0013-7227/92/1304-2373$03.00/O Endocrinology Copyright 0 1992 by The Endocrine

Vol. 130, No. 4 Printed in U.S.A.

Society

Platelet-Derived Growth Factor (PDGF), Epidermal Growth Factor (EGF), and EGF and PDGF ,&Receptors in Human Endometrial Tissue: Localization and in Vitro Action* NASSER

CHEGINI,

MICHAEL

J. ROSSI,

Department of Obstetrics and Gynecology, University Gainesville, Florida 32610

AND of

BYRON

J. MASTERSON

Florida College of Medicine,

Human endometrial tissue and primary stromal ABSTRACT. cell culture contain immunoreactive epidermal growth factor (EGF), platelet-derived growth factor (PDGF)-AB as well as EGF and PDGF-6 receptors. The immunostaining for EGF, EGF receptor, and PDGF j3-receptor were associated with endometrial luminal and glandular epithelial and stromal cells, whereas only the stromal cells contain immunoreactive PDGFAB. The immunostaining intensity of EGF, EGF receptor, and PDGF-AB was similar in both phases of the menstrual cycle, whereas, PDGF-6 receptor immunostaining was highest in proliferative phase and considerably reduced, particularly in luminal and glandular epithelial cells in the secretory phase. In addition primary stromal cell cultures express EGF, PDGF-AB, and contain EGF and PDGF-fi receptors, and very low levels of PDGF-(U receptor. SH-Thymidine incorporation indicate that after 48 h of incubation in serum-free medium approximately 7580% of stromal cells are quiescent. Incubation of quiescent stromal cells with 10% fetal bovine serum stimulate 3H-thymidine incorporation in a time-dependent manner reaching maximal after 3048 h, with a doubling time of 38.2 h. EGF (1.5-15 rig/ml) stimulates ‘H-thymidine incorporation by quiescent stromal

T

cells (P < 0.001). This effect was significantly reduced at concentrations above 15 rig/ml (P < 0.005). PDGF-AI3 (3-10 ng/ ml) and PDGF-BB (0.5-10 rig/ml) also stimulate ‘H-thymidine incorporation in quiescent stromal cells compared to controls (P < 0.005). The action of EGF (15 rig/ml) and PDGF-AB (10 rig/ml) was time dependent, reaching maximal after 36 and 48 h of incubation (P C 0.002). Addition of PDGF-AB (10 rig/ml) to EGF (15 rig/ml) significantly enhanced the action of EGF or PDGF-AI3 used individually (P < 0.001). 17fl-estradiol or progesterone at 1 PM did not stimulate 3H-thymidine incorporation, although they were stimulatory in combination (P < O.OOl), they did not alter the action of EGF or PDGF when added in combination. These observations provide further evidence that human endometrial tissue contains specific immunoreactive EGF receptors. It also demonstrates the presence of immunoreactive EGF, PDGF-AB, and PDGF-j3 receptors in endometrial tissue as well as stromal cells in primary culture. Both EGF and PDGF are mitogenic for endometrial stromal cells, suggesting an autocrine/ paracrine role in modulation of endometrial cell growth and differentiation. (Endocrinology 130: 2373-2385,1992)

Growth factors, such as epidermal growth factor (EGF) and platelet-derived growth factor (PDGF), play an essential role in regulating cell growth and differentiation through their interactions with specific cell surface receptors in their target tissues (6-8). The expression of EGF messenger RNA (mRNA) and presence of immunoreactive EGF and transforming growth factor (Y (TGF-cr) protein have previously been reported in rodent uteri (9-11). Recent reports have demonstrated that human uterine tissue. expresses EGF (12) and TGF-a (12, 13) mRNA and contains immunoreactive TGF-a (14). The presence and cellular distribution of EGF receptor in human and rodent endometrial tissue have also been reported, with the highest EGF receptor concentration in stromal cells (15-22). Recently it has been shown that human uterine tissue expresses PDGF-BB mRNA and very little or no detectable levels

HE cyclical histological and biochemical alterations of the human endometrium during the menstrual cycle have been well documented and correlated with the periodic changes in serum concentrations of sex steroids, estrogen, and progesterone (P4) (l-3). The endometrial response to estrogen and Pq is complex involving many biochemical parameters, which ultimately lead to proliferation and differentiation of all endometrial cell types (l-3). It has been suggested that one of the early events mediating the sex steroid action is the local production of growth factors, protooncogenes and their receptors, acting in an autocrine and/or paracrine manner (4-6). Received October 23,199l. Address correspondence and requests for reprints to: Dr. Nasser Chegini, Department of OB/GYN, University of Florida, P.O. Box 100294, Gainesville, Florida 32610. *Funded in part by University of Florida, Division of Sponsored Research. 2373

