Growth Factor Control of Cultured Rat Uterine Stromal Cell ...

BIOLOGY OF REPRODUCTION 55, 1333-1342 (1996)

Growth Factor Control of Cultured Rat Uterine Stromal Cell Proliferation Is Progesterone Dependent' Marta Piva, Oliver Flieger, and Virginia Rider 2 School of Biological Sciences, Division of Molecular Biology and Biochemistry, University of Missouri-Kansas City, Kansas City, Missouri 641 10 ABSTRACT Uterine stromal cells undergo mitosis and differentiate into the decidua just prior to the expected time of implantation in humans and rodents. We have utilized a culture system that will be suitable for study of the molecular mechanisms regulating stromal cell proliferation. Stromal cells were isolated from the uteri of ovariectomized rats and were cultured in chemically defined medium. Cultured cells express the mesenchymal markers vimentin and desmin. They do not express the epithelial marker cytokeratin. Serum-starved stromal cells were stimulated to proliferate in a time frame consistent with the cell cycle through addition of a panel of growth factors (basic fibroblast growth factor [bFGF], epidermal growth factor, platelet-derived growth factor, transforming growth factor a, insulin-like growth factor I) and hormones to the culture medium. None of the growth factors tested significantly stimulated proliferation in the absence of progesterone. Furthermore, progesterone was the only steroid of those tested that stimulated mitosis in the presence of growth factors. Stromal cell proliferation in response to progesterone and bFGF was dose dependent and saturable. Addition of the progesterone receptor antagonist mifepristone (RU 486) and an inhibitor of tyrosine kinase receptor activation (suramin) abolished stromal cell mitosis. Progesterone receptors and fibroblast growth factor receptor 1 (FGFR1) were identified by immunoblot analysis in proliferating stromal cells. Taken together, these results show that cultured stromal cells maintain progesterone-dependent cell cycle control that ismediated via progesterone receptors. Moreover, the data indicate that bFGF control of stromal cell proliferation is modulated via a specific isoform of FGFR1 containing the three-loop immunoglobulin-like domain. INTRODUCTION The importance of the chronological relationship between the blastocyst and the uterus of eutherian mammals has been confirmed for all species studied [1]. Implantation of the embryo can occur only when both the blastocyst and the endometrium have attained a certain degree of maturity. The period at which the uterus is sufficiently mature to accept the blastocyst is called the time of receptivity. Beyond this restricted period, the intrauterine survival of embryos does not happen and the uterus is said to be refractory or in a state of nonreceptivity. An important correlate of uterine receptivity is proliferation of uterine stromal cells [2]. Stromal cells undergo mitosis and differentiate into the decidua just before the expected time of implantation in humans [3] and rodents [4, 5]. While polypeptide growth factors mediate cell proliferation in a variety of tissues [6], Accepted July 31, 1996. Received March 25, 1996.

'This work was supported by a grant from the University of Missouri Research Board. 2 Correspondence: Virginia Rider, School of Biological Sciences, University of Missouri-Kansas City, 5007 Rockhill Road, Kansas City, MO 64110. FAX: (816) 235-5158.

in the reproductive organs of female mammals a balance of estradiol and progesterone largely controls cell proliferation (reviewed in [7]). Estrogens are associated generally with cell multiplication in uterine and breast tissues, while progesterone is considered more as the hormone promoting cellular differentiation in these organs. However, progesterone is a potent mitogen for stromal cells in the uterus [8] and of lobuloalveolar cells in the mammary gland (reviewed in [7]). Progesterone is required by all mammalian species to prepare the uterus for implantation of the ovum (reviewed in [9, 10]). Hormonal preparation of the endometrium involves cellular proliferation, growth, and differentiation. In the mouse and rat uterus, cell division switches from the luminal and glandular epithelium to the endometrial stroma at Days 3 [11] and 4 [12], respectively. Stromal cells do not divide without progesterone, and proliferation is blocked by progesterone antibodies [13]. Failure of stromal cell division and alteration of progesterone receptor distribution after mifepristone (RU 486) administration suggest that progesterone action is mediated through its receptor protein [14, 15]. While the molecular mechanisms that regulate the proliferation and differentiation of endometrial cells in response to the ovarian hormones are not known, recent evidence suggests that steroid action stimulates the local production of growth factors and their receptors (reviewed in [16]). Growth factors in turn have been implicated in mediating the action of steroid hormones [17]. Stromal cells are dependent upon progesterone for their proliferation and differentiation in the pregnant endometrium [18], in the uterus of ovariectomized steroid-treated rodents [8], and in the immature uterus of neonatal mice [19]. Progesterone is required for proliferation of cultured human [20] and rat [21] uterine stromal cells. While progesterone is necessary, it is not sufficient, since cultured stromal cells require additional factors for cell division [20, 21]. Human stromal cells in culture respond to a panel of growth factors when progesterone is added to the medium [20]. This progesterone dependence is variable, however, since 3 H-thymidine can be incorporated in a dose- and time-dependent fashion in response to epidermal growth factor (EGF), platelet-derived growth factor (PDGF), insulin-like growth factor I (IGF-I), and fibroblast growth factors (FGFs) in the absence of hormone [22-25]. It has been reported that rat stromal cells absolutely depend on progesterone for proliferation in culture; however, additional unidentified factors present in serum are also required [21]. Since we are interested in the hormonal control of stromal cell proliferation, the purpose of the present study was to develop stable cultures of rat uterine stromal cells that respond consistently to mitogens in serum-free medium, and in a time frame consistent with the cell cycle. Rodent stromal cells definitely require progesterone in vivo to proliferate [11, 12]. Therefore, an additional aim for the present study was to determine whether rat stromal cells

1333

1334

PIVA ET AL.

maintain progesterone-dependent proliferation in culture. Such a system will allow dissection of the molecular mechanisms regulating progesterone-dependent control of the cell cycle. In the last five years, increased understanding of cell cycle regulation at the molecular level has emerged (reviewed in [6, 26]). Cell cycle control by progesterone, however, is an area of biology that has not been extensively studied from a regulatory viewpoint. Increased knowledge about the checks and balances that normally ensure orderly cell division in response to progesterone can lead to the development of effective strategies to reduce infertility and combat the uncontrolled cell divisions that occur in breast and uterine cancers.

5255), and anti-pan cytokeratin (1:200 dilution; Sigma #C2931). Secondary antibody for desmin was goat anti-rabbit IgG/fluorescein isothiocyanate (FITC) (1:50 dilution; Sigma). Goat anti-mouse IgG/FITC (1:25 dilution; Sigma) was used to detect vimentin and cytokeratin. Coverslips were washed in PBS-BSA three times and incubated with secondary antibodies for 30 min at 4C. Coverslips were washed in PBS-BSA, and cells were mounted in 50% glycerol. Control cells were incubated without primary and secondary antibodies. Cells were observed using an Olympus microscope (Olympus Corp., Tokyo, Japan) equipped with epifluorescence and photographed with Tmax (Eastman Kodak, Rochester, NY) film.

