SUMMARY
   The presence of antibiotic residues in food from animal origin is one of the most important and contemporary problems in food quality and safety control. Medicines are employed in animal production for therapeutic use and, as growing promoters. The problem arises when antibiotic residues reaches to the consumer through food chain, at health harmful levels, since they have secondary effect, they destroying synergic bacteria in our intestine and provoking allergic reactions. These effects contribute to the raising of resistant bacteria.
   Regarding regulations, in CE regulation No. 470/2009, community procedures for fixing the maximal residual limits (LMR) of active drug substances in commodities from animal origin are established. Thus, for most of the employed veterinary antibiotics, and depending on the tissue, LMR ranges from 5 to 600 µg/kg.
   Currently, tetracycline and sulfonamide determination is carried out mainly by means of liquid chromatography (HPLC) with fluorescent, UV or mass spectrometry (MS) detection. The method is complicated and expensive. Furthermore, the step of sample preparation and clean-up required for obtaining a high purity degree is laborious, and that together with the low rate and high cost of analyses, imply that this methodology is little adequate for carrying out survey and control programs that guarantee food safety. Screening methods are commonly used for detecting residues of antibiotics in food prior to the confirmatory analysis. These methods must be semiquantitative, rapid, robust and relatively cheap. Among screening methods employed for detecting antibiotic residues, chromatographic, microbiologic and immunologic ones are current.
   This Thesis has been focused on the development of immunoreagents and screening analytical systems based on immunochemical methods (ELISA), addressed to detect antibiotic residues, concretely sulfonamides and tetracyclines, in food. The set-up techniques will allow to know the level of these substances in a rapid, sensitive, cost-effective and reliable way, with a minimal sample treatment and without the need of qualified personnel.
   The work was focused first in the synthesis of haptens for raising antibodies against the different compounds from the sulfonamide and tetracycline families. Thus, different molecules were synthesized maintaining the structural basis for each antibiotic group (constant region), and they were functionalized with carboxylic acid groups in order to obtain immunogens, coating conjugates and enzyme tracers, to be used for assays setting-up. In all cases, polyclonal antibodies that have demonstrated to be effective and easy to obtain regarding to both operative technique and response time, were raised by immunizing rabbits.
   The obtained sera were characterized (sensitivity and selectivity) by means of ELISA, establishing the optimal immunoreagent concentration and assay formats. After setting up different assays, the most specific ones were selected, two for sulfonamides (sulfatiazole, STZ and sulfasalazine, SSZ) and one for tetracyclines (chlortetracycline, CTC).
   On the other hand, generic (low selectivity) immunoreagents were raised for some of the most employed sulfonamides and tetracyclines, allowing for the development of two multiresidue immunoassays.
   Furthermore, the physico-chemical factors that have influence on antigen-antibody interaction, such as medium ionic strength, incubation time, pH and surfactant concentration, were optimized. The most sensitive specific immunoassays showed IC50 values of 1,6 µg/L for STZ, 0,51 µg/L for SSZ and 36,40 µg/L for CTC, while in generic assays IC50 was 1,32 µg/L expressed as STZ equivalent, and 6,5 µg/L expressed as tetracycline equivalent. All the developed ELISAs for sulfonamides and tetracyclines showed good sensitivity, at µg/L level, this being lower than LMRs established in regulations.
Finally, the analytical potential of the developed immunoassays was evaluated by applying them to the determination of sulfonamide and tetracycline residues in honey. Results showed excellent recoveries (mean recovery 97%), what makes this methodology interesting for implementing Survey Plans, with minimal sample treatment.