The gap junction communication in Sertoli cells from immature rat testes, cultured either in absence or in presence of follicle-stimulating hormone (FSH), was studied by microinjection of a fluorescent dye and by Fluorescence Recovery After Photobleaching (gapFRAP). The cells cultured for 2-4 days in the absence of FSH showed a flattened "epithelial-like" appearance. They were poorly coupled, as judged by the low frequency of cell-to-cell spread of microinjected Lucifer Yellow, and by the value of the rate constant of dye transfer (k) estimated in gapFRAP experiments. However, when two different subpopulations of cells were separately analyzed, namely the cells forming small groups contacting over part of their circumference ("adjoining cells"), and the cells arranged in tight clusters, we found that the value of k in the latter group was much higher, reaching about 75% of that obtained in the presence of FSH. The cells cultured for two days in a medium containing ovine FSH underwent striking morphological changes and presented a rounded, "fibroblast-like" appearance. They were arranged in networks or in clusters. The frequency of cell-to-cell dye diffusion after microinjection and the rate constant of dye transfer were rapidly increased to the same final level by FSH, although they were initially different in these two groups. A concentration dependence of k, in the range 0.05 to 3 ng/ml, was observed in the cells in networks, contrasting with an all-or-none increase in the cells in clusters. Two days after FSH withdrawal, the dye transfer constant returned to prestimulation control values in the cells in clusters, but not in the cells in networks, which maintained a stable degree of coupling comparable to that of the unstimulated cells in clusters. This observation suggests (i) that an initial promoting effect of FSH already exists in the immature rat testis, which is preserved after enzymatic treatment in the cell clusters, but not in the more dispersed cells, and (ii) that the decreased junctional coupling is re-established in the dispersed cells by FSH, through a synthesis or a membrane insertion of connexin. The effects of FSH were mimicked by a brief exposure to 1 mM dibutyryl-cyclicAMP, but not to 10 nM human chorionic gonadotropin (hCG), indicating that the gap junction communication in Sertoli cells is upregulated by FSH through a specific membrane receptor, with cyclicAMP acting as a second messenger.