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

2374

GROWTH FACTOR REGULATION

of PDGF-AA (23). PDGF occurs in three different isoforms, AA, BB, and AB which bind to two different classes of receptors with different affinities designated as LY-and ,&types (24). Porcine uterus is a major source of PDGF-P receptors, and in situ hybridization has shown the presence of mRNA for the P-receptor preferentially associated with endometrial stromal cells (25, 26). Human uterus also contains PDGF receptors (27), however the cellular distribution and relative contribution of the (Y-and P-receptor types is not known. In rodent, the expression of EGF and EGF receptor in the uterus have been shown to be regulated by 17pestradiol (E2) in ovariectomized animals (9, 10, 21). Recent reports have demonstrated that local administration of EGF could replace the estrogenic effect in rodent uterine tissue (28,29). In human however, no correlation in relation to EGF (12), PDGF-BB mRNA expression (23), and EGF receptor content (15-17) either with phases of the menstrual cycle or pathological conditions have been reported. In this study, we report the presence of EGF, PDGFAB, as well as EGF and PDGF @-receptors in human endometrial tissues and primary endometrial stromal cell cultures. In addition, we present data concerning the mitogenic effect of EGF, PDGF-AB, and PDGF-BB, either alone or in combination with E, and P4, in endometrial stromal cells. Materials

and Methods

Dulbecco’s modified Eagle’s medium (DMEM), Ham’s F-12, EDTA, trypsin-EDTA solution, antibiotic antimycotic solution, and phenol-red free medium were purchased from GIBCO (Grand Island, NY). Fetal bovine serum, Ea, and P4 were from Sigma Chemical Company (St. Louis, MO). Collagenase was obtained from Worthington Biochemical (Freehold, NJ). Culture grade EGF, natural human PDGF-AB, and recombinant human PDGF-BB were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Polyclonal rabbit anti-human recombinant EGF (Biomedical Technologies, Inc. Stoughton, MA), polyclonal anti-human PDGF-AB (Upstate Biotechnology, Inc.), monoclonal antibodies to human PDGF (Y-and PDGF preceptors (Genzyme Corporation, Cambridge, MN), and to porcine PDGF P-receptors (Oncogene Sciences Inc. Manhasset, NY), monoclonal antibodies to human Cytokeratin 19 and human desmin (Dako Corporation, Carpinteria, CA), and monoclonal antibodies to swine and human vimentin were purchased from Dako or Sigma, respectively. Rhodamine (TRITC) conjugated secondary antibodies were purchased from Jackson Laboratories Inc (West Grove, PA). Vectastain ABC Kits were from Vector (Burlington, CA). Methyl-3H-thymidine (83 Ci/mmol) was purchased from Amersham Co. (Arlington Heights, IL). The 75 cm* flasks and the 24 well multiwell plates used in these experiments were purchased from Corning Glass Works (Corning, NY). Eight-well culture slides either glass or plastic were purchased from Nunc Inc. (Naperville, IL). Monoclonal antibodies to the extracellular domain of EGF receptor

CIF ENDOMETRIAL

STROMAL

CELLS

Endo l 1992 Vol1.30. No 4

were gift from Dr. W. A. Dunn, Jr. (Department of Anatomy and Cell Biology, University of Florida). Tissue collection

and endomettinl

stromal cells isolation

Portions of 102 uterine tissues from nonpregnant premenopausal women undergoing hysterectomy for medically indicated reasons (excluding endometrial cancer) were collected at the University of Florida affiliated Shands Hospital. The collection of these tissues for the present study has been approved by the University of Florida Institutional Review Board. Each uterine tissue was examined by the pathologist for histological dating of endometrium. The endometrial histological dating was done according to Noyes et al. (30) and the patient’s last menstrual period. Thirty-five uterine specimens were from proliferative and 67 from secretory phase of the menstrual cycle. The endometrium was removed by scraping the cell layers from myometrium, washed in Ca2+-Mg2+ free PBS containing 1% antibiotic-antimycotic solution, cut into small pieces, and digested in 1% collagenase at 37 C for 30-60 min. The collagenasedigested cells were passed through a metal sieve (Sigma Chemical) to remove the epithelial and undigested tissue fragments. The isolated stromal cells were collected by centrifugation at 465 x g for 5 min, washed several times with DMEM-Ham’s F-12 containing 1% antibiotic-antimycotic solution and cultured. Endometrial

