Development of an indirect enzyme-linked immunosorbent assay and lateral-flow test strips for pefloxacin and its analogues in chicken muscle samples

ABSTRACT An anti-pefloxacin (PEF) monoclonal antibody (mAb), an indirect competitive enzyme-linked immunosorbent assay and lateral-flow test strip methods were developed to detect fluoroquinolone (FQ) residues in chicken muscle samples. Under optimised conditions, the anti-PEF mAb showed reasonable cross-reactivity with nine FQs with a limit of detection of 0.082 ng/mL assayed in 0.01 M phosphate buffered saline (PBS) solution. The intra- and inter-assay recoveries from spiked samples were within the range of 62.42–111.47% and 63.5–113.79%, respectively. The visual cut-off values of the lateral flow test strip in 0.01 M PBS and in food matrices were within the range of 2.5–50 ng/mL and 5–100 µg/kg, respectively. These results show that the anti-PEF mAb immunoassay and lateral flow test strip methods are suitable for simultaneous detection and routine monitoring of FQ residues in food.


Introduction
Antibiotics are subdivided into different groups based on their activity and structures. Among these groups, synthetic fluoroquinolones (FQs) are the most important class of antibiotics that are widely used to treat serious infectious diseases in clinical practice and in veterinary medicine (Chang, Wang, & Tsai, 2010). Pefloxacin, [1-ethyl-6-fluoro-7-(4-methyl-piperazin-1-yl)-4-oxo-quinoline-3-carboxylic acid], which contains a naphthyridine core, is a broad-spectrum FQ antibacterial agent. It shows excellent antibacterial activity against most gram-negative and gram-positive bacteria (Soayed, Refaat, & Noor El-Din, 2014). Clinically, it is widely used in the treatment of respiratory and urinary infections. In agriculture, pefloxacin, along with other antibiotics, is used to antibody was purchased from Jackson ImmunoResearch Laboratories (West Grove, PA, USA). Hypoxanthine-aminopterin-thymidine (HAT) and hypoxanthine-thymidine (HT) supplement and other cell culture reagents were purchased from Life Technologies Corporation (Shanghai, China). Other reagents and chemicals were acquired from the National Pharmaceutical Group Chemical Reagent Co., Ltd. (Shanghai, China).
Eight-week-old female BALB/c mice were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd., (Beijing, China).

Design and synthesis of antigens
The immunogen and coating antigens were prepared by the carbodiimide (EDC) coupling method reported in our previous study (Mukunzi, Isanga, Suryoprabowo, Liu, & Kuang, 2017). Initially, the carrier protein BSA was cationized with 2,2 ′ -ethylenedioxy-bis-ethylamine (EDEA) to form EDEA-BSA. As illustrated in Figure 1, PEF was covalently coupled to EDEA-BSA to form the immunogen PEF-EDEA-BSA. Briefly, PEF (15 mg), EDC (24 mg), and NHS (9.0 mg) were dissolved in 5 mL of 0.05 M MES, pH 4.7, then incubated for 6 h at room temperature under constant stirring. Next, this solution was slowly added to a solution of EDEA-BSA (0.35 μmol) in 10 mL of CB (0.05 M, pH 9.6). The mixture was stirred for 3 h, and uncoupled haptens were removed by dialysis at room temperature against frequent changes of PBS solution every five to six hours for three days. The coating antigen was prepared by the same method using EDEA-OVA as the carrier protein.
UV absorbance was used to evaluate conjugation rate. Conjugates were stored at −20°C.

Immunisation of mice
Eight-week-old female BALB/c mice were acclimatised for one week before immunisation. An emulsion of immunogen PEF-EDEA-BSA was prepared by using 1 mL of FCA and 2 mg of PEF-EDEA-BSA dissolved in 1 mL of sterile 0.9% (w/v) sodium chloride (NaCl). The primary dose consisted of 100 µg and was subcutaneously injected at multiple sites on the back of mice. Three subsequent booster injections of 50 µg immunogen emulsified with FIA were administered at three-week intervals. Seven days following the third boost, blood was collected from the tail vein of the mice. The collected blood samples were tested for anti-PEF activity by indirect competitive ELISA. Lastly, the mouse with the highest anti-PEF activity was intraperitoneally injected with 30 µg of immunogen prepared in 0.2 mL saline and three days later; its spleen was removed and used for hybridoma production.

