Multiplex biosensor immunoassays for antibiotics in the food chain

The use of antibiotics in food-producing animals may result in unwanted residues in food products. The main objective of the present research was to study the development and application of fast and automated multiplex surface plasmon resonance (SPR)-based biosensor immunoassays (BIAs), based on multi-component antibodies and/or combined immunoassays in serially connected flow channels, for the detection of selected antibiotics in the food chain. The scientific challenges to deal with were: the development of multi-sulfonamide monoclonal antibodies (Mabs) against the generic structure of sulfonamides and the evaluation of mutated recombinant antibodies (Rabs) derived thereof, finding of the best BIA format with aminoglycosides as model compounds and solving foreseen matrix and combined immunoassay interferences, and to study the use of antibiotic concentrations in blood serum as predictors for concentrations in edible tissue.
Broiler’s blood serum, easy to collect in slaughterhouses, was chosen for the detection of sulfonamides and quinolones which are frequently used in poultry. With a Mab raised against sulfamethazine (21C7), the BIA could detect at least eight sulfonamides in ten times diluted broiler serum with limits of detection (LODs) far below the desired detection limit. Other less performing Mabs were developed against the generic part of sulfonamides. The best Mab-producing hybridoma cell-line (27G3) was used by the University of Turku to develop better performing mutated Rabs and the mutant-based BIA in broiler serum was found to be the most sensitive towards most of the sulfonamides. The assay was fast (5 min per sample), robust (>1000 runs per chip) and the sample preparation was easy (dilution in buffer only). The Rab-based multi-sulfonamide immunoassay was applied to analyze serum samples from broilers treated with sulfamethoxazole and sulfadiazine and the concentrations found were higher than the concentrations found in tissue by LC-MS/MS. This, and the good correlation with tissue concentrations, made this assay suitable to predict levels in edible tissue. A similar result was obtained with the specific BIA for flumequine.
Unique direct BIAs for the detection of aminoglycosides in milk were developed with Mab-coated chips. However, the inhibition assays with aminoglycosides on the chips were found to be more robust. For the simultaneous detection of five aminoglycosides in milk, the sensor chip surfaces in the four serially connected flow channels were covered with four aminoglycosides. In combination with a mixture of four specific antibodies, gentamicin, neomycin, kanamycin and (dihydro) streptomycin could be detected in milk far below the maximum residue limits (MRLs) and within 7 min.
In conclusion, serum and milk are suitable sample materials for the biosensor detection of antibiotics in the food chain. Such assays are fast, robust, automated, easy to handle, and require simple sample preparations (dilutions in antibody-containing buffer). In principle, such assays can be combined with assays for the detection of anti-pathogens, which broadens the application area in a food safety control system. However, the four-channel biosensor systems are too limited and the antibodies too specific for the simultaneous detection of more antibiotics. More extended multiplex systems (e.g. imaging SPR sensors or multiplex flow cytometry-based systems) need to be explored in which the knowledge obtained in the present research will likely be of great value.