MATERIALS AND METHODS

Cell Proliferation

Animals

Medium with serum. Cells (1-3 x 104) were seeded into 24-well regular tissue culture plastic plates (Falcon, Lincoln Park, NJ). Plating efficiency was greater than 80%. Quiescence was induced by culturing cells in medium containing 10% charcoal-stripped calf serum (CSCS; Sigma) as described previously [21]. Test substances were added to quiescent cultures, and cell proliferation was measured 48 h later using the Cell Titer 96 kit (Promega, Madison, WI) according to the manufacturer's recommendations. This assay relies on the conversion of a tetrazolium salt, 3-[4,5dimethylthiazol-2-yl] diphenyltetrazolium bromide (MTT), into a formazan product by active mitochondrial dehydrogenases of living cells [28]. The sensitivity of this method is comparable to that of 3 H-thymidine incorporation, with less than 5% difference between the two assays [29, 30]. Serum-free medium. Cells (1-3 X 104) were seeded into 24-well plates. To achieve cell cycle synchronization, cells were cultured for 72 h in serum-free, phenol red-free Dulbecco's Modified Eagle medium (Gibco, Grand Island, NY) and MCDB-105 (Sigma) in a 3:1 mixture containing supplements as described previously [20], including 100 U/ml penicillin, 100 jig/ml streptomycin, 5 pig/ml bovine insulin (Sigma), 10 ig/ml human transferrin (Sigma), 50 jig/ml ascorbic acid (Sigma), and 1 mg/ml reagent grade BSA (Sigma). Test substances were added to serum-starved cultures, and proliferation was measured 48 h later using the Cell Titer 96 kit. Test substances. Hormones were purchased from Sigma and dissolved in absolute ethanol. Control cultures were treated with vehicle (ethanol, buffer, or sterile water) alone. Growth factors including basic fibroblast growth factor (bFGF, human recombinant; Upstate Biotechnology [UBI], Lake Placid, NY), EGF (murine, UBI), IGF-I (human recombinant; Collaborative Biomedical Products [CBP], Bedford, MA), PDGF (human natural; CBP), and transforming growth factor ot (TGFa, human recombinant; CBP) were dissolved in accordance with the suppliers' instructions. The concentrations of hormones used were in the physiological range for estradiol-171B and progesterone [20], dexamethasone [31], and 5a-dihydrotestosterone [32]. The concentrations of growth factors have been shown previously to stimulate human stromal cell proliferation in culture [20]. Suramin was a gift from Jim Rusche at Glycan Pharmaceuticals (Boston, MA). Mifepristone (RU 486) was a gift from Romana Nowak at Harvard Medical School (Boston, MA). Statistical analysis. Differences among treatments were examined using a one-way analysis of variance, and significant differences among groups were detected by Dunnett's and Student-Newman-Keuls tests [33] as appropriate.

Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA) between 150 and 175 g in BW were bilaterally ovariectomized. Animals were rested for 10 days after surgery and provided rat chow and water ad libitum. Rats were treated in accordance with the principles and procedures outlined in the NIH Guidelines for Care and Use of Experimental Animals. Uterine horns were removed under anesthesia and opened longitudinally under a dissecting microscope. Stromal cells were isolated using modifications of previously described methods [21, 27]. Briefly, the uteri were incubated in sterile dissociation buffer (DB; PBS without Ca and Mg containing 0.25% trypsin, 20 mM Hepes, 50 U/ml penicillin, 50 pig/ml streptomycin) for 35 min at 37°C. Chemicals were purchased from Sigma Chemical Company (St. Louis, MO). The DB was removed and discarded. Fresh DB was added to the uterine horns, and incubation was continued for 60 min. Released cells were transferred to a sterile 15-ml tube, the uteri were washed with fresh DB, and the washings were added to the 15-ml tube. Fetal bovine serum (FBS; Sigma) was added to 0.1 final volume to inhibit trypsin activity, and the cells were pelleted by centrifugation at 500 X g for 5 min. The pellet was resuspended in growth medium [21] (Medium 199 supplemented with Earle's salts and L-glutamine [Fisher, St. Louis, MO] containing 100 U/penicillin, 100 jig/ml streptomycin, and 10% FBS). Cells were seeded into 60-mm culture dishes (Fisher) and cultured in a humidified atmosphere containing 5% CO 2 at 37°C. Confluent cultures were released from the dishes by incubation in single-strength trypsin-EDTA (Sigma) for 5 min at 37C. Cells were transferred to a sterile 15-ml tube containing 0.1 vol FBS, ceng for 5 min, and resuspended in growth trifuged at 500 medium. Seeding was at a ratio of 1:3. Cells were 95% confluent in 48 h, and viability with the use of 0.2% trypan blue (Sigma) was greater than 95%. Immunofluorescence Cells were cultured on sterile 12-mm coverslips (Fisher) coated with 5% gelatin (275 bloom; Fisher) for 48-72 h. Coverslips were rinsed in PBS twice and fixed in 1% (w: v) paraformaldehyde (Fisher) for 5 min at 4C. The paraformaldehyde was removed, and coverslips were rinsed three times in PBS containing 1% BSA (PBS-BSA). Cells were permeabilized with PBS containing 5% Triton X-100 (Fisher) for 15 min at 220 C. Coverslips were washed three times in PBS-BSA and incubated with primary antibodies at 40C for 12 h. Antibodies were anti-desmin (1:50 dilution; Sigma #D8281), anti-vimentin (1:200 dilution; Sigma #V-

PROGESTERONE-DEPENDENT STROMAL CELL PROLIFERATION

1335

FIG. 1. Immunocytochemical analysis of cultured uterine stromal cells. Uterine stromal cells were seeded onto coverslips and cultured for 48 h. Cells were reacted with antibodies against vimentin (A), desmin (B), and cytokeratin (C)as described in the text. Controls for antibody specificity included a murine myoblast cell line that was negative for vimentin (D) and positive for desmin (E), and a human epithelial cell line that was immunoreactive to the cytokeratin antibody (F). A-C are rat uterine stromal cells; D-E are murine myoblasts; F is keratinocyte epithelial cells. Bars = 200 jim.

Western Blots Total proteins were extracted from stromal cells and Escherichiacoli XLI-Blue strain (Stratagene, La Jolla, CA) as described elsewhere [34], while stromal cell membranes were isolated according to Cohen et al. [21]. For progesterone receptor detection, the protein extract was combined with an equal volume of phosphocellulose (Pharmacia, Piscataway, NJ). Samples were centrifuged in a Fisher microcentrifuge at 12 000 x g for 15 min at 40C. Protein concentration was determined in the unbound fraction using the Bio-Rad (Richmond, CA) reagent [15]. Stromal cell proteins (100 !xg) and isolated membrane proteins (25 jig) were electrophoresed through SDS-polyacrylamide gels (7.5%) by standard methods [35], and proteins were transferred to nitrocellulose membranes (Micron Separation Inc., Westboro, MA). Western blots were prepared as described earlier [15, 36]. Washed blots were reacted with a fibroblast growth factor receptor 1 (FGFR1) polyclonal antibody that recognizes FGFRla and FGFRIP isoforms (UBI), diluted 1:250, and a progesterone receptor monoclonal antibody (20 pig/ml; kindly provided by N. Weigel and D. Edwards, Baylor College of Medicine, Houston, TX). Immunoreactive proteins were visualized after treatment with the TMB substrate kit for horseradish peroxidase (Vector Labs., Burlingame, CA). We have shown previously that these antibodies detect FGFR1 o and FGFR1 3 isoforms [36] and progesterone receptor [15] in rat uterus. RESULTS Morphology and Immunocytochemical Analysis Three separate isolations of uterine stromal cells (UIUIII) resulted in stable cultures with similar immunoreactivity to vimentin, desmin, and cytokeratin. Cells from these

