IC-ELISA and immunochromatographic strip assay based monoclonal antibody for the rapid detection of bisphenol S

ABSTRACT

A highly sensitive and specific monoclonal antibody against bisphenol S (BPS) was prepared. The derived BPS was coupled to keyhole limpet hemocyanin as the immunogen and ovalbumin as the coating antigen. Based on monoclonal antibodies, an indirect competitive enzyme-linked immunosorbent assay (ic-ELISA) was established, and the conditions of action were optimized. The half maximal inhibitory concentration (IC₅₀) of BPS was 0.228 ng/mL and the linear range was 0.064–0.810 ng/mL. In the recovery test for the milk samples, the recovery rates were in the range of 89%–95% and coefficient of variation ranged from 1.0% to 6.0% respectively. In addition, the cut-off value of the immunochromatographic strip detection method in milk samples was 5 ng/mL. Therefore, both methods are suitable for use in milk samples. Furthermore, this immunochromatographic strip detection method is suitable for on-site testing and screening of very large samples.

KEYWORDS:

• Bisphenol S
• monoclonal antibody
• ic-ELISA
• milk

Introduction

Bisphenol S (4,4'-Sulfonyldiphenol), (BPS) is an industrial chemical and a structural analogue of bisphenol A, which is mainly used as a raw material for the production of polycarbonate, epoxy resin, polyesters and phenolic resin, as well as polysulfone and polyethersulfone (Shan, Yao, Wen, & Shao, 2019). In addition, BPS is also a vital raw material for medical polymer materials and various plastics (Qiu et al., 2019), which are widely used as packaging in the food industry (Giacomo et al., 2018; Siracusa, Yin, Measel, Liang, & Yu, 2018). In recent years, BPS has been widely used as a substitute for bisphenol A (Simoneau, Valzacchi, Morkunas, & Eede, 2011). However, many studies have shown that BPS has endocrine disrupting activity along with bisphenol A (Aung, Ferguson, Cantonwine, McElrath, & Meeker, 2019), which can disrupt the reproductive system (Eladak et al., 2015; Ullah, Jahan, Ain, Shaheen, & Ahsan, 2016; Ullah, Pirzada et al., 2018). Japan and Korea have explicitly stipulated that the migration of BPS in food simulators should not exceed 0.05 mg/L. Furthermore, both The European Union Plastics Regulation (EU) No.10/2011 and The China National Standards (GB 9685-2008) have stipulated that the migration limit of BPS in food simulators should not exceed 0.05 mg/kg.

The reported detection methods for BPS include liquid chromatograph-mass spectrometry (LC-MS) (Loureiro, Quirós, & Sendón, 2018), liquid chromatography-tandem mass spectrometry (LC-MS/MS) (Regueiro & Wenzl, 2015; Vázquez Loureiro, Rodríguez-Bernaldo De Quirós, & Sendón, 2018; Xie, Zhao, Zhao, Chen, & Cai, 2018), gas chromatography- mass spectrometry (GC-MS) (Viñas, Campillo, Martínez-Castillo, & Hernández-Córdoba, 2010; Wang, Zhu, et al., 2017), ultra-performance liquid chromatography-high resolution mass spectrometry (UPLC-HRMS) (Shengdong et al., 2017), high performance liquid chromatography (HPLC) (Xiong et al., 2018) and ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-LC-MS/MS) (Grandin et al., 2017). Although the instrumental detection method is precise, the sample pre-treatment process is complex, and the equipment is expensive, and time-consuming. Furthermore, professional technicians are required, as the instrumental methods are difficult to apply to large sample sizes. Compared with instrumental methods, immunoassays can be used for simultaneous detection of multiple samples with good reproducibility, which meets the market demand.

At present, there are few reports looking at the preparation of BSP monoclonal antibodies and the establishment of an immunoassay detection system such as ELISA kits. In this study, we have developed a high specificity and sensitive monoclonal antibody for the detection of BSP. From this an immunological assay for BPS in milk samples was designed.

