Progesterone-Growth Factor Interactions in Uterine Stromal Cells1

Prior to implantation of the mammalian embryo, a sequence of changes takes place in the endometrium in response to the female sex steroids estrogen and progesterone. Ovarian hormone action on uterine cells is necessary in order to change the uterus from a suboptimal environment to one that is receptive to the blastocyst. The best-characterized action of estrogen and progesterone on the mammalian uterus is stimulation of endometrial cell proliferation and differentiation. While it is clear that hormonal control of proliferation is mediated, in part, through control of the expression of a variety of growth factors and/or growth factor receptors [1–3], recent evidence from breast cancer cell studies elegantly shows direct control of cell cycle regulatory proteins by steroid hormones (reviewed in [4]). Studies concerned with understanding the targets of hormone action thus continue to center on the identification of endocrine-dependent genes that regulate uterine cell proliferation and differentiation. Estrogen and progesterone act on target tissues, such as the uterus, by stimulating the transcription of certain genes. Ligand-bound receptors interact with specific hormone response elements along the DNA of target genes and change the rates of transcription [5]. While this mechanism of hormone action has been clearly demonstrated for a variety of individual genes, the potential targets of hormone action required to stimulate endometrial cell proliferation and differentiation are not clearly defined. The purpose of this review is to summarize the current information regarding the effects of progesterone, the hormone of pregnancy, on control of stromal cell proliferation in the uterus. We focus particularly on the rodent models, since the pattern of cell proliferation in these species has been extensively characterized (see [6]). Our understanding of cell cycle control in hormonally responsive cells, particularly uterine cells, lags behind knowledge of other systems [7] because few models have been available for study of the role of steroid hormones and their receptors in this process. The development of stromal cell culture systems in which progesterone-dependent proliferation is maintained [8, 9] provides an important opportunity for new insight into the molecular mechanisms regulating stromal cell proliferation. Formation and regression of the decidua are critical events to ensure successful pregnancy in many eutherian mammals. By understanding hormonal control of cell proliferation, we can best address its cessation for differentiation (stromal cell decidualization) and later regression (apoptosis). In the rodent uterus, epithelial cells proliferate in response to estrogen for the first few days of pregnancy [10]. At Day 3 of pregnancy in the mouse [11] and Day 4 of pregnancy in the rat [12, 13], there is a proliferative switch from epithelial to stromal compartments. Stromal cells do not proliferate when progesterone receptor antagonists are used to block receptor function [13, 14] or when anti-progesterone antibodies are administered in early pregnancy [15]. The number of synchronously dividing stromal cells in the rat endometrium increases in response to progesterone priming for 3 days followed by nidatory estrogen [16]. Stromal cells are not responsive to estrogen in the absence of progesterone pretreatment, and the requirement for estrogen to initiate stromal cell proliferation can be bypassed when the endometrium of progesterone-primed animals is scratched by a needle or pinched with a hemostat. The endometrium of the ovariectomized rat can be sensitized to respond to a decidual stimulus when exposed to the appropriate sequence of sex steroids. In hormonally sensitized rats, progesterone alone increases DNA synthesis in uterine stromal cells, which later differentiate into deciduomal cells [17]. The incorporation of labeled thymidine into polyploid decidual cells increases after progestin and estradiol treatment. However, once decidualization commences, progesterone alone stimulates DNA synthesis in polyploid deciduomal cells [18]. The case for progesterone control of stromal cell proliferation in the human endometrium is less well characterized. While epithelial and stromal cell proliferation is prevalent in the follicular phase of the menstrual cycle (estrogen dominance), there is a second wave of stromal cell proliferation in the last days of the luteal phase [19]. Progesterone receptors are present in stromal cells at this time, suggesting that progesterone action controls this late wave of stromal cell mitosis [20]. During the luteal phase of the menstrual cycle, extensive vascular development and differentiation of stromal cells into decidual cells occur in response to ovarian hormones, in particular progesterone [21]. These biochemical changes in the uterine stroma of women resemble those that take place during the implantation reaction in rodents. However, these cellular alterations occur spontaneously in women, while in rodents, a blastocyst or external stimulus is required. Whether this difference is due to the aggressive nature of the human trophoblast, as suggested by Finn [21], remains to be demonstrated. Only a few maternally expressed factors such as colony stimulating factor-1 [22], leukemia inhibitory factor [23], and progesterone [24], are absolutely necessary for implantation. While interactions between progesterone and estrogen in many species normally affect target tissues, only progesterone is required for the establishment of pregnancy in all mammals studied to date [25]. Even in species such as the mouse, where estrogen action is normally compulsory, epidermal