PubMed1989
Hyper- and hypovitaminosis A both provoke epithelial pathologies in animals and humans. This suggests that a critical level of retinoic acid (RA) is required in vivo for the maintenance of normal architecture and function of these tissues. However, no beneficial, but only adverse effects of RA on epithelia have been so far observed in vitro. For instance, addition of RA to keratinocyte cultures has been shown to inhibit epidermal differentiation while this process is stimulated by serum delipidization, which reduces RA concentration in the medium. Assuming that the previous failure to demonstrate beneficial effects of RA on the epidermal phenotype in vitro was due to culture conditions too far from the in vivo conditions we decided to reevaluate the effect of RA in a culture system optimized for epidermal morphogenesis: the "emerged dermal equivalent." When human keratinocytes were grown in such a system with total fetal calf serum, the resulting epithelium was very similar to normal epidermis. But when delipidized serum was used, the epithelium was abnormal in the direction of excessive maturation (hyperkeratosis). When physiological concentrations of RA (10(-9) and 10(-8) M) were added to the delipidized serum supplement, a normal architecture (orthokeratosis) was restored. However, as classically described in the literature, higher RA concentrations (greater than 10(-7) M) reduced epidermal maturation and produced parakeratosis. Thus, although it is unquestionable that RA reduces the synthesis of epidermal-specific differentiation markers, an optimal epidermal morphogenesis seems to be achieved only in the presence of a critical RA concentration.