Abstract

A homologous RIA for the determination of toxic gizzerosine activity in commercial fish meals has been developed. Three polyclonal antibodies (GR316, GR415, and GR418) were produced in female rabbits and extensively characterized for their potential use in individual RIA. The RIA had lower detection limits of 0.048, 0.78, and 0.39 ng/mg using the three respective antibodies. Because gizzerosine is derived from lysine and histidine, crossreactivity with these amino acids, and with histamine was examined. The antibodies crossreacted with histamine from 21 to 100%.' No crossreactivity with histidine or lysine was observed for any of the three antibodies. Antibody GR415 was chosen for determination of gizzerosine in extracted fish meal samples because crossreactivity of histamine using this antibody was only present at high concentrations, and the Ka value for gizzerosine was 10-fold greater than for histamine. A mild buffer extraction procedure was used, resulting in 98% gizzerosine recovery. Displacement curves from extracted and serially diluted fish meal samples were parallel with gizzerosine standard. Inter-and intra-assay coefficients of variation were 11 and 15%, respectively. We used the RIA for determination of gizzerosine activity in a pool of 23 fish meal samples of known gizzerosine scores (determined with a chick bioassay), and histamine content. The partial correlation coefficient between gizzerosine content determined by the RIA and gizzerosine scores from the bioassay was high (0.83), and significant (P < 0.01). There were also significant correlations between gizzerosine scores and histamine content of the fish meals (0.63, P < 0.01), and between histamine content and gizzerosine levels determined by the RIA (0.59, P < 0.01). The application of the homologous RIA for the determination of gizzerosine activity in commercial fish meals could be of importance for the prevention of gizzard erosions in the poultry industry, and for studying gizzerosine-induced pathology and metabolism.