stromal

cell culture

The isolated endometrial stromal cells were grown as monolayers in 75-cm2 flasks at an initial density of 2 x lo6 cells per flask and maintained in DMEM-Ham’s F-12 (50:50, vol/vol) supplemented with 10% fetal bovine serum (FBS) (vol/vol), and 1% antibiotic-antimycotic solution (vol/vol) for 6 days. The media were changed every 3 days. The cultures were maintained at 37 C in a humidified atmosphere of 5% C02-95% air. After the sixth day of incubation the monolayer cells were rinsed twice with Ca2+-Mp free PBS, trypsinized with 0.1% trypsin-5% EDTA, and subcultured at above density. After the third passage, the cells were seeded in 24 well dishes or 8 well Lab-Tek slides at a density of approximately 2.5 X lo4 cells per well in DMEM-Ham’s F-12 supplemented with 10% FBS. Cells that were grown in 8 well Lab-Tek slides were only used for immunofluorescence or autoradiographic studies. Trypan blue exclusion test was performed to determine the cell viability after trypsinization for each experiment. 3H-Thymidine

incorporation

The stromal cells cultured in multiwell dishes for 48 h as above were washed twice with Ca2+-Mp2+ free PBS and made quiescent by incubation in the presence of serum-free DMEMHam’s F-12 and incubated for various periods. The media were removed and replaced with DMEM-Ham’s F-12 either supplemented with 10% FBS or serum-free, containing 2 &i/ml 3Hthymidine and incubated for various periods ranging from 6 to 72 h. After the incubation the individual wells were washed with Ca*+-M$+ free PBS, trypsinized, and the cell suspensions in 200 ~1volume from each well were divided into three aliquots. One aliquot was used for determination of cell number using Coulter Counter ZM (Coulter Electronics, Hialeah, FL) and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

GROWTH

FACTOR REGULATION “i:

Y

OF ENDOMETRIAL

1 +

1

, /: I ” I ’ t, L( .-‘ i

w ,I

i”

‘#.,

.i ’

,I’

”

--..- _LI_

._

-\

‘i

“I.

i I FIG. 1. A shows a Nomarski interference contrast micrograph of endometrial stromal cells in primary culture and B shows immunofluorescence micrograph of stromal cells stained with antibodies to human vimentin indicating the homogeneity and purity of the culture. Magnification, A = ~127; B = ~117.

the other two aliquots were retained on individual filters (96 well plates Millipore filter device, Millipore, Bedford, MA). The filters were washed once with 10% trichloroacetic acid (TCA). three times with 5% TCA, and twice with 95% alcohol. Radio: isotope content retained on the filters were determined by scintillation spectroscopy using a Beckman LS 5000. Light

microscope

autoradiographic

studies

Endometrial stromal cells grown in eight well Lab-Tek slides under similar condition described above were used to determine their labeling index. The level of 3H-thymidine incorporation by stromal cells after incubation for various period of times ranging from 6-72 h were determined by autoradiography using Kodak NTB-2 emulsion as previously described (31). Immunocytochemical

studies

For immunohistochemistry, portions of 78 of the uteri collected were washed with PBS and cut to small pieces and fixed in Bouin’s solution overnight. The tissue pieces were washed with PBS, dehydrated in a series of graded ethanol, and paraffin

embedded. The tissue sections, 5-6 pm thickness, from 57 of

STROMAL

CELLS

2375

78 uterine tissue specimens processed, were cut on LKB historange, deparaffinized, and processed for immunocytochemical localization of EGF, PDGF-AB, as well as EGF and PDGF @receptors as previously described (11). Frozen sections, 6-10 pm thickness, from 8 uterine tissues were also used for immunolocalization of anti-porcine PDGF B-receptors (32). All the endometrial specimens used for stromal cells isolation and culturing were included in immunohistochemical studies. The primary antibody dilutions used in this study were; EGF (1:50), EGF receptor (l:lOO), PDGF-AB (l:lOO), anti-porcine PDGFP-receptor (50 rg immunoglobulin G/ml) and anti-human PDGF-a and -/3 receptor antibodies (5 pg immunoglobulin G/ ml). Immunofluorescence

