Various matrix metalloproteinases (MMPs) participate in the menstrual breakdown of the human endometrium. MMP-9/gelatinase B is proposed as a major factor because it degrades many extracellular matrix constituents, including in the vasculature. Although globally under ovarian steroids control, endometrial MMP-9 seems expressed differently than other MMPs, and conflicting publications prevent a clear understanding of its regulation. We therefore quantified MMP-9 expression in the cycling human endometrium, defined its localization, and analyzed its regulation by estradiol and progesterone and by LEFTY-A/endometrial bleeding-associated factor in explant cultures. In fresh tissues, a major increase in MMP-9 mRNA expression occurred at menstruation, after a larger increase in LEFTY-A mRNA. MMP-9 was immunodetected in all cell types throughout the cycle, especially in foci of stromal cells during menstruation. MMP-9 synthesis by these cells was confirmed in cultured explants. In proliferative explants, ovarian steroids slightly decreased MMP-9 mRNA. They had no consistent effect on MMP-9 release in culture medium but strongly inhibited proMMP-9 activation. Addition of recombinant LEFTY-A to explants induced MMP-9 in most samples, a response prevented by ovarian steroids. We propose that endometrial MMP-9 activity is overall controlled by the ovarian steroids and locally adjusted through a network of modulators, including LEFTY-A.

TGF-betas have been reported to mediate the repression by progesterone of several matrix metalloproteinases in the human endometrium, thereby preventing menstrual breakdown. Because of conflicting reports on the expression profiles, source, and regulation of the TGF-beta system in this tissue, we investigated by real-time RT-PCR and ELISA the expression of the three TGF-betas (total and mature forms) and their two receptors throughout the menstrual cycle, and their regulation by ovarian steroids in cultured explants including their microdissected epithelial and stromal compartments. Regulation by cAMP and MAPK was further investigated. This comprehensive study on a large collection of endometrial samples evidenced a differential regulation of TGF-beta isoforms expression, both in vivo and in explant culture. In vivo, TGF-beta2 increased by about 5-fold at the mid-late secretory phase then declined after menstruation; TGF-beta3 increased at menstruation and remained high during the proliferative phase; TGF-beta1 was maximal at menstruation. In explants cultured without ovarian steroids both TGF-beta2 and -beta3 were preferentially expressed in the stroma. Ovarian steroids strongly repressed both TGF-beta2 and -beta3 in stroma but only TGF-beta2 in glands. cAMP prevented inhibition by ovarian steroids of TGF-beta2 but not -beta3. In presence of ovarian steroids, MAPK inhibitors (p38 and ERK pathways) stimulated TGF-beta3 but inhibited TGF-beta2 expression. In conclusion, TGF-beta2 and -beta3 are differentially expressed during the menstrual cycle and regulated by progesterone in epithelial vs stromal cells. The opposite regulation of TGF-beta2 and -beta3 by cAMP and MAPK could account for their distinct expression in vivo.

Both phase- and gene-specific modulators finely tune in space, time, and amplitude the global control of MMPs and TIMPs mRNAs by estradiol and progesterone in the cycling human endometrium.

Recent studies strongly suggest that matrix metalloproteinases (MMPs) play a key role in the initiation of menstrual bleeding in the human endometrium upon the fall of ovarian steroid serum concentrations by inducing the degradation of the extracellular matrix of this mucosa. MMPs are also involved in abnormal endometrial bleeding and have been identified in endometriotic foci. In all cases, they are associated with areas of extracellular matrix breakdown. This paper reviews the literature on the regulation by estradiol and progesterone of the expression and activation of MMPs, and of the expression of their tissue inhibitors (TIMPs), (i) in the endometrium in situ during normal cycle, (ii) during artificial cycles in spayed monkeys, and (iii) in cultures of endometrial explants or purified cells. Whereas progesterone consistently decreases the activity of endometrial MMPs, its effects vary in intensity, duration, and pattern between MMPs as well as among experimental systems. The contribution and limitations of the various investigations are therefore discussed. The focal heterogeneity points to additional local controls of the expression and activation of MMPs in human endometrium, acting beyond the general inhibitory role of progesterone, for example, by cytokines. Focal changes in type or abundance of sex steroid receptors also could be responsible for spatial variation in the expression of MMPs in the endometrium and endometriotic lesions.