three isolations were frozen, stored in liquid nitrogen, and passaged more than 30 times, with consistent and predictable responses to mitogens. Stromal cells in culture are mostly fibroblast-like in appearance, although some cells were enlarged and more polygonal in shape (Fig. 1, A and B). Binucleate cells constituted approximately 5% of the population. Stromal cells were positive for vimentin (panel A) and desmin (panel B). We were unable to detect cytokeratin immunoreactivity (panel C) in any of the cultures tested (passages 4, 7, 18, and 26). In the absence of primary and secondary antibodies, no immunofluorescence was detected (not shown). Specificity of these antibodies was confirmed further by similar culture and immunocytochemical analysis of a murine myoblast cell line (C2C 12; ATCC CRL 1772, Rockville, MD) and a keratinocyte epithelial cell line (a gift of L. Hutt-Fletcher). Myoblasts were not reactive to vimentin antibody (panel D) but were positive for desmin (panel E). The epithelial cells (panel F) reacted uniformly to the cytokeratin antibody. Progesterone and bFGF Stimulate Stromal Cell Proliferation Quiescence was induced by culturing stromal cells in medium containing 10% CSCS for at least 48 h. Estradiol173 (10 nM), progesterone (1 gIM), and a panel of various growth factors were added to the quiescent cells (Fig. 2). Cell proliferation was measured 48 h later using the Cell Titer 96 kit. In the presence of CSCS, which contains no measurable amounts of estrogens or progesterone, stromal cells do not proliferate (Fig. 2, basal control). Of the panel of growth factors tested, only bFGF stimulated proliferation significantly (p < 0.05) over that in the control cultures. None of the growth factors tested stimulated mitosis to the same extent as the medium containing 10% FBS (positive

1336

PIVA ET AL. -

0 0 0 0

ul

A

M ul

1.20

0.80

0.96

0.64

0.72

-

-

C

~

~

~

I

0

0.48

0.48

0.32

0.24

0.16 0.00

A Control

0.00

I bFGF

EGF

PDGF TGF-alpha

I 0'a,

. q

.

Q, " sx V

.

x

F(Cs

FIG. 2. Cultured stromal cells require bFGF and steroid hormones for proliferation. Quiescence was induced in uterine stromal cells by culture in medium containing 10% CSCS. Two separate isolations of stromal cells (UI, passages 20-26; UII, passages 23-24) were treated with a panel of growth factors (bFGF, 50 ng/ml; EGF, 50 ng/ml; PDGF, 1.25 ng/ml; TGFt, 10 ng/ml) in medium containing 10% CSCS, estradiol-17 (10 nM), and progesterone (1 F.M). The control cultures received medium containing either 10% CSCS (basal control) or 10% fetal calf serum (FCS, positive control). Proliferation was measured 48 h after addition of the test substances using the Cell Titer 96 kit. Values shown are the means +_SEM of results from three independent experiments in triplicate. * p < 0.05; ** p < 0.01, Dunnett's test.

B

,

*L+ 5 StiX

-f

I

1.00

,I

0.80

S

C S .

,

0.60

I I,

0.40

I I, I

I

0.20 0.00 0

control). The results were similar between the two different isolations. It was reported in previous studies that human stromal cells retain progesterone-dependent proliferation in chemically defined medium [20]. A serum-free culture system provides an attractive method to identify components required for proliferation under chemically defined conditions. To test whether bFGF and steroid hormones could stimulate stromal cell proliferation in the absence of serum, cells were cultured in phenol red-free, serum-free medium for 72 h. To stimulate proliferation, estrogen, progesterone, bFGF, and combinations of these treatments were added to serum-starved cells (Fig. 3A). Negative controls contained 0.1% (v:v) ethanol in medium alone. Proliferation was measured 48 h after addition of the test substances using the Cell Titer 96 assay. Of the combinations of test substances added, only bFGF in combination with progesterone (bFGF+P) or with progesterone and estradiol (bFGF+ P+E) stimulated stromal cell proliferation significantly (p < 0.01) over that in the negative control group. Progesterone and estradiol alone or in combination did not stimulate stromal cell proliferation significantly (p > 0.05). In addition, bFGF did not stimulate stromal cell mitosis in the absence of progesterone, even in combination with estradiol (p > 0.05). Taken together, these results show that progesterone is necessary for stromal cell proliferation but that it is not sufficient. To ensure that the MTT assay was measuring cell proliferation and not the rescue of dying cells, stromal cells were cultured under standard serum-starved conditions. The cells were collected at times 0, 24, and 48 h after the addition of bFGF (50 ng/ml) and progesterone (1 ELM) or serum-free medium without proliferative additives (Fig. 3B). In both groups, the absorbance was increased as compared with that at the 0 time point (Fig. 3B). After 48 h of culture, proliferation was significantly increased (p < 0.01) in cell cultures treated with bFGF and progesterone over

24

48

Time (h)

FIG. 3. Uterine cell proliferation requires progesterone and bFGF. A) Serum-starved uterine stromal cells (UII, passage 22) were cultured in phenol red-free, serum-free medium containing the following agents: progesterone (P) added at 1 M; 17J-estradiol (E)added at 10 nM; and bFGF added at 50 ng/ml. Control cultures contained medium with ethanol vehicle. Cells were measured for the proliferative response 48 h after addition of test substances. Data are expressed as the means + SEM for two independent experiments assayed in triplicate. **p < 0.01, Dunnett's test. B) Serum-starved stromal cells were stimulated with 1 piM progesterone and 50 ng/ml bFGF (broken line) or provided medium with vehicle alone (solid line), and the absorbance was measured using the MTT assay at 0, 24, and 48 h. The time 0 was determined from the basal number of cells SEM of one at the time of mitogen addition. Data represent the mean experiment in triplicate.

that in the control conditions. Taken together, these results show that the MTT assay was measuring proliferation and not the rescue of dying cells.

Proliferation Is Dose Dependent and Saturable To determine whether the proliferative response was dependent upon the concentration of bFGF and/or hormone in the culture medium, serum-starved stromal cells were treated with 50 ng/ml bFGF and a range of progesterone concentrations (0.1-10 M). Alternatively, cultures contained 1 M progesterone and various concentrations of bFGF (range 5-100 ng/ml). Proliferation was measured 48 h after addition of the test substances through use of the Cell Titer 96 assay (Fig. 4). Stromal cell proliferation was stimulated significantly (p < 0.01) over that in the control cultures at the three concentrations of progesterone tested. Furthermore, significant differences in the proliferative response among the three concentrations of progesterone were obtained. The number of cells between 0.1 pM and 1 F.M was significantly different (p < 0.05), and increased further (p < 0.01) between 0.1 F.M and 10 p.M progester-

PROGESTERONE-DEPENDENT STROMAL CELL PROLIFERATION 1.00 f-

1.00 te)

OO.8

8 c

,;

(1)~~~ (9

(a)

(c)

*

0.60 0

0.40

OAO40

0.00

(b)

+

0.0o

020

(d)

0.80

(hi -1

(d)

,

1337

I

I 'oz"4 V 1 ~

1l··\ sf4

,

0.20

0.00 l

FIG. 4. Control of stromal cell proliferation by bFGF and progesterone is dose dependent and saturable. Serum-starved uterine stromal cells (UII, passage 26) were cultured in phenol red-free, serum-free medium as described in the text. Cultures contained medium with 50 ng/ml bFGF and three concentrations of progesterone (P)(range 0.1-10 p.M) or contained 1 M progesterone and three concentration of bFGF (range 5-100 ng/ml). Control cultures (basal) contained ethanol vehicle. Proliferation was measured 48 h after addition of test substances using the Cell Titer 96 assay. Data represent the means SEM of two independent assays in triplicate. **p < 0.01 compared with the control cultures. (a)vs. (b) and (e)vs. (f), p < 0.05; (a)vs. (c), (d)vs. (e), (d)vs. (b), and (d)vs. (f), p < 0.01, StudentNewman-Keuls test.