Cell fusion, screening and production of mAb
The mouse antiserum specific for the PEF molecule was selected for cell fusion. Mouse spleen lymphocytes were fused with myeloma cells using polyethylene glycol (PEG) to form hybrid cells Kong et al., 2015;Kuang et al., 2013). One week after cell fusion, the ic-ELISA method was used to screen for positive cells with high anti-PEF activity. The selected cell lines were subcloned and screened repeatedly five times at intervals of six to seven days. At the end of the cell fusion process, the hybrid cell line 2G3 was selected for bulk antibody production. The cells were intraperitoneally injected into mature female BALB/c mice primed with paraffin oil. Ascites fluid which contained the antibodies was harvested and purified by the caprylic acid-ammonium sulphate precipitation method (Kuang et al., 2013). Further purification was performed by dialysis against 0.01 M PBS in order to remove ammonium sulphate salts and other small impurities. The purified antibody was quantified and stored at −20°C.

ELISA procedure
The ic-ELISA was developed using the checker-board titration procedure to enhance the interaction between coating antigen PEF-EDEA-OVA (0.3, 0.1, 0.03 and 0.01 μg/mL) and anti-PEF mAb (1.0, 0.5, 0.25 and 0.125 μg/mL) concentrations. Microtiter plates were coated with 100 μL/well and incubated at 37°C for 2 h. The plates were washed three times with PBST (0.01 M PBS containing 0.05% Tween-20 v/v), and blocked with 200 μL/well of the blocking buffer. After incubation at 37°C for 2 h, the plates were washed twice with PBST and each well was loaded with 50 μL of PEF standard solution followed by the addition of 50 μL of anti-PEF-mAb. The plates were incubated for 30 min at 37°C. Afterwards, the plates were washed three times, and secondary antibody (HRP-labeled goat anti-mouse IgG) diluted 1:3,000 with antibody dilution buffer was added (100 μL/well). The microtiter plates were incubated for 30 min and washed four times. Immediately, freshly-prepared TMB substrate solution was added to the plates (100 μL/well) and allowed to react in the dark for 15 min at 37°C. The colour development was stopped by adding 50 µL/well of 2 M Sulphuric acid (H 2 SO 4 ) and the absorbance values were determined at 450 nm by a microtiter plate reader.

Effects of physical-chemical parameters on ELISA performance
Immunoassay methods are generally improved by adjusting various parameters of the assay buffer. In this study, the effects of pH, salt concentration and organic solvent contents were investigated. A pH range of 6.0-9.0, sodium chloride in the range of 0-3.2% (w/v), and methanol and acetonitrile concentrations in the range of 0-40% (v/v) were tested. The maximum absorbance (A max ), half-inhibition concentration (IC 50 ), and maximal A max /IC 50 ratio were used as conclusive criteria. Then, based on the optimum conditions identified, a standard curve was established using the four-parameter sigmoidal logistic dose response generated by OriginPro 8.5 (OriginLab Corporation, Northampton, MA, USA).

Cross-reactivity determination
The sensitivity and specificity of the anti-PEF mAb were evaluated by studying cross-reactivity (CR) with related FQ analogues (Table 1). Standard concentrations were prepared under the optimum assay buffer conditions and each IC 50 value was determined using the ic-ELISA procedure. The CR was calculated using the following equation.

CR(%) =
(IC 50 PEF) (IC 50 of FQ analogue) × 100. Preparation of the lateral-flow test strip Preparation of colloidal gold particles Colloidal gold nanoparticles (GNPs) were prepared using the trisodium citrate reduction method (Kong et al., 2015;Xu et al., 2016). All solvents were prepared with double-distilled water and then filtered through a 22 µm transfer membrane. With constant stirring, the chloroauric acid (25 mL of 0.1 g/L) solution was heated to boiling point, then 1% w/v sodium citrate tribasic dihydrate solution (1.0 mL) was added. The resulting mixture was kept stirring for 30 min until it developed a colour like red wine. The solution was then cooled to room temperature, and stored at 4°C. Transmission electron microscopy analysis showed that the GNP had a closely uniform particle size of 15 nm.

