Although retinoids may exert their action via binding to nuclear retinoic acid receptors (RARs), other mechanisms of action are not excluded. For example, the anti-acne drug, isotretinoin, lacks affinity for the receptors, but is a very potent inhibitor of endogenous vitamin A metabolism in human epidermal cells. To further extend this observation, we studied the effect of 12 different retinoids on the metabolism of [3H]retinol ([3H]ROH) in HeLa cells, previously shown to produce constant levels of 3,4-didehydroretinol (ddROH). The cells were cultured in the presence of the unlabeled retiniods for 20 h, followed by 4 h incubation with [3H]ROH. The accumulation of [3H]ROH and [3H]ddROH in cellular extracts was analysed by HPLC. Addition of 10(-10) to 10(-5) M of four naturally occurring isomers of retinoic acid caused a 4- to 6-fold increase in [3H]ROH accumulation and an 80% decrease in [3H]ddROH. Addition of synthetic retinoids with a terminal carboxyl (CD270, CD271, CD367 and Ro 13-7410) decreased the [3H]ddROH accumulation with about 70%, but hardly at all affected the accumulation of [3H]ROH. We conclude that cultured HeLa cells appear to be useful for screening retinoids for their effects on vitamin A metabolism showing that a terminal carboxylic acid is a prerequisite for any major effects on metabolism to occur. Whether this effect is due to interaction with RARs or to competitive inhibition of vitamin-A-metabolizing enzymes demands to be studied.

For unknown reasons, the epidermal concentrations of retinol and its metabolite 3,4-didehydroretinol become characteristically changed during treatment with certain synthetic retinoids. In the present study we investigated [3H]retinol uptake and conversion to 3,4-[3H]didehydroretinol in cultured human skin exposed to six different retinoids. The tissue accumulation of [3H]retinol increased by 50-130% following addition of 1.3 microM of all-trans-retinoic acid, 13-cis-retinoic acid, or 3,4-didehydroretinoic acid (putative metabolites of retinol), and by 13-55% after addition of acitretin, arotenoid Ro 13-7410, or 4-(N-hydroxyphenyl)retinamide. The formation of 3,4-[3H]didehydroretinol was markedly inhibited (minus 80-90%) by the three putative retinol metabolites, moderately inhibited (minus 60-70%) by the aromatic retinoic acid analogues acitretin and arotinoid, and only slightly inhibited (minus 25%) by 4-(N-hydroxyphenyl)retinamide. Addition of citral and ketoconazole, substances known to interfere with the oxidative metabolism of vitamin A, diminished the formation of 3,4-didehydroretinol by 70 and 95%, respectively, but only marginally affected the retinol values. We conclude that the increase in retinol and marked reduction in 3,4-didehydroretinol concentrations in epidermis especially during treatment with 13-cis-retinoic acid are most likely due to diminished conversion of retinol to 3,4-didehydroretinol, but whether or not this reflects feedback inhibition of retinol metabolism or direct inhibition of a specific enzyme remains to be established. Nonetheless, in vitro analysis of 3,4-didehydroretinol biosynthesis appears to be a sensitive means for screening compounds for their ability to interfere with the cutaneous vitamin A metabolism.

Vitamin A (retinol) is an anti-keratinizing agent essential for normal epithelial differentiation. In order to examine the epidermal distribution of vitamin A and provitamin A (beta-carotene), we took advantage of the extraordinarily thick snout epidermis of the cow which can be cut horizontally into at least 6 layers, representing keratinocytes at different stages of maturation. Extracts of saponified samples were analyzed for retinol and beta-carotene by reversed phase high-performance liquid chromatography. The highest retinol concentration (0.8 micrograms/g protein; n = 3) was recorded closest to the dermis; progressively decreasing amounts of retinol were found in the upper parts of epidermis. Maximum values of beta-carotene (1.0 micrograms/g; n = 7) were found in the lower parts of epidermis; substantially lower levels were seen at the dermal transition zone and in the upper parts of epidermis. The results suggest that the endogenous concentration of vitamin A in snout epidermis is inversely related to the degree of cellular differentiation.