one. Cells treated with 5 ng/ml bFGF did not proliferate significantly (p > 0.05) over those in the control cultures, while a concentration of 25 ng/ml of the growth factor effectively stimulated mitosis (Fig. 4). No significant increase (p > 0.05) occurred with addition of bFGF at concentrations greater than 25 ng/ml when the concentration of progesterone was 1 pxM (compare b, e, and f). Growth Factor Stimulation of Proliferation Is Progesterone Dependent A number of growth factors have been implicated in the control of uterine cell proliferation [16]. To investigate whether growth factors besides bFGF stimulated progesterone-dependent stromal cell proliferation, serum-starved stromal cells were treated with a panel of growth factors (Fig. 5). Cultures contained 1 pM progesterone and the amounts of growth factor indicated in the figure legend. Proliferation was measured 48 h after addition of the test substances using the Cell Titer 96 assay. Of the panel of growth factors tested, EGF, TGFat, and bFGF stimulated stromal cell division significantly (p < 0.01) over that in the control, progesterone, PDGF, and IGF-I cultures. Cell proliferation in response to progesterone alone, or to progesterone with added IGF-I and PDGF, was not significantly greater (p > 0.05) than that in the control cultures. We tested a range (0.25-6.25 ng/ml) of PDGF and found that at the higher dose (6.25 ng/ml), proliferation was stimulated significantly (p < 0.01) over that in the control cultures (data not shown). However, the response to this higher amount of PDGF was significantly less (p < 0.01) than the absorbance obtained with bFGF, TGFa, and EGF (data not shown). Growth factors tested in the absence of progesterone did not stimulate proliferation over that in the control cultures (Fig. 3A and data not shown). Furthermore, addition of two growth factors (bFGF and TGFa) to the stromal cell cultures in the absence of progesterone did not stimulate proliferation significantly (p > 0.05) compared to

I

I

Iif

q

o

2~o

/P

FIG. 5. Growth factors in combination with progesterone stimulate stromal cell proliferation. Serum-starved uterine stromal cells (UII, passages 18-19) were cultured in phenol red-free, serum-free medium containing 1 M progesterone (P). Control cultures contained medium with ethanol vehicle. Growth factors were added at the following concentrations: PDGF, 1.25 ng/ml; IGF-I, 100 ng/ml; EGF, 50 ng/ml; TGFoa, 10 ng/ml; bFGF, 50 ng/ml. Proliferation was measured 48 h after addition of test substances using the Cell Titer 96 assay. Data represent the meand - SEM of two independent experiments in triplicate. ** p < 0.01 compared with the control cultures. (a)vs. (b), (a)vs. (c), and (a)vs. (d), p < 0.01, StudentNewman-Keuls test.

TGFa + progesterone or bFGF + progesterone (data not shown). Hormonal Control of Stromal Cell Proliferation Is Progesterone Specific To test the specificity of progesterone action in controlling uterine stromal cell proliferation, serum-starved stromal cells were treated with progesterone, 1713-estradiol, dexamethasone, and 5a-dihydrotestosterone in addition to bFGF (Fig. 6). Cells were cultured with 50 ng/ml of bFGF and the concentration of steroid hormone indicated in the legend to Figure 6. Progesterone was the only hormone that 0.80 (d)

0.64

(c)

(a)

O() T

I

0.48

2 I

0.32 0.16 0.00

bFGF

bFGF+E

bFOF+D

bFGF+T

bFGF+P

FIG. 6. The hormonal control of stromal cell proliferation is progesterone specific. Serum-starved uterine stromal cells (UII, passage 19) were cultured in phenol red-free, serum-free medium. Cultures contained 50 ng/ml bFGF and ethanol vehicle or 50 ng/ml bFGF and the indicated steroid hormones. The concentrations of hormones added were estradiol171 (E), 10 nM; dexamethasone (D), 10 nM; 5-dihydrotestosterone (T), 10 nM; progesterone (P), 1 p.M. Proliferation was measured 48 h after addition of test substances using the Cell Titer 96 assay. Data represent the mean SEMs of two independent experiments in triplicate. **p < 0.01 compared with bFGF. (a)vs. (d) and (b)vs. (d), p < 0.01; (c) vs. (d), p < 0.05, Student-Newman-Keuls test.

1338

PIVA ET AL. stimulated proliferation significantly (p < 0.01) over that in the control cultures containing bFGF alone. Progesterone stimulated stromal cell proliferation significantly (p < 0.01) compared with that in cultures containing estrogen and dexamethasone. The difference in cell number between cultures containing 5ot-dihydrotestosterone and progesterone was also significant (p < 0.05). Addition of these hormones without bFGF in the culture medium did not stimulate stromal cell proliferation (data not shown). These results confirm progesterone-dependent control of stromal cell proliferation. Proliferating Stromal Cells Express Progesterone and FGF Receptors

FIG. 7. Immunoblots reveal expression of both forms of progesterone receptors and one isoform of FGFR1 in proliferating uterine stromal cells. A)Total proteins were isolated from the rat uterus at Day 6 of pregnancy and from proliferating uterine stromal cells (UII, passage 19) cultured in serum-free medium containing bFGF (25 ng/ml) and progesterone (1 p.M). Bacterial E. coli XL1 -Blue proteins were prepared as detailed in Materials and Methods. Samples (100 pIg each) were fractionated by SDS-PAGE and transferred to a nitrocellulose membrane. Immunoreactive proteins were visualized using a monoclonal progesterone antibody and the indirect peroxidase detection method. Two immunoreactive proteins were evident in rat uterine proteins from Day 6 pregnant animals (lane 3) and in proteins from proliferating uterine stromal cells (lane 2). The molecular masses (shown by arrows) of these proteins were similar to those reported for the A and B forms of progesterone receptor. No immunoreactivity was detected in the bacterial proteins (lane 1, negative control). Arrowheads indicate position of molecular weight markers of 205, 121, 86, 50.7, and 33.6 (x 103), respectively. B) Serum-starved uterine stromal cells (UII,

passage 19) were cultured in phenol red-free, serum-free medium containing 25 ng/ml bFGF and 1 M progesterone. Membrane proteins were isolated 48 h after addition of proliferative agents. Membrane proteins from rat liver (negative control) and stromal cells were prepared as described in Materials and Methods. The samples (25 Ig) were fractionated by SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was reacted with an FGFR1 antibody that recognizes FGFR1 and FGFR1 isoforms, and the immunoreactive receptor was visualized using an indirect peroxidase detection method. A protein with an apparent molecular mass of approximately 175 kDa was evident in the stromal cell