Materials and methods

Chemicals

BPS was purchased from J&K Scientific Ltd. (Beijing, China), N-hydroxysuccinimide,1-ethyl-3-(3-dimethylaminopropyl) carbodi-imide, N,N-dimethylformamide, 2-morpholino-ethanesulfonic acid, Keyhole limpet hemocyanin (KLH), ovalbumin (OVA), Freund’s complete adjuvant, Freund’s incomplete adjuvant, enzyme immunoassay-grade horseradish peroxidase (HRP)-labelled goat anti-mouse immunoglobulin, gelatin, 3,3’,5,5’-tetramethyl-benzidine, Tween-20 and polyethylene glycol 1500 were all purchased from Sigma-Aldrich (St. Louis, MO, USA). RPMI-1640 cell culture medium, 100 × HT (hypoxanthine-thymidine supplement), 50 × HAT (hypoxanthine-aminopterin-thymidine supplement), and fetal calf serum were acquired from Gibco BRL (Paisley, UK). All other chemicals and reagents were obtained from the National Pharmaceutical Group Chemical Reagent Co., Ltd. (Beijing, China).

Apparatus

The instruments used in this study consisted of a UV/Vis scanner purchased from Bokin instruments (Tsushima, Japan), a Multiskan MKS microplate reader obtained from Thermo Labsystems Company (Beijing, China), a vortex machine purchased from Shanghai Huxi Analysis Instrument Factory Co. Ltd (Shanghai, China) and a membrane dispenser obtained from Xinqidian Gene-Technology Co. Ltd (Beijing, China).

Buffers and solutions

We prepared the following solutions: boracic acid buffer (BB, pH8.8), coating buffer (CB; 0.05 M carbonate bicarbonate, pH9.6), blocking buffer (0.2% gelatin in CB), washing buffer (PBST; 0.05% Tween-20 (v/v) in 0.01 M PBS), and stop solution (2 M sulfuric acid). The substrate buffer consisted of a mixture of solutions A (0.3% H₂O₂, v/v in phosphate buffer) and B (0.01% TMB, w/v in glycol) at a 5:1 (v/v) ratio.

Hapten Synthesis

In this study, a new BPS hapten was synthesized using methyl 4-bromobutyrate. BPS contains two phenolic hydroxyl groups. This method involved the derivation of one hydroxyl group into a carboxyl group with linked arms, which was used to couple to the carrier protein (Figure 1). The specific steps are described as follows: 1 g of 4,4'-dihydroxydiphenyl sulfone was added into a round-bottomed flask, then approximately 6 mL of pure water was added and heated to 94°C to dissolve the 4,4'-dihydroxydiphenyl sulfone. Next, 0.24 g of solid sodium hydroxide was added. After one hour, of heating and reflux, 4-methyl bromide was added. The mixture was left to react for 30 h, and then hydrochloric acid was added to regulate the pH to approximately pH3-4. The reaction was extracted three times with ether, and the organic phase was merged with saturated salt water. The saturated salt water was stewing after full oscillation, then the organic phase was separated into the water phase. The organic phase was dried with anhydrous sodium sulfate, and after a period of standing, the sodium sulfate was filtered out and the solvent was removed by vacuum evaporation. The crude product was purified by column chromatography, and the purified compound A was obtained.

Figure 1. The synthetic route for the hapten of bisphenol S.

The 0.55 g of compound A produced was dissolved in 10 mL of methanol. During stirring, the prepared KOH solution (n_(KOH): n_{compound A} = 6:1) was added. After stirring for 48 h at room temperature, hydrochloric acid was added to adjust the pH value of the reaction system to approximately pH3.0, and a quenching reaction was carried out. The reaction system was extracted, dried, decompressed and distilled, and the purified compound B was obtained by column chromatography, which represented the BPS hapten, which was characterized using LC-MS.

