Assessment of retinoid-induced differentiation of F9 embryonal carcinoma cells with an enzyme-linked immunoadsorbent assay for laminin: statistical comparison of dose-response curves.

A convenient procedure, using enzyme-linked immunoadsorbent assay of laminin, to measure retinoid-induced F9-cell differentiation into parietal endoderm was developed. Dose-response curves were fitted with the Allfit program, a statistical method for the analysis and simultaneous comparison of sigmoidal curves, which has been modified for use with a microcomputer. The procedure was standardized with respect to time of retinoid incubation, time-course of laminin production, effects of dibutyryl cAMP, and nature of individual dose-response curves. Retinoic acid produced a half-maximal response at 1.3 nM. Retinol was 175-fold less potent than retinoic acid and required 72 h to effect a maximum response, in contrast to 48 h for retinoic acid. Six oxidized and/or isomerized metabolites of retinoic acid, including 13-cis-retinoic acid, were less potent than retinoic acid, but were more potent than retinol. The dose-response curves had identical slopes with the exception of those obtained with 13-cis-4-oxo- and 4-oxo-16-hydroxyretinoic acids, the only metabolites tested with two structural alterations relative to retinoic acid. Multiple functional group alterations were synergistic in deactivating retinoic acid. The synthetic retinoids 13-cis-N-ethylretinamide and 4-hydroxyphenylretinamide and the steroid hormone 1,25-dihydroxycholecalciferol were inactive.