Biological response to progesterone and bFGF should occur through interaction with their high-affinity receptors [37, 38]. To determine whether these receptors were present in proliferating uterine stromal cells, Western immunoblots were employed that used antibodies whose ability to recognize progesterone and FGF receptors in rat uterus had been previously verified [15, 36]. Serum-starved uterine stromal cells were treated with 1 LM progesterone and 25 ng/ml bFGF for 48 h. These proliferating cells expressed two forms of the progesterone receptor (Fig. 7A). The A form of the receptor migrated with an apparent molecular mass of 120 kDa in total rat uterine proteins (lane 3) and in proliferating uterine stromal cells (lane 2). The B form of the receptor from the cultured cells (lane 2), however, migrated at a slightly smaller molecular mass than the 78-kDa protein detected in uterine tissue (lane 3, and ref. [15]). This is probably due to the large volume of stromal cell protein sample loaded on the gel. No immunoreactivity was seen with the use of equivalent amounts of E. coli protein extracts (lane 1). Immunoblot analysis of stromal cell membrane proteins revealed an immunoreactive protein at 175 kDa (Fig. 7B, lane 2). The molecular mass of this protein was consistent with FGFRla, an isoform of the receptor that contains three immunoglobulin-like loops in the external domain. The 120-kDa isoform that we identified previously [36] at comparable amounts in rat uterine membranes isolated from pregnant animals at Days 4 and 6 was not evident in these pure stromal cell membrane preparations. No immunoreactivity was observed in liver membranes at equal quantities of membrane proteins (lane 1), consistent with previous findings [36]. Inhibitors of Receptor Function Abolish the Proliferative Response Previous work in our laboratory [15] showed that administration of the progesterone receptor antagonist RU 486 inhibited stromal cell proliferation in pregnant rats. Suramin is an inhibitor of the interaction of bFGF with its tyrosine kinase receptors [39-42]. We reasoned that the proliferative response to ligands should be eliminated by the blocking of receptor function. Serum-starved stromal cells were treated with phenol red-free, serum-free medium containing proliferative agents (1 pIM progesterone, 25 ng/ml bFGF) and various concentrations of RU 486 (Fig. 8A) or

membrane proteins (shown by arrow). Lane 1, rat liver membrane proteins (negative control); lane 2, cultured proliferating uterine stromal cell membrane proteins. Arrowheads indicate position of molecular weight markers of 205, 121, 86, 50.7, and 33.6 (x 103), respectively.

PROGESTERONE-DEPENDENT STROMAL CELL PROLIFERATION

A

A

1.oo

1.00 F

1.00 0 a

0s0O O 0

0.60

I

1339

I

C

0.40

0.60

(a)

F

(b)

-T

0.40AO

0.20

(C

0.20

(d) 0.00

0

B

]

0.1

1

5 10 RU 486 (M)

50

l

0.00

Control

100

40

200

-

500

Control

Suramln (pM)

125

F

100

I-

75

I-

60

F

e-

90 _

25

F

E

80 I'

B

(a) -5-

8

0

150 120

r

[

T

T

n

_

bFGF Progeterone RU 486

+

l M

+

I pm 10 M*

+

10

+

+

-

10M low 10

FIG. 8. Inhibition of progesterone receptor function blocks stromal cell proliferation. A) Serum-starved uterine stromal cells (Ull, passages 20 and 21) were cultured in phenol red-free, serum-free medium containing 25 ng/ml bFGF and 1 JAM progesterone to stimulate proliferation. Control cultures (basal) contained medium and vehicle alone. Proliferation was measured 48 h after the addition of mitogens and the receptor antagonist using the Cell Titer 96 assay. Data represent the means SEM of two independent experiments in triplicate. (a)vs. (b)and (a)vs. (c), p < 0.01, Student-Newman-Keuls test. B) Serum-starved uterine stromal cells (Ull, passages 20 and 21) were treated in phenol red-free, serum-free medium containing RU 486 and progesterone. Data represent the means SEM of two independent experiments. (a)p < 0.01; (b) p > 0.05, StudentNewman-Keuls test.

suramin (Fig. 9A). Positive control cultures contained hormone and growth factor alone. Negative control cultures contained medium and vehicle alone. Addition of the progesterone receptor antagonist RU 486 at a 10 M concentration reduced significantly (p < 0.01) the proliferative response to 64.7% compared with that in the positive control cultures (Fig. 8A). Although the number of cells was reduced at 5 F.M of RU 486, the decline was not significantly different (p > 0.05) from those obtained without RU 486. To determine whether the effects of RU 486 were cytostatic or cytotoxic, cells were treated with 10 pIM RU 486 and 10 ptM progesterone for 48 h (Fig. 8B). At equimolar concentrations of hormone and RU 486, the progesterone receptor antagonism was overcome (p > 0.05), suggesting that RU 486 exerts cytostatic rather than cytotoxic action in vitro. By MTT assay, addition of suramin to stimulated stromal cells cultures inhibited proliferation significantly (p < 0.01) at suramin concentrations of 100, 200, and 500 p.M (Fig. 9A). At the lower doses of suramin (0-40 .M), cell proliferation was not inhibited (p > 0.05) compared with that

in the positive control cultures. Since suramin affects respiration and energy balance [43], the number of cells in

30 A V

L.

Surmlin (pM) bFGF

m

. 10o + +

- 200 + +

- 500 + +

- 500

FIG. 9. Suramin has multiple effects on stromal cell proliferation. A) Serum-starved uterine stromal cells (Ull, passages 20 and 21) were cultured in phenol red-free, serum-free medium containing 25 ng/ml bFGF and 1 M progesterone. Five concentrations of suramin were used to inhibit FGFR1 activation. Control cultures (basal) contained medium alone. Proliferation was measured 48 h after the addition of the test substances using the Cell Titer 96 assay. Data represent the means SEM of two independent experiments in triplicate. (a) vs. (b), (a) vs. (c), and (a) vs. (d), p < 0.01, Student-Newman-Keuls test. B) Since suramin affects mitochondrial enzymatic activity, the numbers of cells in cultures containing bFGF (25 ng/ml) and progesterone (1 M) were counted. Data represent the means SEM of two independent experiments. Suramin at 500 M concentration inhibited significantly (** p < 0.01) stromal cell proliferation in response to bFGF and progesterone. At 100 FM suramin, cell number was increased significantly (* p < 0.05) over that in the control groups. Cultures containing 500 1iM suramin in the absence of bFGF were not significantly different (p > 0.05) from the bFGF-free control cultures. See text for details.

cultures containing mitogens was counted and compared with the number of cells in cultures containing mitogens and the three concentrations of suramin (Fig. 9B). At the two lower doses of suramin (100, 200 RM) there was no significant inhibition (p > 0.05) of cell number between the two groups (Fig. 9B). In fact, at 100 luM suramin, cell number was increased significantly (p < 0.05) over that in the control group. Thus, suramin reduced the mitochondrial enzymatic activity of uterine stromal cells (as measured by the MTT assay) rather than exerting cytostatic action at concentrations of 100 and 200 M. However, suramin at 500 p.M inhibited significantly (p < 0.01) stromal cell proliferation in response to bFGF and progesterone (Fig. 9B). Since cell viability as measured by the trypan blue exclusion test [42] was greater than 95% in these counting ex-

1340

PIVA ET AL.