Preparation and Identification of the complete antigen

The synthesized BPS was conjugated to bovine serum albumin (BSA) using the EDC coupling procedure, to form the complete antigen (BPS-KLH), as previously described (Kong, Liu, Song, Kuang, & Xu, 2017; Wang, Xie, et al., 2017). The BPS hapten (1.2 mg) was dissolved in 400 μL of dimethylformamide, then N-hydroxysuccinimide (2.7 mg) and 1-ethyl-3-(3-dimethylaminopropyl) carbodi-imide (2.2 mg) were added to the mixture. The mixture was stirred at room temperature for 8 h and then slowly added to 10 mg of KLH which was diluted with BB, and then stirred overnight at room temperature. The BPS-KLH conjugate was dialyzed against 0.01 M PBS for 3 days with six changes of dialysate and stored at −20°C until used. BPS was conjugated to OVA by the same method as used for the coating antigen, for the detection of serum titres and suppression of mice by ic-ELISA. The conjugates were identified by ultraviolet spectroscopy (Yamada, Satoh, Yamashita, Kambegawa, & Iwata, 1997).

Production of the BPS monoclonal antibody (mAb)

Female BALB/c mice (6–8 weeks of age) were immunized with complete antigen (BPS-KLH). Tail vein blood sampling was carried out after the third immunization and serum was analyzed using an ic-ELISA (Liu, Xu, Suryoprabowo, Song, & Kuang, 2018). The mouse with the highest inhibitory rate and titre was selected for cell fusion. The selected hybridoma cell lines were cultured and intraperitoneally injected into mice for ascites production (Kong, Liu, Song, Zheng, et al., 2017). After purification with octanoic acid-ammonium sulfate precipitation (Liang et al., 2014; Wang, Zheng, et al., 2017) and dialysis, monoclonal antibodies against BPS were obtained.

Development of the ic-ELISA method

The complete antigen (BPS-OVA) was diluted with CB (0.05 M, pH9.6) and added to 96-well plates (100 μL per well), incubated at 37°C for 2 h, washed three times, and blocked with blocking buffer (200 μL per well). After a 2 h incubation at 37°C, it was washed, dried, and stored at 4°C until used. Fifty micro litres of BPS standards and 50 μL of BPS mAb were added to each well and incubated at 37°C for 30 min. Then goat anti-mouse IgG-HRP was added to each well and incubated at 37°C for 30 min. After three washes, the substrate buffer was mixed with solutions A and B at a 5:1 (v/v) ratio and was added into each well and incubated at 37°C for 15 min. Following the addition of stop solution (2 M H₂SO₄, 50 μL per well), the absorbance was measured using a microplate reader at a wavelength of 450 nm. A standard curve of BPS mAb against BPS standard concentrations was then constructed.

Optimization of the ic-ELISA method

The methanol content, pH value and ionic strength (NaCl content) of the standard diluent were optimized under the same conditions. The OD_max and IC₅₀ values of the standard curve were used to evaluate the advantages and disadvantages of each optimization condition. Methanol was added to the PBS to make a final concentration of 0%, 5%, 10% and 20% (v/v), respectively. The effect of pH value was estimated using buffer solutions with different pH values ranging from 4.1 to 9.0. To optimize the ionic strength, PBS containing different concentrations of NaCl (0%, 0.8%, 1.6%, 3.2%) were tested.

Sensitivity of the BPS mAb

The half maximal inhibitory concentration (IC₅₀) means that the half maximum inhibitory concentration of the antagonist was measured. The optimum conditions for the action of the BPS mAb were determined by the above optimization process. Under these conditions, as shown in Figure 2, an indirect competitive inhibition curve for the BPS mAb was drawn by using Origin software and the IC₅₀ value was obtained.

Figure 2. Indirect competitive inhibition curve of bisphenol S monoclonal antibody under the optimized conditions.

Determination of cross-reactivity

The analogues and metabolites of the target drug were determined by cross-reaction. The IC₅₀ of the antibody was determined according to the procedure used for the ic-ELISA. The cross-reaction rate was calculated using the following formula: CR = (IC₅₀ of the original drug/IC₅₀ of the drug to be tested) × 100% (Xu, Pan, Yu, Wei, & Wang, 2017). Generally, a cross-reaction rate of less than 10% can be regarded as having no cross-reaction.

Detection of BPS in milk using a recovery experiment

Milk samples were obtained from a local supermarket in Wuxi, and a 1 mL milk sample was diluted eight times with known BPS concentrations (0.5, 1 and 2.5 ng/mL). The ic-ELISA method was then used to detect BPS in the samples and establish a corresponding standard curve to calculate the added recovery rate (Wang, Wang et al. 2013).