growth factor can substitute for estrogen in promoting blastocyst implantation during embryonic delay [26]. Since progesterone is the hormone of pregnancy, it is imperative that its molecular mechanisms of action on uterine cell proliferation and differentiation be elucidated. Progress in this area is expected since mice lacking the progesterone receptor exhibit defects in all reproductive tissues [27]. The uterine stromal cells in the progesterone receptor “knockout” mouse are unable to undergo a decidual cell reaction, and the uterus exhibits hyperplasia and inflammation. Since progesterone is a pleiotropic regulator of multiple aspects of reproduction in vivo [27], development of an in vitro system that can specifically address progesterone control of cell proliferation is an essential complement to the progesterone receptor knockout model. Recent progress in understanding the mechanisms involved in cellular differentiation has been possible because of our increased appreciation of the regulation of cell proliferation and cell cycle progression [28]. The convergence of these two fields is critical because differentiation programs are normally initiated as cell proliferation decreases and arrest occurs. This concept, supported by experimental evidence in a variety of cell types, clearly indicates that increased understanding of endocrine control of cell cycle progression is essential before cellular differentiation and apoptosis can be systematically studied in hormone target tissues. Current research in our laboratory is therefore focused on understanding the role of progesterone in regulating stromal cell transit through the cell cycle. Mammalian cells use a small family of related cyclin-dependent kinase (cdk) complexes to regulate progression through the cell cycle [29]. Some of these function during G1 while others are important during the G2 to M transition. Cyclin-dependent kinase complexes are heterodimeric proteins composed of two subunits. One is a catalytic protein that phosphorylates multiple substrates via its kinase activity. The second subunit is a cyclin protein that regulates substrate specificity. The current model for the action of growth factors during G1 of the cell cycle predicts that cyclin D and cyclin E levels increase, resulting in the formation of different cdk complexes that drive cells through the restriction point of G1 into DNA replication [30]. Cyclins are rapidly degraded once the peak concentration is reached, and this degradation is essential before cells go into the next phase of the cycle. Regulated cyclin destruction via the ubiquitin-mediated degradation pathway is an important regulatory mechanism for cell cycle progression [31]. A large number of growth factors and their receptors have been mapped to the various cell types in the uterus of the human and of rodents and domestic animals, and reviews summarizing their distribution are available [1–3]. It is clear that some steroidal effects on the endometrium are mediated by endocrine-dependent synthesis of growth factors and their receptors. Newly synthesized growth factors presumably then act on endometrial cells via mechanisms identified in other cell types. Since hormones may also act directly and stimulate expression of genes required for cell cycle transit, identifying hormone-dependent versus growth factor-mediated changes in cell cycle progression is essential. For example, estrogen and progesterone increase c-jun and c-fos mRNAs in the ovariectomized rat uterus [32, 33]; but it is not clear whether the resulting proteins regulate cell cycle genes, the synthesis of growth factors involved in cell cycle progression, or perhaps both [32]. The difficulties in resolving these issues using an in vivo model are apparent because expression of these protooncogenes does not correlate strictly with cell proliferation in the uterus [34]. The challenge of defining the molecular mechanisms of hormone action on the endometrium are further suggested from the recent studies of Cooke and colleagues ([35]; and see minireview in this issue [36]). The reconstitution experiments using stromal and epithelial cells from estrogen receptor-deficient mice clearly show that epithelial cell proliferation in the mouse is mediated via estrogen receptor action from the stroma. Similarly, analysis of progesterone receptor distribution in a variety of mammalian species indicates an absence of progesterone receptor in epithelial cells at the time of implantation [13, 37]. Thus, epithelial cell differentiation in response to progesterone may be mediated, in part, through stromal cell progesterone receptors. Formal proof for this concept is required, but together these findings reemphasize the importance of the stroma as an epithelial cell modulator in uterine development and the sensitization of the endometrium for implantation. Importantly, these newly emerging concepts about the mechanisms involved in hormonal control of cell proliferation [9, 27, 36] are derived from different experimental approaches, indicating the importance of multifaceted research strategies to study complex biological problems. In humans and in rodents, stromal cells proliferate and differentiate into large, polyploid cells with characteristic morphology [38]. A variety of markers exist that are associated with differentiation of rat [39] and human [40] decidual cells. While these markers have proved useful, their expression does not correlate strictly with the fully differentiated phenotype. Cultured stromal cells expressing “decidual markers” continue to proliferate. Differentiation programs are normally initiated as cell proliferation and growth arrest occur [28]. Therefore, activation of genes that are