations led to the EU Commission in 2004 to establish a specific migration limit (SML) for BPA from food contact plastic materials of 600 ng/g. On the other hand, due to its widespread use and to the endocrine disrupting and toxic properties of BPA, the revision and the optimization of existing methods and the research of new others are necessary for both control of compliance with current legislation and for assessment of human exposure. Separation, identification and quantification of BPA have frequently been performed by gas chromatographymass spectrometry (GC-MS) and liquid chromatography with mass spectrometry, fluorescence and electrochemical


510 detection. These chromatographic methods have recently been reviewed[10] and they are highly sensitive and specific. However, sample preparation still constitutes the key step for the determination of BPA in food and it is the origin of the main drawbacks of the available methodologies. Immunoassays are easy to perform, do not require qualified personnel and offer a cheap and quick alternative to traditional chromatographic methods. Recently, ELISA methods based on monoclonal[11, 12] and polyclonal[13–15] mammalian antibodies and chicken immunoglobulins[16] were developed for the determination of BPA. The limits of detection ranged from 0.1 to 200 ng/mL, depending on the immunogen and the type of antibody. Other methods such as immunosensors[17–20] and chemical sensors[21, 22] have arisen as new tools complying with the current requirements for the analysis of micropollutants in the environment. In the present work we describe the production of monoclonal antibodies to BPA and the subsequent development of an enzyme-linked immunosorbent assay for this compound. The application of this immunoassay to the determination of BPA in canned vegetables is also presented, and its analytical performance is discussed to prove its suitability for the simple detection of the contaminant with a high sample throughput.

Materials and methods Chemicals and instruments Fluka and Sigma-Aldrich products were purchased from Sigma-Aldrich Qu´ımica, S.A. (Madrid, Spain). Analytical standard of bisphenol A, bis-4(hidroxyphenyl)-methane, buthilated hidroxy-anisole, 4-hidroxydiphenylmethane, 4,4-ethylidenebisphenol and 4-cumylphenol were from Sigma-Aldrich. Working stock solutions were prepared in N,N-dimethylformamide and stored at −20◦ C. Analytical grade solvents and silica gel 60, 0.06-0.2 mm, for column chromatogrphy, were from Scharlau (Barcelona, Spain). Ethyl chloroacetate was from Fluka. 4,4-Bis(4-hydroxyphenyl)-valeric acid (BPVA), ethyl 3bromopropionate, ethyl 6-bromohexanoate, Ovalbumin (OVA), o-phenylenediamine (OPD), shephadex G-50, and Freund’s complete and incomplete adjuvants were obtained from Sigma-Aldrich. Bovine serum albumin fraction V (BSA), enzyme-immunoassay grade horseradish peroxidase (HRP), hybridoma fusion and cloning supplement (HFCS), and polyethylene glycol (PEG) 1500 were purchased from Roche Diagnostics (Barcelona, Spain). Peroxidase-labeled rabbit anti-mouse immunoglobulins were obtained from Dako (Glostrup, Denmark). Culture media (high-glucose Dulbecco’s Modified Eagle’s medium with GLUTAMAX I and sodium pyruvate, DMEM), fetal calf serum (Myoclone Super Plus), and supplements were

Moreno et al. from Gibco (Paisley, Scotland). Culture plasticware was from Corning (New York, USA). P3-X63-Ag8.653 mouse plasmacytoma line was from American Tissue Type Culture Collection (Rockville, MD). Polystyrene culture plates (High Binding Plates, cat. no. 3590) were from Costar (Cambridge, USA). ELISA plates were washed with a 96 PW microplate washer from SLT Labinstruments GmbH (Salzburg, Austria), and absorbances were read in dual-wavelength mode (490–650 nm) with a Emax microplate reader from Molecular Devices (Sunnyvale, U.S.A.). 1H nuclear magnetic resonance (NMR) spectra were obtained with a Varian Gemini 300 spectrometer (Sunnyvale, CA), operating at 300 MHz for 1H and at 75 MHz for 13C. Chemical shifts are given relative to tetramethylsilane. Ultraviolet-visible reported spectra were recorded on a UV-160A Shimadzu spectrophotometer (Kyoto, Japan).