Labelling of the mAb with colloidal gold

The BPS mAb was labelled with gold nanoparticles (Guo, Xu, Song, Liu, & Kuang, 2018; Liu, Peng et al., 2017). One millilitre of colloidal gold solution, 0.1 M K₂CO₃ solution and 8 μg of mAb were added to a centrifuge tube in turn, and then mixed. The labelling reaction was carried out at room temperature for 2 h. After labelling, 10% (w/v) of BSA solution dissolved in 100 μL of ultrapure water was added slowly and stirred at room temperature for 2 h to block the uncoupled antibody binding sites on the colloidal gold particles. Thereby, prevent the colloidal gold particles from adsorbing each other and keep the stability of the system. Then the reaction was centrifuged at 9,500 rcf for 20 mins, and the supernatant was discarded to remove excess antibody and BSA. The precipitation was re-diluted in 1 mL of labelled detergent, washed, and centrifuged repeatedly. The final supernatant was discarded, and the precipitate was resuspended with 100 μL of heavy suspension and stored at 4°C until use.

Preparation of the immunochromatographic strip

The structure of the colloidal gold strip is shown in Figure 3. The colloidal gold strip consisted of a PVC backboard, sample pad, gold pad, water absorbent pad and cellulose acetate film (NC film). The NC membrane contains both a detection line (T line) and control line (C line). The coating antigen (0.25 mg/mL) was sprayed onto the T line and the goat anti-mouse antibody (1 mg/mL) was sprayed onto the C line. The sample pad is located at the end, near the T line, and the upper part of the back plate, which was treated with 0.01 M PBS containing 1% BSA, 1% sucrose, and 0.2% Tween 20, and dried at room temperature for 4 h. The gold pad with gold labelled monoclonal antibody was above the NC film and below the sample pad. A Water absorbent pad was at the end near the C line. The sample pad was immersed in the sample pad treatment solution, dried and stored for later use.

Figure 3. Principle of the immunochromatographic strip assay for bisphenol S.

Principle of the immunochromatographic strip assay

The immunochromatographic strip assay is based on an antibody–antigen reaction and competitive immunoreactions (Kong et al., 2019; Wang, Gu et al., 2018). As shown in Figure 3, for sample determination, the sample solution was first mixed with colloidal gold labelled mAb and reacted for 5 min. The mixture was then added to the sample pad, where it migrated towards the absorbent pad by capillary action. The results could be read by eye after 5 min.

For a positive sample, BPS will compete with the coating antigen adsorbed onto the T-line and combine with the gold labelled mAb. The higher the concentration of BPS in the positive sample, the weaker the T-line colour, until the colour of T-line disappears. The gold labelled mAb will bind with the goat anti-mouse antibody on the C line to make the C line become red whether BPS is present or absent within the sample. For a negative sample, the gold labelled mAb will definitely bind to the coating antigen at the T-line, making the T-line become red. If the C-line is colourless however, it means that the strip is invalid.

Sample analysis

Liquid milk selected for sample analysis, was purchased from a local supermarket in Wuxi. According to the linear detection range and IC₅₀ value of the antibody, a suitable known concentration of BPS was added to 1 mL of milk and diluted eight times and tested.

Results and discussion

Hapten

As a small molecule, BPS must be coupled with a macromolecular carrier protein to produce immunogenicity. The BPS molecule contains two active characteristic hydroxyl groups, which may be shielded by the process of its insertion into the protein vector, and thereby resulting in a loss of antibody specificity to the shielded groups. In order to fully expose the characteristic groups in small molecule compounds, one phenolic hydroxyl group is retained, and a carboxyl group is derived by introducing a straight chain spacer arm at the opposite end to the hydroxyl group. This maintains the original molecular structure of the hapten so that the artificial antigen is expose to the surface and can be recognized by immune active cells of the animal, thereby stimulating the body to produce antibodies with high affinity and specificity for the assay.