Preparation of colloidal gold-labeled mAb
The PEF colloidal gold-labeled mAb was prepared following the methods previously described . The colloidal gold solution was neutralised to pH 7.0 with 0.1 M K 2 CO 3 . To this solution (10 mL), anti-PEF mAb (0.4 mL) was slowly added and after 35 min, 1 mL of 10% (w/v) BSA was added. The mixture was stirred for 2 h. The resulting product was centrifuged for 45 min at 8000×g to remove gold aggregates. After centrifugation, the bottom layer (red gold-labeled) was collected and washed with 0.02 M phosphate buffer containing 5% sucrose, 1% BSA, and 0.5% PEG 6000 (pH 7.4). The conjugates were reconstituted to 1 mL in gold-labeling resuspension buffer (0.02 M PBS, 5% sucrose, 2% sorbitol, 1% mannitol, 0.1% PEG, 0.1% tween, and 0.04% NaN 3 ) and stored at 4°C.

Preparation of NC capture membranes
The coating antigen (PEF-EDEA-BSA) and goat anti-mouse IgG were used as the capture reagents in the control line on the test strip. The antigen and goat anti-mouse IgG coatings were sprayed onto the NC membrane at 1 µL/cm using a dispenser to form the test and control lines on the strip. The capture and control reagents were also sprayed onto the glass fibre membrane to prepare the conjugate pad, which was later dried at 37°C for 2 h. The NC membrane coated with capture reagents was then pasted onto the centre of the plastic backing plate PVC and the conjugate pad, sample pad, and absorbent pad were laminated and pasted onto the backing plate. Finally, the plate was cut into 2.8 mm-wide strips with a strip cutter.
Application of the lateral flow test strip in 0.01 M PBS and spiked chicken meat samples As illustrated in Figure 2, the results of the test strips were evaluated as follows: a positive test was obtained when a sample with a strong target analyte was loaded onto the sample application pad and flowed through the strip to produce a coloured line in the control zone (Figure 2(B)); a negative result occurred when a sample with a weak or without a target analyte was loaded onto the sample application pad and flowed through the strip to produce coloured lines in both control zone and test zone (Figure 2(A)). When no line appeared either at the control line or the test line, the test was invalid Mukunzi et al., 2016). In this work, 50 µL of PEF-gold-labeled mAb was mixed with 150 µL of sample solution (FQ analyte dissolved in PBS or spiked into chicken muscle samples) and allowed to react for about 5 min, then added to the sample pad. The solution was allowed to migrate to the absorbent pad, and the test results were obtained within 10 min.

Sample preparation and analysis
Chicken muscle samples were purchased from a local supermarket and authenticated to be free from FQ residues by HPLC-MS/MS. The samples were homogenised, and aliquots of 2 g were placed into 15 mL test tubes containing 2 mL of 80% methanol solution and spiked with different concentrations of each FQ drug (PEF, NOR, CIPRO, ENO, FLE, PF, OFL, SARA, and LOM; Table 2). The spiked samples were thoroughly vortexed, mixed on a shaker for 15 min, and then centrifuged at 8445×g. The supernatants were collected and dried in a vacuum drier at 45°C; the resulting dried matter was reconstituted with 2 mL of 0.01 M PBS and aliquots were analysed by either ic-ELISA or the lateral flow strip test method.

Results and discussion
Design and synthesis of the coating antigen Pefloxacin, like other small molecules, is not able to stimulate an immune response except when it is attached to a large protein carrier such as BSA, keyhole limpet hemocynin (KLH) , or OVA. Therefore, it is fundamentally necessary to modify its structure in such a way that it will produce a desirable immune response. Based on the molecular This new antigen (PEF-EDEA-BSA) was spectrophotometrically analysed and displayed the spectrum range from 300 to 350 nm highlighting the presence of PEF and it also showed a large peak at 280 nm emphasising the combined spectrum of PEF and EDEA-BSA molecules ( Figure 3). This indicated that the conjugation was successful. The coating antigen PEF-EDEA-OVA was analysed using the same approach, which revealed a characteristic UV pattern similar to the PEF-EDEA-BSA antigen.