Hapten synthesis The structures of the haptens used in this work are depicted in Figure 1. BPAA, BPAB and BPAH were prepared by introduction of an alkyl chain spacer, ending in a carboxylic acid, by O-alkylation of the hydroxyl group of the BPA molecule. BPVA is commercially available.

Synthesis of 6-[4-(1-(4-hydroxyphenyl)-1-methylethyl) phenoxy] hexanoic acid (BPAH) To 30 mL of dry acetone, stoichiometric amounts (10 mmols) of 2,2-bis(4-hydroxyphenyl) propane, potassium carbonate, and ethyl-6-bromohexanoate were added. After reflux for 12 h, the mixture was filtered and the solvent was removed under reduced pressure. The residue was dissolved in 30 mL of ethyl acetate, washed with 1M HCl (2 × 10 mL), 1 M NaHCO3 (2 × 10 mL), and 4 M NaCl (2 × 10 mL), and finally dried over Na2 SO4 . After evaporation of the solvent, the residue was subjected to silica gel column chromatography (dichloromethane/methanol, 15:1) to give BPA ethyl ester (1.8 g, 57%). An amount measuring 4 mL of 4 M NaOH (22 mmols) was added to the residue dissolved in 8 mL of ethanol and the solution was stirred while heated under reflux for 1.5 h. Then, the solution was slightly concentrated and 10 mL of distilled water was added. The aqueous phase was washed with ethyl acetate (2 × 15 mL) and acidified with concentrated hydrochloric acid, extracted with ethyl acetate (3 × 10 mL), and dried over Na2 SO4 . The oily product obtained after solvent evaporation was crystallized from a mixture of hexane and ethyl acetate (3:1) to obtain 430 mg of the pure hapten: 1 H NMR (DMSO) ∗ : 1.52–1.56 (m, 6 H, CH2 -CH2 -CH2 ), 1.54 (s, 6 H, 2 CH3 ), 2.19 (t, 2 H, CH2 -COOH), 3.89 (t, 2 H, O-CH2 ), 6.62–7.08 (s, 2 H, aromatic). From appropriate spacers, the following compounds were also synthesized by O-alkylation of BPA:

511

Immunoassay of bisphenol A in canned vegetables CH3 OH

C

OH

CH3

Bisphenol A

CH3


OH

C

CH3 O

CH2COOH

OH

C

CH3

CH3

BPAA

BPAB

C

CH2CH2CH2COOH

CH3

CH3 OH

O

O

CH2CH2CH2CH2CH2COOH

OH

C

OH

CH2

CH3

CH2

BPAH

COOH

BPVA

Fig. 1. Chemical structures of bisphenol A and of the haptens used to develop the immunoassay.

2-[4-(1-(4-hydroxyphenyl)-1-methylethyl) phenoxy]) acetic acid (BPAA) 1 H NMR (DMSO) ∗ : 1.55 (s, 6 H, 2 CH3 ), 4.61 (s, 2 H, OCH2 ), 6.63–7.10 (m, 8 H, aromatic). 4-[4-(1-(4-hydroxyphenyl)-1-methylethyl) phenoxy] butyric acid (BPAB) 1 H NMR (DMSO) ∗ : 1.54 (s, 3 H, 2 CH3 ), 1.91 (q, 2 H, -CH2 -), 2.34 (t, 2 H, CH2 -COOH), 3.90 (t, 2 H, O-CH2 ), 6.62–7.09 (m, 8 H, aromatic). Preparation of hapten-protein conjugates Preparation of immunizing conjugates. BPAH hapten was covalently attached to BSA using the modified active ester method[23] and BPAA, BPAB, and BPVA were attached to BSA using the mixed-anhydride method.[24] According to the first method, 15 µmol of BPAH was reacted with stoichiometric amounts of N-hydroxysuccinimide and dicyclohexylcarbodiimide in 0.1 mL of DMF overnight at room temperature. After centrifuging, 0.080 mL of the clear supernatant containing the active ester was dissolved in 0.310 mL of DMF and slowly added to a solution of 15 mg of BSA in 0.6 mL of 0.2 M borate buffer, pH 9. The initial hapten to protein molar ration in the mixture was 50:1. The mixture was allowed to react at room temperature for 3 h with stirring, and finally the conjugate was purified by gel filtration on Sephadex G-50 using 100 mM sodium phosphate buffer, pH 7.4, as eluant. The extent of coupling of the hapten to BSA was estimated by the deter-