closely linked to molecular mechanisms involved in cell cycle arrest are expected to initiate a differentiation program. In the presence of charcoal-stripped calf serum (i.e., no steroids), stromal cells do not proliferate significantly [8]. Tessier et al. [41] showed that stromal cells cultured in serum incorporate significantly more [3H]thymidine than cells cultured in charcoal-stripped serum. Progesterone restored the proliferative response when added to charcoal-stripped serum. We have extended those earlier observations of progesterone-dependent stromal cell proliferation and developed a serum-free culture system. It is essential to use a chemically defined culture system in order to identify growth factor-directed versus hormone-dependent cell cycle regulation. Stromal cells are cultured in Dulbecco’s Modified Eagle’s medium containing supplements as described [9]. Under these conditions the cells become quiescent 48–72 h after the start of culture. Even though no net increase in cell number is seen, there may be some proliferation that is balanced by cell loss. Analysis of stromal cells cultured for 72 h in serum-free medium using flow cytometry showed that approximately 80% are in G1 phase of the cell cycle (Fig. 1). The percentage of cells in G1 phase in different experiments ranged from 73% to 83% with use of these same serum-free culture conditions. Stromal cells propagated in medium containing 10% fetal bovine serum (FBS) were asynchronous, and the percentage of cells in G1 72 h after plating was 32% (Fig. 1). However, cell cycle phase parameters in asynchronous cultures depended on the time after plating. At 48 h after plating, 50% of the cells were at G1, while as the cells approached confluence at 96 h after plating, 25% were at G1. Synchronization of uterine stromal cells at G1 phase of the cell cycle. Uterine stromal cells (UII, passage 17) were cultured in serum-free medium (serum starved) or medium 199 containing 10% FBS (asynchronous) for 72 h. Between 1–2 × 106 cells from each culture were analyzed. Cells were collected and stained for at least 4 h with a one-step propidium iodide solution (0.005% propidium iodide, 0.02% ribonuclease, 0.3% sodium phosphate, and 0.1% sodium citrate in distilled water) using standard methods [55]. The cells were analyzed on an EPICS 752 flow cytometer (Coulter Electronics, Hialeah, FL) with a laser output of 400 nW at 488 nm. Red fluorescent data were collected and DNA histograms analyzed with the Modfit program (Verity Software, Topsham, ME). Stromal cells cultured in serum-free medium for 72 h were stimulated to synchronously reenter the cell cycle by addition of basic fibroblast growth factor (bFGF) and progesterone [9]. It should be kept in mind, as the flow cytometry analyses showed (Fig. 1), that all of the cells do not require stimulation into the cell cycle because perfect synchrony is never achieved using cultured cells. However, progesterone and bFGF significantly (p < 0.01) increased stromal cell proliferation in a dose-dependent fashion [9]. Of the growth factors tested in this culture system, epidermal growth factor, transforming growth factor α, and bFGF stimulated stromal cell division significantly (p < 0.01) over that in the control cultures containing serum-free medium alone. None of the growth factors stimulated proliferation significantly in the absence of progesterone, and addition of two growth factors did not stimulate proliferation as compared to progesterone with growth factor [9]. This progesterone-dependent proliferation was specific, because dexamethasone, estradiol-17β, and 5α-dihydrotestosterone did not stimulate proliferation significantly [9]. More recently, we demonstrated that cells treated with bFGF or progesterone alone are unable to progress through G1 and enter DNA replication (Fig. 2). Stromal cells cultured with progesterone and bFGF incorporated [3H]thymidine significantly (p < 0.01) compared to cells treated with progesterone alone, fibroblast growth factor (FGF) alone, or serum-free medium (control) (Fig. 2). To ensure that stromal cells were progressing through the cell cycle, the number of cells over the stimulation period was counted (Fig. 3). Cell number increased slightly at 24 h in cultures containing serum-free medium and serum-free medium with progesterone and bFGF. However, this increase was significantly greater (p < 0.01) at 48 h in cell cultures containing progesterone and bFGF as compared with the control cultures. Analysis of cell cycle phases using flow cytometry showed that the accumulation of cells in S phase, at the expense of a reduction of cells in G1 phase, occurred 24 h after addition of bFGF and progesterone (Fig. 4). These results are consistent with a single stromal cell transit through the cell cycle resulting in the 2-fold increase in cell number over this same time period (compare Figs. 3 and 4). Taken together, these data indicate that progesterone and bFGF stimulate mitosis and not just the rescue of cell death. Synergistic action between progesterone and bFGF stimulates stromal cell progression through G1 and entry into DNA replication. Quiescent uterine stromal cells (UIII, passage 18) were stimulated with either FGF (50 ng/ml), medroxyprogesterone acetate (MPA, 1 μM), the two agents together (FGF+MPA), or serum-free medium alone (Control) for 24 h. The cells were pulsed for 2 h with [3H]thymidine (1 μCi/well). Incorporation of [3H]thymidine was determined by scintillation counting. Results are expressed as the mean ± SEM for 6 independent observations for each treatment. *p < 0.01. Progesterone and bFGF stimulate stromal cell proliferation. Serum-starved stromal cells (passages 17–20) were stimulated with 1 μM progesterone