periments, the results indicate that suramin at 500 F.M exerts cytostatic effects on uterine stromal cells. This proposal is substantiated by comparison between a control experiment without bFGF and the 500 p.M dose of suramin in the same conditions (Fig. 9B). No significant difference (p > 0.05) in cell number between the two groups was obtained, suggesting that suramin effects are cytostatic rather than cytotoxic. DISCUSSION Hormonal control of cell division is a cornerstone of reproductive biology because cell cycle control mechanisms are influenced by sex steroids (reviewed in [7]). Results from the present investigation demonstrate that progesterone-dependent stromal cell proliferation is maintained in vitro and confirm earlier reports from rat [21] and human [20] studies. Moreover, the doubling time of these cells cultured in progesterone and bFGF was consistent with the time frame reported previously for experiments in which the cells were cultured in medium supplemented with 10% FBS ([21], Fig. 3B). Most cultured mammalian cells require growth factors alone to regulate the ordered progression through the cell cycle [6]. Progesterone-dependent cell cycle controls may have evolved in uterine stromal cells because of the important chronological relationship between the blastocyst and the uterus. Thus, cell cycle regulation by progesterone ensures that decidualization occurs temporally with ovulation in both humans and rodents. This synchrony connects the hormone of pregnancy (progesterone) with preparation of the endometrium for blastocyst implantation. Development of stable uterine stromal cell cultures with cell doubling times (Fig. 3B) in frame with the eukaryotic cell cycle now provides a unique and important opportunity to increase understanding about the control of stromal cell proliferation and decidual growth by progesterone. Under the conditions used in this study, and as reported previously [44], IGF-I does not stimulate rat uterine cell proliferation. These results must be interpreted cautiously, however, because human stromal cells that are unresponsive when cultured under similar culture conditions [20] reportedly can proliferate in response to IGF-I [45, 46]. In addition to bFGF, other growth factors such as EGF and TGFa are mitogenic agents for uterine stromal cells cultured in serum-free medium containing progesterone. The lack of response in rat uterine stromal cells in medium containing 10% charcoal-stripped fetal calf serum (Fig. 2) compared to serum-free medium (Fig. 5) is due to the presence of these growth factors in serum. While progesterone is necessary, it is not sufficient to stimulate cell proliferation in cultured rat or human stromal cells. Previous work in our laboratory has shown that bFGF and FGF receptors (FGFRI) are present both spatially and temporally in proliferating stromal cells in the rat uterus at Days 4-6 of pregnancy [15, 36]. FGFR1 immunolocalizes to the perinuclear region only in uterine stromal cells, implicating bFGF function in these cells [36, 47]. More recent data have shown that the distal immunoglobulin-like loop of FGFRla mediates a differential targeting of this isoform to a perinuclear location [48]. The mechanism for nuclear translocation is unknown, but preliminary evidence suggests that the intact receptor is translocated and that it maintains functional tyrosine kinase activity [48]. FGFRIP3 is an isoform of the receptor that lacks the most external immunoglobulin-like domain [38]. Proteins isolated from the uterus of pregnant rats contain both FGFRloe and FGFR1 3

at comparable levels [36]. This variant isoform (FGFR13) of the receptor, however, does not translocate to the nucleus and does not stimulate tyrosine kinase activity [48]. The present results demonstrate that proliferating uterine stromal cells in culture express predominantly FGFRl(o (Fig. 7B). Taken together, these results support further the view that perinuclear localization of FGFRI in uterine stromal cells is due to activation of FGFR1. Addition of suramin at 500 p.M concentration to uterine stromal cell cultures inhibited the growth-stimulating activity of progesterone and bFGF (Fig. 9A). For some growth factors, suramin blocks activity by aggregating the ligand, thereby reducing the amount of growth factor available for interaction with receptor [39, 42, 49]. Stimulation of proliferation by suramin has been reported previously [40, 50, 51] and in our study was strictly dose dependent at 100 pIM concentration. This may be the optimum concentration of suramin-bFGF dimers resulting in FGFR1 activation. Our results suggest that suramin at 500 pIM concentration specifically inhibits bFGF binding to FGFR1 and thereby inhibits stromal cell proliferation. However, suramin has been reported to reduce the testosterone-dependent accumulation of FGFR1 mRNA in mouse mammary cancer cell line cells [52], and we can not exclude the possibility that FGFR1 is down-regulated in suramin-treated stromal cells. Moreover, suramin inhibits DNA polymerase a in lymphoid cells [53], plasminogen activator activity in transformed GM 7373 fetal bovine aortic endothelial cells [42], and cathepsin D expression in breast cancer cell lines [40]. In the breast cancer cells, suramin inhibited both estrogen- and growth factor-stimulated proliferation [40]. Although the mechanisms have not been elucidated, overexpression of cathepsin D in murine cancer cells decreases the cellular secretion of growth inhibitors [54], suggesting an indirect repression. Control of cell proliferation by progesterone is an area of biology that has not been extensively studied from a mechanistic viewpoint. Evidence is accumulating to suggest that cross-talk between growth factors and steroid hormones generates diversity in hormone signal transduction pathways [55-57]. Dopamine activates progesterone receptors in the absence of progesterone, probably by a cAMPdependent phosphorylation mechanism [55]. Epidermal growth factor and TGFot stimulate estrogen-dependent transcriptional activity in a dose-dependent manner utilizing a consensus estrogen response element [56]. Transcriptional activation of the estrogen-responsive vitellogenin reporter gene requires both the EGF and estrogen receptors [56]. Speculation regarding the nature of cross-talk between growth factors and hormones has been advanced. Since growth factor binding to high-affinity tyrosine kinase receptors results in the phosphorylation of target substrates (reviewed in [58]), one or more of these substrates may act on steroid receptors, stimulating hormone-dependent gene transcription [17]. Alternatively, phosphorylated substrates could serve as cofactors providing differential combinations at the ligand-binding domain of the receptor [57] or at the hormone response element (reviewed in [59]). Growth factors and hormones stimulate the transcription of other regulatory proteins such as the AP-1 activators. In addition, a consensus AP-1 site has been reported in the bFGF gene [60]. Thus, overlap and convergence of several signaling pathways may be required for full tissue response to hormone action [56]. Stromal cells proliferate and differentiate into decidual cells with characteristic morphology and polyploid nuclei [61, 62]. Decidual cells develop gap junctions between ad-

PROGESTERONE-DEPENDENT STROMAL CELL PROLIFERATION jacent cells [63] and express characteristic markers of decidualization [64-67] including the intermediate filament proteins desmin and vimentin [64, 65]. Our immunocytochemical analyses show that cultured stromal cells express vimentin and desmin, in agreement with earlier reports [64, 65]. In contrast, Cohen et al. [21] reported no immunoreactivity of isolated rat stromal cells with desmin. The reason for this difference is not clear, but the difference may be due to the additional separation of stromal cells using Percoll gradients [21], which we have not employed. An essential basis for studying cellular differentiation, including decidualization, is identification of markers that are expressed at defined stages of the differentiation process. When uterine cells are removed from their normal complex environment and cultured, alterations in specific gene expression occur [68]. These observations suggest that localized inhibitory mechanisms must exist in situ that are not functional when decidual cells are isolated from endometrial components including the epithelial cells, extracellular matrix, and basement membrane [68]. In view of the interesting results from that elegant study, it is important to distinguish markers of differentiation from genes that are "derepressed" in cultured cells. We propose that expression of vimentin and desmin in cultured stomal cells does not indicate a differentiated phenotype because these cells can continue to proliferate. A pivotal point for the initiation of a differentiation program is the activation of genes closely linked to molecular mechanisms involved in cell cycle arrest. This assumption is consistent with data from other systems in which an inverse relationship between cell proliferation and differentiation has been established [6, 69]. It is also supported by the recent study of Finn and colleagues [70], who report that uterine sensitivity is influenced by the phase of the stromal cell cycle. The decidual-sensitive period correlates temporally with later stages of the cell cycle (S or G2) where an appropriate stimulus can direct the cells toward differentiation rather than proliferation [70]. Development of stable, progesterone-dependent uterine stromal cells in culture will now facilitate identification of progesterone control points for cell cycle progression. While nontransformed lines of rat uterine cells exist [71], these cells contain a temperaturesensitive mutant of the SV40 large T antigen. Rat uterine stromal cells provide a system free of foreign DNA as an important model to study cell cycle control mechanisms.