mination of the number of protein free amino groups before and after conjugation, using 2,4,6-trinitrobenzenesulfonic acid (TNBS) as the titration reagent. The hapten to protein molar ratio was evaluated as 33. According to the mixedanhydride method, 22.5 µmol of BPAA, BPAB, and BPVA were treated with a 20% molar excess of tri-n-butylamine and isobutyl chloroformate in 0.150 mL of DMF for 30 min at room temperature. To a volume of the reaction mixture containing 12.5 µmol of activated haptens was added DMF up to 0.4 mL, and was slowly added to a solution of 15 mg of BSA in 0.6 mL of 0.2 M borate buffer, pH 9. The initial hapten to protein molar ration in the mixture was 55:1. After stirring for 2 h at room temperature, the conjugate was purified as above. The hapten to protein molar ratio in the conjugate was evaluated as 22, 27 and 25 for BSA-BPAA, BSA-BPAB and BSA-BPVA, respectively.

Preparation of the coating conjugates. BPAH hapten was covalently attached to OVA using the mixed-anhydride method. BPAA, BPAB, and BPVA haptens were attached to OVA using the modified active ester method. Following the same procedures as before, BPA haptens were first activated and then conjugated to OVA (15 mg/mL) using a hapten/OVA molar ratio of 15–20:1. Coating conjugates were purified by gel filtration. The extent of coupling of the hapten to OVA was determined by the TNBS protocol, described above. The hapten/protein molar ratio of the conjugates was evaluated as 7 for all of them.

512 Preparation of enzyme conjugates. Using the modified active ester procedure described above, all BPA haptens were first activated and then conjugated to HRP (5 mg/mL) using a hapten/HRP molar ratio of 14:1. Enzyme tracers were purified by gel filtration and stored at 4◦ C in a 1:1 mixture of saturated ammonium sulfate and PBS (phosphate buffered saline: 10 mM sodium phosphate, 137 mM NaCl, 2.7 mM KCl, pH 7.4).


Production of monoclonal antibodies to BPA BALB/c female mice (8–10 weeks old) were immunized with BSA-BPAA, BSA-BPAB, BSA-BPAH and BSABPVA conjugates. The first dose consisted of 100 µg of conjugate intraperitoneally injected as an emulsion of PBS and complete Freund’s adjuvant. Three subsequent injections were given at three-week intervals using the immunogen emulsified in incomplete Freund’s adjuvant. After a resting period of at least three weeks from the last injection in adjuvant, mice selected to be spleen donors for hybridoma production received a final soluble intraperitoneal injection of 100 µg of conjugate in PBS, four days prior to cell fusion. Cell fusion procedure was carried out as previously described.[25] Eight to eleven days after cell fusion, culture supernatants were screened for the presence of antibodies that recognized BPA. The screening consisted of the simultaneous performance of a non-competitive and a competitive indirect enzyme-linked immunosorbent assay (ELISA), to test the ability of antibodies to bind the OVA conjugate of the immunizing hapten and to recognize BPA, respectively. The first screening was performed using 1 µM BPA as competitor. Next, wells showing antibodies that sensitively recognized BPA (signal inhibition > 50%) were further characterized, this time using 100 nM BPA. Selected hybridomas were cloned by limiting dilution. Stable antibody-producing clones were expanded and cryopreserved in liquid nitrogen. Antibodies were directly purified from late stationary phase culture supernatants by saline precipitation with saturated ammonium sulfate (1:1, v/v), followed by affinity chromatography on a HiTrap Protein G column (GE Healthcare, Uppsala, Sweden). Purified MAbs were stored at 4◦ C as ammonium sulfate precipitates. ELISA procedures ELISA plates were coated overnight with conjugate or antibody solutions in 50 mM carbonate buffer, pH 9.6. A volume of 100 µL per well was used throughout all assay steps. All incubations were carried out at room temperature. After each incubation, plates were washed four times with washing solution (0.15 M NaCl containing 0.05% Tween 20). Conjugate-coated format. Competitive immunoassays based on the homologous conjugate-coated format were

