Progestin-epidermal growth factor interactions regulate decidualization-related TF expression

Frederick Schatz, Ph.D.

The absence of estrogen and progesterone canonical response elements from the TF gene promoter suggests that autocrine and/or paracrine factors mediate ovarian steroid effects on TF expression. Progestin-epidermal growth factor (EGF) interactions modulate growth as well as prolactin, laminin and fibronectin expression in cultured stomal cells. The cellular effects of EGF are initiated by binding to its cell surface-sequestered receptor (EGFR), a structural homologue of the c-Erb b oncogene protein product. Ligand binding induces the EGFR to homodimerize or to heterodimerize with other members of the EGF/ErbB receptor family, which triggers complex downstream intracellular gene-activating phosphorylation pathways.

Levels of EGFR in cycling human endometrium was similar to that seen for TF. Thus, IHC staining for the EGFR was diffuse in follicular phase stromal cells, but increased in conjunction with the progression of decidualization in stromal cells of mid-late luteal phase specimens. Western blotting supported these in vivo observations by demonstrating that cultured stomal cells contain a doublet at 170 kDa that corresponds to the EGFR. The magnitude of this doublet increased two to three-fold in incubations with E2 + MPA compared with E2 alone. These observations with cultured HESCs are consistent with higher EGFR levels reported in luteal than follicular phase human endometrial extracts.

That EGF and transforming growth factor–alpha (TGF-a) bind to and activate the EGFR with equal affinity, prompted comparison of their effects on TF expression in cultured HESCs. Northern blotting revealed that neither EGF, nor TGF-a, nor E2 + MPA was sufficient to alter TF mRNA levels. By contrast, TF mRNA levels were elevated several-fold when E2 + MPA was added with either of the EGFR agonists. Similarly, Fig 1 shows that immunoreactive TF levels were unaffected by incubation of HESCs with either EGF or E2 + MPA, but increased several-fold after incubation with EGF and E2 + MPA. Figure 1 also shows that neither transforming growth factor–beta (TGF-b) nor interleukin-1 beta (IL-1b), affected TF expression whether added alone or with E2 + MPA. Neither TGF-b nor IL-1b activates the EGFR, emphasizing the role of the EGFR in elevating TF levels during progestin-regulated decidualization.

Figure 1 – Growth factor, cytokine and steroid effects on TF levels in cultured stromal cells. The stromal cells were pre-incubated for 4 days in serum-supplemented medium with 0.1% ETOH (C), or 10-8M E2 (E) + 10-7 M MPA (P). They were then incubated for 36 hours in a defined medium containing corresponding C or E+P with and without 50ng/ml EGF, or 2ng/ml TGFb or 1.0 U/ml I1-1b. Ordinate = secreted levels of ir TF by ELISA normalized to cell protein. Means +/- SEM, n=4; C vs. E+ P + EGF (P<0.03 by the Mann-Whitney Rank Sum Test).

Human endometrial endothelial cells (HEECs): an in vitro model for studying the hemostatic/thrombotic balance
The expression of TF transforms the endothelium from an anti-coagulant to a pro-coagulant, pro-thrombotic surface. Although endothelial cell biology has relied on studies with cultured human umbilical venous endothelial cells (HUVECs), endothelial cell phenotype is species, organ and vessel-type specific and controlled by perivascular cells. Recently, we isolated endothelial cells from the vasculature of cycling endometrium. Markers for potentially contaminating epithelial, stromal, smooth muscle and bone marrow-derived cells were absent from human endometrial endothelial cells (HEECs), whereas HEECs displayed prominent immunostaining for several endothelial cell-specific surface markers. Cultured HEECs also exhibited such endothelial cell-specific characteristics as tube formation on Matrigel and uptake and sequestration of acetylated-LDL.

To study potential prothrombotic changes in human endometrial endothelium, the protein kinase c agonist, phorbol myristate acetate (PMA) was evaluated on TF expression in HEECs. PMA elevated TF levels by fifteen and six-fold after two and twenty hours respectively (Fig 2A), whereas HEEC-secreted PAI-1 levels was constitutive (Fig. 2B). This differential response in HEECs contrasts with the elevation of both TF and PAI-1 levels in other endothelial cell types by growth factors and cytokines. While TF and PAI-1 each promote hemostasis, their overexpression elicits thrombosis. Therefore, the selective enhancement of TF expression in HEECs, emphasizes their importance in evaluating hemostatic-thrombotic regulatory mechanisms in human endometrium.

Figure 2 – PMA effects on immunoreactive TF and PAI-1 levels in HEECs. Confluent fifth passage HEECs from 5 preparations were incubated with 0.1% ethanol (vehicle control) or 25ng/ml PMA for two hours and 19 hours. A) ir TF analyzed by ELISA in the cell pellet. B) ir PAI-1 analyzed by ELISA in the medium * Control versus PMA, p < 0.01, n=5; ** Control versus PMA, p < 0.05, n=5 by Mann Whitney test.