6. 7. 8. 9.

10.

11. 12. 13.

14. 15.

16. 17.

18. 19.

20. 21. 22.

23.

ACKNOWLEDGMENTS 24. The authors thank Jim Swafford (UMKC) for computer graphic assistance. We are grateful to James Rusche (Glycan Pharmaceuticals) for the suramin and to Romana Nowak (Harvard University) for the RU 486. We acknowledge Lindsey Hutt-Fletcher and Susan Turk for their advice on cell culture and for providing the epithelial cell line. The authors thank N. Weigel and D. Edwards for the progesterone monoclonal antibody. We thank the reviewers for their excellent comments and suggestions.

25.

26.

REFERENCES 1. Chang MC. Development and fate of transferred rabbit ova or blastocyst in relation to the ovulation time of recipients. J Exp Zool 1950; 114:197-225. 2. Psychoyos A. Uterine receptivity for nidation. Ann NY Acad Sci 1986; 476:36-42. 3. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950; 1:3-25. 4. Martin L, Finn CA. Hormonal regulation of cell division in epithelial and connective tissues of the mouse uterus. J Endocrinol 1968; 41: 363-371. 5. Tachi C, Tachi S, Lindner HR. Modification by progesterone of estra-

27. 28.

29. 30.

1341

diol-induced cell proliferation, RNA synthesis and estradiol distribution in the rat uterus. J Reprod Fertil 1972; 31:59-76. Murray A, Hunt T. The Cell Cycle: An Introduction, Ist ed. New York: Oxford University Press; 1993: 89-116. Clarke CL, Sutherland RL. Progestin regulation of cellular proliferation. Endocr Rev 1990; 11:266-301. Martin L, Finn CA, Trinder G. DNA synthesis in the endometrium of the progesterone-treated mice. J Endocrinol 1973; 41:363-371. Weitlauf HM. Biology of implantation. In: Knobil E, Neill J, Ewing JJ, Greenwald GS, Markert CL, Pfaff DW (eds.), The Physiology of Reproduction. New York: Raven Press Ltd; 1988: 231-262. Heap RB, Rider V, Wooding FBP, Flint APE Molecular and cellular signaling and embryo survival. In: Sreenan JM, Diskin MG (eds.), Embryonic Mortality in Farm Animals. Dordretch: Martinus Nijhoff Publishers; 1986: 46-73. Finn CA. Endocrine control of endometrial sensitivity during the induction of the decidual cell reaction in the mouse. J Endocrinol 1966; 36:239-248. Galassi L. Autoradiographic study of the decidual cell reaction in the rat. Dev Biol 1968; 17:75-84. Rider V, Wang M-Y, Finn CA, Heap RB, Feinstein A. Anti-fertility effect of passive immunization against progesterone is influenced by genotype. J Endocrinol 1986; 108:117-121. Cullingford TE, Pollard JW. RU 486 completely inhibits the action of progesterone on cell proliferation in the mouse uterus. J Reprod Fertil 1988; 83:909-914. Rider V, Psychoyos A. Inhibition of progesterone receptor function results in loss of basic fibroblast growth factor expression and stromal cell proliferation during uterine remodelling in the pregnant rat. J Endocrinol 1994; 140:239-249. Guidice LC. Growth factors and growth modulators in human uterine endometrium: their potential relevance to reproductive medicine. Fertil Steril 1994; 61:1-17. Ignar-Trowbridge DM, Nelson KG, Bidwell MC, Curtis SW, Washburn TF, McLachlan JA, Korach KS. Coupling of dual signaling pathways: epidermal growth factor action involves the estrogen receptor. Proc Natl Acad Sci USA 1992; 89:4658-4662. Finn CA, Martin L. Patterns of cell division in the mouse uterus during early pregnancy. J Endocrinol 1967; 39:593-597. Bigsby RM, Cunha GR. Effect of progestins and glucocorticoids on deoxyribonucleic acid synthesis in the uterus of the neonatal mice. Endocrinology 1985; 117:2520-2526. Irwin JC, Utian WH, Eckert RL. Sex steroids and growth factors differentially regulate the growth and differentiation of cultured human endometrial stromal cells. Endocrinology 1991; 129:2385-2392. Cohen H, Pageaux JF, Melinand C, Fayard JM, Laugier C. Normal rat uterine stromal cells in continuous culture: characterization and progestin regulation of growth. Eur J Cell Biol 1993; 61:116-125. Chegini N, Rossi MJ, Masterson BJ. Platelet-derived growth factor (PDGF), epidermal growth factor (EGF), and EGF and PDGF 3-receptors in human endometrial tissue: localization and in vitro action. Endocrinology 1992; 130:2373-2385. Hammond MG, Oh S-T, Anners J, Surrey ES, Halme J. The effect of growth factors on the proliferation of human endometrial stromal cells in culture. Am J Obstet Gynecol 1993; 168:1131-1138. Mellor SJ, Thomas EJ. The actions of estradiol and epidermal growth factor in endometrial and endometriotic stroma in vitro. Fertil Steril 1994; 62:507-513. Tang X-M, Rossi MJ, Masterson BJ, Chegini N. Insulin-like growth factor I (IGF-I), IGF-I receptors, and IGF binding proteins 1-4 in human uterine tissue: tissue localization and IGF-I action in endometrial stromal and myometrial smooth muscle cells in vitro. Biol Reprod 1994; 50:1113-1125. Massague J, Roberts JM. Cell multiplication. Cell cycle 1995: constructing cell physiology with molecular building blocks. Curr Opin Cell Biol 1995; 7:769-772. McCormack SA, Glasser SR. Differential response of individual uterine cell types from immature rats treated with estradiol. Endocrinology 1980; 106:1634-1649. Mossman T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983; 65:55-63. Tada H, Shibo O, Kuroshima K, Koyama M, Tsukamoto K. An improved colorimetric assay for interleukin-2. J Immunol Methods 1986; 93:157-165. Kitamura T, Tange T, Teresawa T, Chiba S, Kuwaki T, Miyagawa K, Piao YE Miyazono K, Urabe A, Takaku E Establishment and char-

1342

31.

32.

33.

34. 35. 36.

37. 38. 39. 40. 41.

42.

43.

44.

45.

46. 47.

48.

49.

50.

PIVA ET AL.