Moreno et al. performed as follows: ELISA plates were coated overnight with OVA-hapten conjugates. Then, serum, culture supernatant, or antibody dilutions in PBST (PBS containing 0.05% Tween 20) were added and incubated for 1 h. Next, plates were incubated for 1 h with peroxidase-labeled rabbit anti-mouse immunoglobulins diluted 1/2000 in PBST. Finally, peroxidase activity bound to the wells was determined by adding the substrate solution (2 mg/mL OPD and 0.012% H2 O2 in 25 mM citrate and 62 mM sodium phosphate, pH 5.35). After 10 min, the reaction was stopped with 2.5 M sulfuric acid, and the absorbance at 490 nm was read and recorded. For competitive assays, the procedure was the same except that after coating a competition step was introduced by adding 50 µL of the standard or sample followed by 50 µL of the appropriate concentration of antibody (culture supernatant or purified MAb).

Antibody-coated format. In this format, plates were coated overnight with antibodies at 1 µg/mL. Next, the competition was established for 1 h between BPA standards and the selected dilutions of the enzyme tracers. Finally, peroxidase activity was measured as above. Competitive curves were obtained by plotting absorbance values against the logarithm of analyte concentration. Sigmoidal curves were fitted to a fourparameter logistic equation[26] using a commercial software (Sigmaplot, Jandel Scientific). BPA concentration of samples was determined by interpolation of the mean absorbance values on the standard curve run in the same plate.

Preparation of standards and spiked samples From a 1 mM stock solution of BPA in DMF, serial dilutions in this solvent from 200 µM to 2.56 nM were prepared. From these concentrations, standards from 400 nM to 5.12 pM were prepared by 1/500 dilution in PBS. As the assay procedure involved the addition of the same volume of the appropriate immunoreagent concentration, bisphenol A standards in the final assay ranged from 200 nM (45.66 ng/mL) to 2.56 pM (0.00058 ng/mL). Cans of artichoke, sweet corn, and peas were obtained from a local supermarket. The liquid portion of these cans was collected, filtered and spiked with BPA at 20, 50, and 100 ng/mL. For ELISA determination, spiked samples were daily prepared by dilution in PBS, and each dilution was analyzed in triplicate. The BPA concentration in the samples was calculated by averaging values obtained at each dilution.

513

Immunoassay of bisphenol A in canned vegetables Table 1. Summary of cell fusion and hybridoma selection results. Number of wells Immunizing hapten

Fusion No.

Serum I50 (µM)

Seeded

Positive (conjugate)

Competitive (analyte)

No. of cloned hybridomas

1 2 3 1 2 3 1 2 3 1 2 3

9.1 0.7 0.5 7.5 0.5 0.1 0.2 0.05 0.02 78.8 7.3 4.6

570 960 384 384 480 384 672 576 576 384 576 384

1 1 3 6 1 11 18 1 8 13 5 7

1 0 0 1 0 6 16 0 5 0 1 0

0 0 0 1 0 5 4 0 4 0 1 0

BPAA


BPAB

BPAH

BPVA

Results and discussion Immunoassay development Haptens depicted in Figure 1 were conjugated to BSA to obtain immunizing conjugates and to OVA and HRP to obtain assay conjugates. Polyclonal antibodies from mice immunized with BSA conjugates bound competitively BPA with I50 values between 0.02 and 78.8 µM. Three cell fusion experiments were performed for each of the BSA conjugates. Table 1 summarizes the results of cell fusions and hybridoma selection. Screening of hybridoma cultures was carried out by simultaneous non-competitive and competitive ELISAs using BPA as competitor. Supernatants were assayed in ELISA plates coated with two OVA-conjugate concentrations (0.1 and 1.0 µg/mL). Hybridomas giving a