Complete antigen and coating antigen

As a small molecule with a relative molecular mass of 250.27, BPS is unable to induce an immune response. To induce the organism to produce antibodies, BPS must be coupled to a carrier protein. The carboxyl group of the BPS hapten was activated by the carbodiimide method and coupled to both KLH and OVA to synthesize the immunogen (BPS-KLH) and coating antigen (BPS-OVA). BPS-KLH and BPS-OVA were measured by absorbance spectra analysis within the wavelength range of 200–500 nm. The ultraviolet absorption curves are shown in Figure 4. As shown in the figure, KLH has a characteristic absorption peak at 278 nm, whereas the BPS hapten has a characteristic absorption peak at 262 nm. The peak for BPS-KLH is between 262 and 278 nm. The maximum characteristic absorption peak shift of the BPS hapten occurs due to its coupling to KLH.

Figure 4. The ultraviolet absorption spectrum of BPS-KLH (a) and BPS-OVA (b).

Optimization of assay conditions

The methanol content, pH value and ionic strength (NaCl content) in standard diluent were optimized. The best condition was obtained by comparing the A_max value and IC₅₀ value under various conditions. A_max reflects the binding properties of the coating antigen and antibody in the absence of standards, whereas IC₅₀ is a measure of antibody sensitivity. A highly sensitive mAb possesses lower IC₅₀ values.

Firstly, the effect of methanol content on the assay is shown in Figure 5(a). The IC₅₀ value of the standard diluent containing 20% methanol is approximately two times higher when compared with |diluent containing no methanol. The results showed that the presence of methanol could affect the sensitivity of the BPS monoclonal antibody. In addition, the A_max value decreased with an increase in methanol content. In view of the adverse effects of methanol on the sensitivity and A_max value of the BPS monoclonal antibody, a methanol-free PBS buffer was selected as the standard diluent.

Figure 5. Influence of methanol content (a), pH (b), and NaCl content (c) on IC₅₀ value and maximum absorbance (A_max) for the ic-ELISA of bisphenol S.

Secondly, the influence of pH value on the BPS monoclonal antibody is shown in Figure 5(b). It can be seen from the curves in Figure 3(b), that A_max values generally decreased even if the standard diluent was acidic or alkaline, while the IC₅₀ value almost changed. Therefore, the most suitable standard diluent pH value was 7.4. Thirdly, the effect of ionic strength on the detection method is shown in Figure 5(c). The appropriate ionic strength is necessary for the formation of antigen–antibody complexes. In the absence of ions the antibody curve was linear, and its titre was low. However, high ionic strength resulted in a decrease in A_max value, whereas the IC₅₀ value was not significantly affected. Therefore, a 0.8% NaCl content for the BPS standard diluent was selected.

As shown in Figure 2, under optimal conditions for the standard diluent (PBS buffer, pH 7.4), the indirect competitive inhibition curve is drawn with a logarithmic value of BPS standard concentration as abscissa, and absorbance value at 450 nm as ordinate. The curve had a linear range of 0.064–0.810 ng/mL, and the IC₅₀ value of the curve was 0.228 ng/mL.

Cross-Reactivity

The cross-reaction rate between drugs and their structural analogues can be used to describe the specificity of monoclonal antibodies and was used here also. Moreover, the lower the cross-reaction rate, the higher the specificity of the monoclonal antibody for its target. Data displayed in Table 1 showed that the rate of cross-reaction of the BPS mAb’s with other structural analogues was below 0.0228%, whereas with BPS, it was 100%. It is suggested therefore, that the BPS mAb is a monoclonal antibody with high specificity. The main reason for this is that the monoclonal antibody recognized the structure of diphenyl sulfone of BPS and a hydroxyl group on one side of the benzene ring. The structures of other analogues were quite different from that of BPS, which explained the reason for the antibody being unable to recognize these structures, and the cross-reaction rates, therefore, were low.

Table 1. Cross-reactivity results of bisphenol S monoclonal antibody.

