ISOLATION, CULTURE, AND CHARACTERIZATION OF THE 2 CELL SUBPOPULATIONS FORMING THE RAT DECIDUA - DIFFERENTIAL GENE-EXPRESSION FOR ACTIVIN, FOLLISTATIN, AND DECIDUAL PROLACTIN-RELATED PROTEIN

The decidual tissue of the pregnant rat is formed primarily by two markedly different decidual cell populations located in the mesometrial and antimesometrial sites of the uterus. The antimesometrial decidua functions as an endocrine organ, secreting PRL-like hormones and expressing activin and follistatin. In contrast, the mesometrial decidua has no apparent endocrine activity, but secretes abundant alpha(2)-macroglobulin. To determine whether the profound difference in morphology and gene expression is inherent to each cell population or whether it results from the position of these cells in the uterus and their subjection to regionally restricted information, we isolated the giant antimesometrial decidual cells from the small mesometrial decidual cells by elutriation. We examined, using Northern and reverse transcription-polymerase chain reaction analysis, the expression of PRL-related protein (DPRP), follistatin, activin, and alpha(2)-macroglobulin in the two cell populations just after cell separation and after primary culture. We also examined, using coculture, the possible cross-talk between the two cell populations. In culture, the cells maintained the morphological appearance seen in the decidua; the mesometrial cells remained less differentiated, whereas the antimesometrial cells formed giant cells. In addition, the antimesometrial cells, which have been shown previously to be the only site of DPRP messenger RNA (mRNA) expression, remained, after several days of culture, the only cells able to express this gene. The coculture experiment suggested that the absence of DPRP mRNA in the mesometrial cells is not due to inhibitory signals from the antimesometrial cells, but, rather, to the inherent inability of the mesometrial decidual cells to express the DPRP gene. The cellular expression pattern of follistatin was also not changed by cell separation. However, in sharp contrast to that of DPRP and follistatin, the cell-specific expression of activin-beta A and alpha(2)-macroglobulin was markedly affected by cell separation. A remarkable switch in activin-beta A cell expression was observed when mesometrial cells were separated from antimesometrial cells and maintained in culture. Activin-beta A mRNA, which was barely detectable in these cells in situ, became expressed in higher levels than in antimesometrial cells. The isolated antimesometrial cells also acquired the ability to express the alpha(2)-macroglobulin gene, whose expression is highly restricted to the mesometrial cells in vivo. No down-regulation of antimesometrial alpha(2)-macroglobulin mRNA could be induced by mesometrial cells in a coculture experiment. However, alpha(2)-macroglobulin mRNA levels in mesometrial cells were clearly up-regulated by antimesometrial cell signal(s). In summary, we successfully separated the two decidual cell populations and developed primary cultures of both cell types, either alone or in coculture systems. Our findings that both decidual cells are able to express follistatin, alpha(2)-macroglobulin, and activin-beta A strongly support the notion that both cells arise from a single stem cell population. The restricted in situ expression of each gene in a defined cell population suggests the presence of a local tonic inhibitory mechanism. The cell-specific expression of DPRP makes this hormone an excellent marker for antimesometrial cells and suggests that the DPRP expression is programmed at a very early stage during stromal cellular differentiation.