and 50 ng/ml bFGF (dashed line) or were provided medium with vehicle alone (solid line). The number of cells was counted at 0, 24, and 48 h after mitogen addition. The Time 0 was determined from the basal number of cells at the time of progesterone and bFGF addition. After 48 h of culture, cell number increased significantly (p < 0.01) in cell cultures treated with bFGF and progesterone over that in the control conditions. Data shown are the mean ± SEM of triplicate samples. Temporal analysis of cell cycle phases in uterine stromal cells. Cells (UII, passage 17) were synchronized by culture in serum-free medium. The cells were stimulated to reenter the cell cycle by addition of serum-free medium containing bFGF (50 ng/ml) and progesterone (1 μM). Cells were collected at the indicated times, and cell cycle progression was monitored by flow cytometry. A major concern regarding the usefulness of this model is the apparent lack of estrogen-dependent proliferation. While a nidatory amount of estradiol is required to initiate stromal cell proliferation in vivo, no direct effect of estrogen on cell cycle progression in uterine stromal cells has been demonstrated. Therefore, our interpretation of the existing evidence is that estrogen functions in two important aspects of stromal cell proliferation (see Fig. 5). First, estrogen may be necessary in intact animals to stimulate the synthesis of progesterone receptor rather than to direct cell cycle progression. Consistent with this interpretation, estrogen induces transcription of the progesterone receptor via an intragenic response element [42], and Kraus and Katzenellenbogen [43] showed that estrogen is required for progestin action in the ovariectomized rat uterus because the progesterone receptor content is low. We have clearly demonstrated that proliferating uterine stromal cells express both the A and B forms of the progesterone receptor [9]. Moreover, addition of progesterone receptor antagonist RU-486 to stromal cell cultures containing bFGF and progesterone significantly inhibited proliferation [9]. Since the serum-free medium used in our studies is phenol red free, it is not clear how progesterone receptor expression is regulated in these cells. The simplest interpretation is that increased expression of progesterone receptor occurs during the prior passage of these cells in medium containing 10% serum. progesterone receptor synthesis may be the of a between the estrogen receptor to an mechanism as shown for cells in culture since is a in the serum-free medium used in our cell proliferation of estradiol to stromal cell cultures containing progesterone and bFGF the proliferative response as compared to that with progesterone and bFGF but the difference was not [9]. We are whether stromal cells in culture express estrogen receptors. model for progesterone-dependent stromal cell proliferation potential control for cell cycle progression. Cells are into S and M phases by the formation of Progesterone may stimulate the synthesis of in G1 and G2 phases of the cell cycle. of the complex kinase to and Estrogen may be necessary for the synthesis of growth factors and progesterone receptor in vivo, but its action is not required for stromal cell proliferation in culture. for The second important function of estrogen in vivo may be to stimulate the of growth factors synthesis is not required when are added to the culture system. We have shown that proliferating stromal cells in culture express a single of fibroblast growth factor the potential of these cells to respond to added bFGF. of to stromal cell cultures containing bFGF and progesterone significantly inhibited proliferation, indicating that bFGF was to its receptor [9]. In the ovariectomized rat uterus, bFGF levels increase in response to estrogen and progesterone [32]. Stromal cells in culture do not proliferate in response to progesterone alone suggesting either that an amount of bFGF is in these cells or that the mechanisms for bFGF in cultured stromal cells are hormones act as transcription thus it is that cell cycle genes are one of the targets for progesterone Estrogen and progesterone stimulate the expression of and in breast cancer cells, and the of the response is to that when transcription of these genes is stimulated by growth factors In breast cancer cell cyclin is responsive to both estrogen and progesterone However, the response to progesterone is of greater results in our laboratory indicate that cyclin is one target for progesterone receptor in quiescent uterine stromal cells stimulated to proliferate with progesterone and bFGF After DNA stromal cells differentiate into decidual cells with characteristic changes in morphology [38]. cells express and and are by In the mouse uterus, a of stromal cells their DNA but do not undergo mitosis are as the to at and increased from 4 to is While is not in mammalian cells, it has been in cells and in during differentiation In the of factor DNA replication from mitosis Results from cell cycle analysis of uterine stromal cells are expected to new insight into the role of progesterone and its receptor in the control of cell cycle progression and stromal cell differentiation. Progesterone is the only hormone required for the establishment of pregnancy in all its mechanism of action in endometrial cells is from a The results emerging from cell cycle analysis can be used as a to the molecular targets for progesterone-dependent proliferation. Progress in this area is essential to identify the targets for progesterone-dependent cell cycle progression and to the for studies to systematically endocrine control of cellular differentiation and We for The are to and present of the laboratory research has to this