acterization of a unique human cell line that proliferates dependently on GM-CSF, IL-3, or erythropoietin. J Cell Physiol 1989; 140:323334. Lockwood CJ, Krikun G, Papp C, Aiger S, Nemerson Y, Schatz E Biological mechanisms underlying RU 486 clinical effects: inhibition of endometrial stromal cell tissue factor content. J Clin Endocrinol & Metab 1994; 79:786-790. Narukawa S, Kanzaki H, Inoue T, Imai K, Higuchi T, Hatayama H, Kariya M, Mori T Androgens induce prolactin production by human endometrial stromal cells in vitro. J Clin Endocrinol & Metab 1994; 78:165-168. Steel RGD, Torrie JH. Analysis of variance I: the one-way classification. In: Principles and Procedures of Statistics, Ist ed. New York: McGraw-Hill Book Company; 1960: 99-131. Carlone DL, Rider V. Embryonic modulation of basic fibroblast growth factor in the rat uterus. Biol Reprod 1993; 49:653-665. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227:680-685. Rider V, Piva M, Cohen ML, Carlone DL. Alternative splicing and differential targeting of fibroblast growth factor receptor I in the pregnant rat uterus. Endocrinology 1995; 136:3137-3145. Beato M. Gene regulation by steroids hormones. Cell 1989; 56:335344. Johnson DE, Williams LT. Structural and functional diversity in the FGF multigene family. Adv Cancer Res 1993; 60:1-41. Middaugh CR, Mach H, Burke CJ, Volkin DB, Dabora JM, Tsai PK, Bruner MW, Ryan JA, Marfia KE. Nature of the interaction of growth factors with suramin. Biochemistry 1992; 31:9016-9024. Vignon F Prebois C, Rochefort H. Inhibition of breast cancer growth by suramin. J Natl Cancer Inst 1992; 84:38-42. Braddock PS, Hu D-E, Fan T-PD, Stratford IJ, Harris AL, Bicknell R. A structure-activity analysis of antagonism of the growth factor and angiogenic activity of basic fibroblast growth factor by suramin and related polyanions. Br J Cancer 1994; 69:890-898. Takano S, Gately S, Neville ME, Herblin WF, Gross JL, Engelhard H, Perricone M, Eidsvoog K, Brem S. Suramin, an anticancer and angiosuppressive agent, inhibits endothelial cell binding of basic fibroblast growth factor, migration, proliferation, and induction of urokinase-type plasminogen activator. Cancer Res 1994; 54:2654-2660. Rago R, Mitchen J, Cheng A-L, Oberley T, Wilding G. Disruption of cellular energy balance by suramin in intact human prostatic carcinoma cells, a likely antiproliferative mechanism. Cancer Res 1991; 51:6629-6635. Aronica SM, Katzenellenbogen BS. Progesterone receptor regulation in uterine cells: stimulation by estrogen, cyclic adenosine 3',5'-monophosphate, and insulin-like growth factor I and suppression by antiestrogens and protein kinase inhibitors. Endocrinology 1991; 128: 2045-2052. Frost RA, Mazella J, Tseng L. Insulin-like growth factor binding protein-1 inhibits the mitogenic effect of insulin-like growth factors and progestins in human endometrial stromal cells. Biol Reprod 1993; 49: 104-111. Irwin JC, De las Fuentes L, Guidice LC. Growth factors and decidualization in vitro. Ann NY Acad Sci 1994; 734:7-18. Prudovsky I, Savion N, Zhan X, Friesel R, Xu J, Hou J, McKeehan WL, Maciag T. Intact and functional fibroblast growth factor (FGF) receptor-i trafficks near the nucleus in response to FGF-I. J Biol Chem 1994; 269:31720-31724. Prudovsky IA, Savion N, LaVallee TM, Maciag T. The FGF- I dependent intracellular traffick differential of FGFR-1 a and FGFR-113. Mol Biol Cell 1995; 6S:124a. Schrell UMH, Gauer S, Kiesewetter F, Bickel A, Hren J, Adams EF Fahlbusch R. Inhibition of proliferation of human cerebral meningioma cells by suramin: effects on cell growth, cell cycle phases, extracellular growth factors, and PDGF-BB autocrine growth loop. J Neurosurg 1995; 82:600-607. Michel P, Van Velthoven R, Metein M, Gras S, Etievant C, Pasteels

51.

52.

53. 54. 55.

56.

57. 58. 59. 60. 61.

62. 63. 64. 65.

66. 67.

68.

69. 70. 71.

JL, Kiss R. Influence of suramin alone or in combination with DHT and PDGF on the cell proliferation of benign and malignant human prostatic tissues in organ cultures. Anticancer Res 1993; 11:20752078. Rubio GJ, Pinedo HM, Virizuela J, van Ark-Otte J, Giaccone G. Effects of suramin on human lung cancer cell lines. Eur J Cancer 1995; 31A:244-251. Kasayama S, Saito H, Kouhara H, Sumitani S, Sato B. Suramin interrupts androgen-inducible autocrine loop involving heparin binding growth factor in mouse mammary cancer (Shionogi Carcinoma 115) cells. J Cell Physiol 1992; 154:254-261. Spigelman Z, Dowers A, Kennedy S, Disorbo D, O'Brien M, Barr R, McCaffrey R. Antiproliferative effects of suramin on lymphoid cells. Cancer Res 1987; 47:4694-4698. Liaudet E, Derocq D, Rochefort H, Garcia M. Transfected cathepsin D stimulates high density cancer cell growth by inactivating secreted growth inhibitors. Cell Growth Differ 1995; 6:1045-1052. Power RF, Lydon JP, Conneely OM, O'Malley BW. Dopamine activation of an orphan of the steroid receptor superfamily. Science 1991; 252:1546-1548. Ignar-Trowbridge DM, Teng CT, Ross KA, Parker MG, Korach KS, Mclachlan JA. Peptide growth factors elicit estrogen receptor-dependent transcriptional activation of an estrogen-responsive element. Mol Endocrinol 1993; 7:992-998. Forman BM, Umesono K, Chen J, Evans RM. Unique response pathways are established by allosteric interactions among nuclear hormone receptors. Cell 1995; 81:541-550. Schlessinger J, Ullrich A. Growth factor signaling by receptor tyrosine kinases. Neuron 1992; 9:383-391. Lamb P, McKnight SL. Diversity and specificity in transcriptional regulation: the benefits of heterotypic dimerization. Trends Biochem Sci 1991; 16:417-422. Shibata F, Baird A, Florkiewicz RZ. Functional characterization of the human basic fibroblast growth factor gene promoter. Growth Factors 1991; 4:277-287. Welsh AO, Enders AC. Light and electronmicroscopic examination of the mature decidual cell of the rat with emphasis on the antimesometrial decidua and its degeneration. Am J Anat 1985; 172:1-29. Leroy F, Bogaert C, Van Hoeck J, Delcroix C. Cytophotometric and autoradiographic evaluation of cell kinetics in decidual growth in rats. J Reprod Fertil 1974; 38:441-449. Finn CA, Lawn AM. Specialized junctions between decidual cells in the uterus of the pregnant mouse. J Ultrastruct Res 1967; 20:321-327. Glasser SR, Julian J. Intermediate filament protein as a marker of uterine stromal cell decidualization. Biol Reprod 1986; 35:463-474. Mani SK, Julian J, Lampelo S, Glasser SR. Initiation and maintenance of in vitro decidualization are independent of hormonal sensitization in vivo. Biol Reprod 1992; 47:785-799. Glasser SK, Lampelo S, Munir IM, Julian J. Expression of desmin, laminin and fibronectin during in situ differentiation (decidualization) of rat uterine stromal cells. Differentiation 1987; 35:463-474. Hochner-Celnikier D, Ron M, Eldor A, Segal S, Paltu Z, Fuks Z, Vlodavsky I. Growth characteristics of human first trimester decidual cells cultured in serum-free medium: production of prolactin, prostaglandins and fibronectin. Biol Reprod 1984; 31:827-836. Gu Y, Gibori G. Isolation, culture and characterization of the two cell subpopulations forming the rat decidua: differential gene expression for activin, follistatin and decidual-related prolactin protein. Endocrinology 1995; 136:2451-2458. Derynck R, Wagner E. Cell differentiation. Curr Opin Cell Biol 1995; 7:843-844. Finn CA, Pope M, Milligan SR. Control of uterine stromal mitosis in relation to uterine sensitivity and decidualization in mice. J Reprod Fertil 1995; 103:153-158. Arslan A, Guillermina A, Aingg HH. Characterization and co-culture of a novel nontransformed cell lines derived from rat endometrial epithelium and stroma. In Vitro Cell Dev Biol 1995; 31:140-148.