The Immunocytochemistry of Cytokeratin in Fish Tissues

An increasing interest in fish species as sentinels of environmental pollution and in carcinogenesis research has led to the identification of diagnostically challenging neoplasms of uncertain cellular origin and the need for additional diagnostic methods. To determine the potential of using commercially available antibodies to intermediate filament proteins on paraffin-embedded fish tissues for immunocytochemistry in tumor diagnosis, the application of three antikeratin antibodies to normal adult tissues from two fish species was assessed. Multiple tissues from 12–14-in, striped bass (Morone saxtills) and 6-month-old medaka (Oryzias latipes) of both sexes were fixed in Bouin's or formalin fixatives. Formalin-fixed neoplasms from several mammalian species, including cat, dog, hedgehog (Atelerix albiventris, Erinaceus europaeus), rhesus macaque (Macaca mulatta), and sloth bear (Melursus ursinus), were also used as positive controls. Using a strepavidin horseradish peroxidase method on paraffin-embedded tissues, the broad spectrum antibodies AE1/AE3 (Boehringer Mannheim, Indianapolis, IN) and MAK-6 (Triton Biosciences, Alameda, CA), which recognize most of the 19 human cytokeratins, and CAM 5.2 (Becton Dickinson, Mountain View, CA), which recognizes cytoker-atins present in human liver, were used as primary antibodies. Epithelia from skin, gills, cornea, bile ducts, renal tubules, gastrointestinal tract, and thymus were strongly positive with AE1/AE3 and MAK-6 in striped bass, but nonepithelial tissues such as bone and muscle were negative. Skin, gills, cornea, and portions of the gastrointestinal tract were strongly positive in medaka with the same antibodies, whereas bile duct, renal, and intestinal epithelia were less so. Tissue digestion improved the intensity of staining, and fixation with Bouin's fixative improved results somewhat compared with formalin fixation. Although CAM 5.2 elicited positive reactions in mammalian hepatocytes and biliary, intestinal, and renal epithelia, all tissues from either fish were negative despite manipulations of conditions such as strength of primary antibody concentration and duration of tissue digestion. Although variations in results among mammals and fish and between the two fish species may indicate differences in tissue cytokeratin immunogenicity, biochemistry, or content, the broad crossreactivity seen suggests good potential for the use of diagnostic immunocytochemistry in fish carcinogenesis research.