strong positive response in the absence of analyte (A490 > 2.5) were re-assayed by diluting the culture supernatant. Finally, fifteen hybridomas were cloned and stabilized. None of them was derived from BPAA hapten. MAbs produced by each of those hybridomas were purified on a small scale from culture supernatants and were subsequently characterized in terms of affinity to BPA, using homologous and heterologous assay haptens. Following bidimensional titration experiments, competitive ELISAs were performed at subsaturating conditions with all of the MAbs in the conjugate-coated format. Seven selected MAbs were additionally characterized in the antibody-coated format. The most sensitive combinations obtained in both ELISA formats are shown in Table 2. These ELISAs, showing I50 values in the 1.0–21.4 nM range, were performed using both homologous and heterologous haptens.

Table 2. I50 values for Bisphenol A with different immunoreagent combinations. Conjugate-coated format Monoclonal antibody BPAB-11 BPAH-11 BPAH-12 BPAH-14 BPAH-15 BPVA-21 BPAB-31 BPAB-32 BPAB-33 BPAB-34 BPAB-35 BPAH-31 BPAH-32 BPAH-33 BPAH-34 ∗

Antibody-coated format

OVA-hapten

I50 (nM)

HRP-hapten

I50 (nM)

BPAB BPAH BPAH BPAH BPAH BPVA BPAB BPVA BPAA BPAH BPVA BPAH BPAH BPAH BPVA

1.0∗ 1.6 5.8 3.0 2.0 21.4 1.5∗ 3.7 5.3 12.2 4.6 2.1 6.2 2.1 1.0∗

BPAB – – BPAH BPAH – BPAB BPAB – – – BPAH – – BPAB

1.2∗ – – 3.0 2.7 – 1.5∗ 4.9 – – – 1.5∗ – – 1.6∗

Values in bold type indicate the most sensitive combinations.


514

Moreno et al.

Fig. 2. Standard curve for the bisphenol A immunoassay performed with MAb BPAB-11 in the homologous conjugate-coated format.

Considering the sensitivity attained by each combination, both formats provided immunoassays of similar affinity. MAbs showing the lowest I50 values, that is, BPAB-11, BPAB-31, and BPAH-34, were selected for specificity studies.

Specificity of the bisphenol a immunoassays MAb specificity was evaluated by performing competitive assays with several compounds structurally related to bisphenol A in the conjugate-coated format. Crossreactivity data for each compound are summarized in Table 3 and they were calculated according to Equation 1. Percentage of cross − reactivity = (I50 of Bisphenol A/I50 of other compound) × 100 (1)

Analytical characteristics of the ELISA for BPA A standard curve for the BPA competitive immunoassay, based on the MAb BPAB-11 in the conjugate-coated format, is shown in Figure 2. This ELISA was selected because it was the most sensitive immunoassay performed using an homologous hapten. The sigmoidal curve was obtained by averaging four individual standard curves performed in the course of BPA determination in canned vegetables. For standards, the assay working range, calculated as the analyte concentrations providing a 20–80% inhibition of the maximum signal was 0.1–0.8 ng/mL. The mean BPA I50 value was 0.27 ng/mL and the detection limit, estimated as the analyte concentration that reduced absorbance to 90% of the maximum, was 0.05 ng/mL. The sensitivity of this ELISA is similar to, and in some cases higher than, other immunoassays previously reported.[12,14,17]

Analytical performance of the BPA immunoassay A similar recognition pattern was obtained for the three MAbs. They cross-reacted to a greater or lesser extent with the closely related compounds 4,4-ethylidenebisphenol and 4-cumylphenol. In this sense, cross-reactivity values in the 6–19% range and in the 20–74% range were obtained, respectively. On the other hand, bis-(4hydroxyphenyl)-methane was poorly or not recognized (cross reactivity value of 3.5% for the best of the cases). Cross-reactivity of the other tested compounds were