The decidual prolactin receptor and its regulation by decidua-derived factors.

Decidualization of the endometrial stroma in the rat gives rise to two different cell populations, located either mesometrially or antimesometrially in the uterus. We have previously shown that the rat decidua is the site of production and action of a PRL-like hormone. In this investigation we examined, using reverse transcription-PCR, whether and which type of PRL receptor (PRL-R) messenger RNA (mRNA) is expressed in the decidua, whether the receptor is confined only to one cell population, and whether the PRL-R expression is regulated by decidua-derived factors. The results indicate that the uterus of pseudopregnant rats does not express the PRL-R and that decidualization does not trigger a rapid appearance of PRL-R mRNA. It is only 3 days after the induction of decidualization that the long form of the PRL-R was first expressed. Thereafter, mRNAs for both the short (PRL-RS) and the long (PRL-RL) form became detectable in both antimesometrial and mesometrial decidua, although PRL-RL mRNA was much more abundant than PRL-RS. As development proceeded, PRL-R mRNA decreased and disappeared specifically from the antimesometrial decidua, whereas the mesometrial decidua continued to express this receptor mRNA. Concomitant with down-regulation of the PRL-R in the antimesometrial tissue was a rather abrupt expression of activin A. In contrast, the mesometrial tissue that maintained high levels of PRL-R mRNA expressed little activin A, but produced an activin-binding protein, alpha(2)-macroglobulin (alpha(2)MG). To determine whether activin A and alpha(2)MG regulate PRL-R expression, antimesometrial and mesometrial cells were separated by elutriation and maintained in culture in the presence or absence of activin A, alpha(2)MG, or follistatin. Just after cell separation, both cell populations expressed PRL-R, but not activin A. Within 6 h, activin A mRNA and protein became highly expressed in the mesometrial cells, whereas PRL-RL mRNA became undetectable. In contrast, activin A mRNA was at very low levels in the antimesometrial cells, and no activin A protein could be detected in the medium for at least 12 h. In these cells PRL-RL mRNA remained elevated. Addition of activin A to antimesometrial cells caused a marked down-regulation of PRL-RL mRNA expression, whereas addition of alpha(2)MG and follistatin to mesometrial cells prevented the disappearance of PRL-R. In summary, the results of this investigation 1) indicate that decidualization of the endometrial stroma induces the appearance of both forms of the PRL-R mRNAs; 2) show differential expression of the PRL-R mRNA in the two-cell population forming the decidua; 3) establish that this differential expression is due to two key decidual molecules, activin A and alpha(2) macroglobulin; and 4) demonstrate that activin A can cause the decidual cells to lose the PRL-R and that the disappearance of the decidual PRL-R can be prevented by addition to the culture of two activin binding proteins, follistatin and alpha(2) MG.