Optimisation of ic-ELISA
The optimum conditions for an ic-ELISA are usually established by evaluating different parameters in order to enhance the sensitivity of the antibody. Using the checkerboard titration method, it was determined that the optimal concentrations of coating antigen (PEF-EDEA-OVA) and anti-PEF mAb were 0.03 and 0.125 μg/mL, respectively. Other parameters of pH, NaCl and organic solvent contents were assessed in order to simulate the characteristic behaviour of food matrices. The evaluation of pH effects showed that from pH 5.0 to 8.0 there was a progressive increase in A max value (1.519-1.802) and a slight decrease at pH 9.0 (A max value 1.653). On the other hand, the IC 50 value was decreased as the A max /IC 50 ratio increased (Figure 4(a)). Ultimately, pH 8.0 was selected for subsequent experiments. Increasing salt concentration (0-0.1.6%) showed no significant change in A max value until the salt concentration reached 3.2% when the A max value dropped sharply down from 1.858 to 1.378. A salt concentration of 1.6% was selected because it exhibited a desirable IC 50 value (Figure 4(b)). Organic solvent tolerance was optimised also to assess the effects of methanol and acetonitrile, since they are often used in the extraction of FQ residues from food samples (Chen & Jiang, 2013). It was observed that increasing the methanol concentration from 0% to 5% had no significant effects on A max , while the IC 50 value changed slightly from 0.213 to 0.216 ng/mL. Concentrations of methanol of 10% and above significantly increased the IC 50 value. The A max was negatively affected by the increase in acetonitrile concentration, but the IC 50 was not greatly affected until the concentration of acetonitrile reached 40% (Figure 4(d)).

Determination of cross-reactivity
The capability of the anti-PEF mAb to recognise other FQs was determined by ic-ELISA and the results are shown in Table 2. In comparison with earlier immunoassays developed against FQs (Cao, Lu, et al., 2011;Cao, Sui, et al., 2011;Jinqing, Haitang, & Figure 3. Characterization of PEF-EDEA-BSA conjugate by UV spectrophotometric. Ziliang, 2011;Tochi et al., 2016;Wang, Zhang, Ni, Zhang, & Shen, 2014), the anti-PEF mAb and ic-ELISA developed in this study showed a great improvement in terms of IC 50 (0.218 ng/mL) and could effectively bind 9 out of 22 FQs investigated with an IC 50 below 2 ng/mL. This broad selectivity was only observed for PEF, NOR, CIPRO, ENO, FLE, PF, OFL, SARA, and LOM; the majority of drugs which are classified into the second generation of FQs (King, Malone, & Lilley, 2000).

Immunoassay analysis of spiked chicken muscle samples
Based on the CR and IC 50 , the compounds PEF, NOR, CIPRO, ENO, FLE, PF, OFL, SARA and LOM were spiked into and recovered from spiked chicken muscle samples and analysed by ic-ELISA. To evaluate whether the matrix was affecting the ic-ELISA, standard curves were generated in PBS and compared with standard curves generated using the control matrices from chicken muscle sample extracts. Comparison of their IC 50 and A max showed no significant difference. Table 2 shows the spiked levels and recovery data for each drug. The mean recoveries for the nine selected FQs ranged from 62.42% to 111.47% and 63.5% to 113.79% with %CV less than 13.59% and 12.2%, respectively for intra and inter-assays. In summary, the spiked samples showed good agreement between the spiking level and the concentration detected. Therefore, the developed ic- ELISA method is appropriate for simultaneous detection of FQs in chicken muscle samples.

Conclusion
In this study, we developed a PEF-monoclonal antibody, ic-ELISA and lateral flow test strip for the detection of FQ residues in chicken muscle samples. Under optimised conditions, the mAb showed an IC 50 value of 0.2 ng/mL, with an limit of detection value of 0.082 ng/mL. Based on the sensitivity of the anti-PEF antibody, ic-ELISA and lateral flow test strip methods were capable of detecting nine target FQ substances in chicken muscle samples. The results were within the range of 0.2-4 ng/mL and 5-100 µg/kg, respectively for ic-ELISA and lateral flow test strip methods. Therefore, these rapid and effective tools are suitable for sensitive and on-site mass sample screening and can play an important role in preliminary and semi-quantitative analytical methods.

Disclosure statement
No potential conflict of interest was reported by the authors.