Names	Molecular structure	IC50 (ng/mL)	Cross-reaction rate (%)
Bisphenol S		0.228	100
Bisphenol A		>1000	<0.0228
Tetrabromobisphenol A		>1000	<0.0228
Bisphenol C		>1000	<0.0228
Bisphenol F		>1000	<0.0228

Milk sample analysis

Matrix effect

In the analysis of the samples, matrix interference may affect the accuracy of detection. Therefore, the samples needed to be pre-treated before analysis (Wang, Xu, Zhang, & He, 2008; Zhao et al., 2017). Liquid milk contains a certain amount of fat, protein and lactose in addition to water, which may affect the recognition of antigens and antibodies, resulting in the precision and accuracy of the results being affected. In order to assess the effect of the matrix on the performance of the BPS mAb, indirect competitive inhibition curves of antibodies were established with PBS diluted eight times in milk. As shown in Figure 6, the curves were almost overlapping, indicating that the matrix effect was not significant. Therefore, milk samples can be diluted eight times with PBS and analyzed by the ic-ELISA procedure.

Figure 6. Comparison of indirect competition inhibition curve for bisphenol S obtained from standards prepared in PBS and 8-fold dilution liquid milk.

Recovery Experiments

The results of recovery rate and coefficient of variation are shown in Table 2. The recovery rate of BPS in milk ranged from 89% to 95%, which was within the allowable range. In addition, the coefficient of variation was lower than 15%. The results demonstrated that the ic-ELISA method using the BPS mAb is feasible for sample detection.

Table 2. Recovery rates and coefficient of variation of bisphenol S spiked in milk samples.

Sample	Spiked level (ng/mL)	Mean ± SD (ng/mL)	Recovery rate (%)	CV (%)
Milk	0.5	0.475 ± 0.011	95	2.3
	1.0	0.890 ± 0.053	89	6.0
	2.5	2.325 ± 0.024	93	1.0

Sample analysis using the immunochromatographic strip assay

Milk is the best source of calcium and protein for the human body. Therefore, the supervision and management of the quality and safety of dairy products are particularly important, to ensured that people drink high-quality milk. In addition, milk is suitable for people of all ages, and milk products available on the market are abundant and easy to purchase. To verify the feasibility of the immunochromatographic strip assay method, common liquid milk samples from a market were selected. As seen in Figure 7(a), the milk was spiked with BPS (0, 0.1, 0.25, 0.5, 1 and 2.5 ng/mL). As the concentration of BPS standard increased, the colour of the T line became gradually lighter. The standard concentration corresponding to the disappearance of the C line was the cut-off value, and the results were directly observed using the naked eye (Liu, Chu, & Yu, 2016). When the concentration of the BPS standard was 2.5 ng/mL, the T line disappeared completely. Therefore, the cut-off value for PBS was 2.5 ng/mL. Equally, as shown in Figure 7(b), the milk samples were spiked with BPS (0, 0.25, 0.5, 1, 2.5 and 5 ng/mL) and the cut-off value seen by the naked eye in milk sample was 5 ng/mL.

Figure 7. Image of bisphenol S detection by immunochromatographic strip in PBS (a) and milk sample (b). (a), 1–6 corresponds to concentration of bisphenol S spiked in PBS buffer solution were 0, 0.1, 0.25, 0.5, 1 and 2.5 ng/mL. (b), 1–6 corresponds to concentration of bisphenol S standard spiked in milk samples were 0, 0.25, 0.5, 1, 2.5 and 5 ng/mL.

Conclusion

The molecular structure of BPS was modified and conjugated with a carrier protein KLH. Monoclonal antibodies were obtained by immunization, antiserum detection and cell fusion. Under the optimized ic-ELISA analysis condition, the IC₅₀ value of the BPS mAb was 0228 ng/mL and the linear range was 0.064–0.810 ng/mL. The BPS mAb did not cross-react with other bisphenol structural analogues. The recovery rate of BPS in milk samples was in the range of 89%–95% and the coefficient of variation ranged from 1.0% to 6.0%. The cutoff value of the immunochromatographic strip assay method in milk samples was 5 ng/mL. Our findings confirm therefore, that the ic-ELISA method and immunochromatographic strip assay method established in this study are feasible for the detection of BPS in milk samples.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work is financially supported by National Key R&D Program [grant number 2018YFF0214803-04], the National Natural Science Foundation of China [grant numbers 21631005, 21874058, 51802125, 21771090, 21673104], and grants from Natural Science Foundation of Jiangsu Province, MOF and MOE [grant numbers BK20180029, BK20180605, CSE12N1708, BE2016307].