Extensive evidence demonstrates pronounced effects of relaxin on the differentiation of human endometrial cells in vitro. In vivo data in rhesus monkeys suggest a role for relaxin in the development of endometrial vascular architecture. In women, pregnancy can be established and maintained in the absence of circulating relaxin. Thus, local synthesis by the endometrium is necessary if relaxin plays a physiological role in human endometrial function. Although relaxin protein and the prorelaxin C peptide have been localized to human endometrium, no data for relaxin synthesis have been provided to date. We therefore assessed relaxin mRNA and protein levels in cultured, defined human endometrial cells. Reverse transcriptase-polymerase chain reaction (RT-PCR) techniques were used to demonstrate the presence of relaxin mRNA in human stromal and glandular epithelial cells. Secretion of the protein into the media of cultured cells of both types was also detected. Relaxin stimulated the expression of vascular endothelial growth factor in glandular epithelial and stromal cells that were isolated from tissue that had been taken during the secretory phase of the cycle. Relaxin inhibited the expression of procollagenase from both glandular epithelial cells, with a more marked inhibition demonstrated from cells that were isolated from tissue that had been taken during the secretory phase, and from stromal cells. These data demonstrate that human endometrial cells synthesize relaxin, and they support the concept that relaxin fosters endometrial conditions that are required for implantation in women.

Prolactin (PRL) and insulin-like growth factor-binding protein (IGFBP-1) are two major secretory proteins of human endometrial/decidual cells. We have characterized the mRNA of PRL and IGFBP-1 and studied the effect of progestin, medroxyprogesterone acetate (MPA), anti-progestin (RU486), and relaxin (RLX) on the levels of these two mRNA transcripts in a long-term culture of human endometrial stromal cells. Northern blot analysis showed that the size of PRL mRNA was 1.15 kb and that of IGFBP-1 mRNA, 1.6 kb. Primer extension of endometrial/decidual IGFBP-1 mRNA showed two transcription initiation sites identical to those found in HepG2 human hepatoma cell line. The levels of mRNA in control samples remained low, approximately 2 pg PRL and approximately 5 pg IGFBP-1/microgram RNA at various times of culture. When stromal cells were treated with MPA for 28 days, PRL mRNA gradually increased 100-fold whereas IGFBP-1 mRNA exponentially increased approximately 1000-fold compared to control values and leveled after 25 days in culture. The timing of maximal stimulation was shortened by withdrawing MPA or by replacing MPA with RU486. After removal of MPA, levels of both mRNAs increased and each peaked after approximately 10 days, with PRL showing a 2-fold and IGFBP-1 a 20-fold increase compared to cells treated with MPA continuously. Replacing MPA by RU486 caused a rapid increase of PRL mRNA (2-3-fold) in 2-3 days followed by a gradual reduction to less than 20% of peak levels over the next 3 days. IGFBP-1 mRNA levels increased 30- and 100-fold in 1-2 days followed by a reduction to less than 20% of peak levels over the next 24 h. The reduction of mRNA levels by RU486 was reversed when cells were rechallenged with MPA. Relaxin alone caused a transient stimulation of PRL and IGFBP-1 mRNA. Maximal stimulation occurred between 10 and 20 days of culture and was 100-fold for PRL and 1000-fold for IGFBP-1 relative to control values. Cells treated with MPA and RLX in sequence had higher mRNA levels than cells treated with MPA continuously or cells subjected to MPA withdrawal. Maximal mRNA levels reached 0.4 ng PRL and approximately 8 ng IGFBP-1/microgram total RNA, approximately 0.04% and 0.8% of cellular RNA. The mRNA levels under various hormonal manipulations were similar to the previously published synthesis and secretion patterns of PRL and IGFBP-1 proteins in this system.(ABSTRACT TRUNCATED AT 400 WORDS)

Estradiol-17beta dehydrogenase activity in proliferative human endometrium (average of 1.5 nmole of estrone formed from estradiol/mg protein/h) was stimulated as much as as 6-fold during incubations of tissue slices in culture medium containing progesterone. Stimulation was already detectable at 7 h and the highest activity values were reached at 48-72 h of incubation in the presence of excess progesterone. Maximal stimulation was achieved with concentrations of the hormone of 0.25 mug/ml or higher. At concentrations approximately equal to midluteal plasma levels (20 ng/ml) more than 50% of the maximal response was observed. Norgestrel (17alpha-ethynyl-18-methyl-19-nortestosterone) was also effective in inducing enzymatic activity. The similarity of the effects obtained with progesterone (a possible substrate for estradiol dehydrogenase) and the synthetic progestin indicates that the stimulation of enzymatic activity was not due to substrate induction. Addition of estradiol to the culture medium had no influence on the activity of the enzyme. The induction of estradiol dehydrogenase by progesterone was inhibited by puromycin or actinomycin D. These observations indicate that progestational agents increase the rate of de novo synthesis of the enzyme. Stimulation of endometrial estradiol dehydrogenase was also observed after 2-3 day oral administration of medroxyprogesterone acetate to women in the follicular phase. In contrast, the enzymatic activity in endometrium obtained from women taking estrogens was found to be as low as in normal proliferative tissue. These in vitro and in vivo results point to progesterone as the agent responsible for the 10-fold increase in endometrial estradiol dehydrogenase activity observed during the luteal phase in menstruating women. Data obtained from superfusion studies of estrogen dynamics in endometrium indicate that changes in enzyme concentrations may play a physiologic role in the regulation of tissue levels of estradiol.