microscopic

studies

Stromal cells grown on Lab-Tek slides for 48 h using DMEM-Ham’s F-12 supplemented with 10% FBS were washed with Ca’+-Me free PBS and immediately fixed with ice-cold methanol for 5 min. The slides were either used immediately for immunofluorescence observations or stored at -20 C until needed. The purity and homogeneity of stromal cells isolated and cultured for 48 h or after the 3rd passage were analyzed using monoclonal antibodies to human cytokeratin 19 (epithelial cells), desmin (smooth muscle cells), and vimentin (stromal cells). The slides were treated with 0.1% Tween 20 in PBS for 5 min, washed in PBS, and subsequently exposed to normal serum 1:50 dilution for 20 min, 1:lOO dilution of primary antibodies for 1 h and then to TRITC conjugated second antibodies for 30 min at dark. The stromal cells were also immunostained with antibodies to EGF, PDGF-AB, as well as EGF, human PDGF-a, and human PDGF-P receptors with similar dilution to those used for the immunohistochemical studies. The cells were extensively washed with PBS and viewed in Olympus BH2 microscope equipped for fluorescence microsCOPY. Statistical

analysis

Data are reported as the mean + SEM for the indicated experiments and analyzed using analysis of variance. All the experiments were repeated at least three times in triplicate.

Results Isolation and characterization

of endometrial

stromal

cells

The isolated endometrial stromal cells were homogenous and pure single cell populations devoid of any remnant of glandular or luminal epithelial or myometrial smooth muscle cells. Morphological observations of stromal cells after 48 h of incubation and third passage in primary culture indicated a monolayer of flat, spindle and fibroblast-like characteristic cells, suggesting their endometrial origin (Fig. 1A). Immunofluorescent microscopic observations using specific antibody to human vimentin (a class of intermediate filament proteins present in fibroblasts) indicated that over 95% of the cells were positively immunostained (Fig. 1B) suggesting the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

GROWTH

2376

FACTOR

REGULATION

OF ENDOMETRIAL

STROMAL

CELLS

Endo. 1992 Vol130.No4

A

Time

(days)

B

0 0

12

24

Time

(hours)

FIG. 2. A shows the growth curve of stromal cells incubated in the presence of 10% FBS. Cell numbers were determined for three separate cultures at each time point. Each point represents the mean cell number for a single culture in triplicate. The values are significantly different from nonstimulated controls (P < 0.0001). B demonstrates the time dependency of 3H-thymidine incorporation by stromal cells. The cells cultured for 48 h in the presence of 10% FBS, then incubated in serum+), and exhibited a gradual free conditions (left-hand side, +reduction in DNA synthesis. Addition of 10% FBS to quiescent stromal cells stimulates ‘H-thymidine incorporation in a time-dependent man0). Each point represents mean + SEM ner (right-hand side, Oof three separate triplicate experiments.

purity and homogeneity of the initial isolated stromal cells and their frbroblastic nature. Less than 5% of the cell populations were immunostained with either cytokeratin 19 or desmin (figures not shown). The immunofluorescent microscopic observation using the above antibodies on stromal cells after their third passage, also indicated similar patterns of immunostaining, suggesting that the endometrial stromal cells under the culture conditions used in this study retained their fibroblastic character. DNA synthesis and growth of endometrid

stromal cells

The stromal cell cultures in the presence of 10% FBS grow exponentially with a doubling time of 38.2 h (Fig.

FIG. 3. Light microscope autoradiographic Nomarski interference contrast photographs of stromal cells. Quiescent cells were labeled with 3H-thymidine for 48 h either in serum-free conditions (A) or in the presence of 10% FBS (B). Note the presence of few labeled nuclei in A, but extensive nuclear labeling indicating significantly higher 3Hthymidine incorporation in B. Magnification, ~127.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

GROWTH

FACTOR REGULATION

ZA). The stromal cells grown under these conditions reach confluency after 8 days. Stromal cells grown in the presence of 10% FBS for 48 h, then incubated in serumfree medium showed a significant reduction in 3H-thymidine incorporation in a time-dependent manner (Fig. 2B). 3H-Thymidine incorporation by quiescent stromal cells stimulated by 10% FBS, significantly increased by 24 h, and reached maximal after 48 h of incubation compared to nonstimulated serum free cultures (Fig. 2B). Stromal cell cultures derived from endometrial tissues at either phase of the menstrual cycle proliferated equally in the presence of 10% FBS, and did not change at various passage levels (up to 10 passages in this study). The labeling index also indicated a considerable reduction in nuclear labeling (75-80%) after incubation in serum-free conditions, and a gradual increase in labeling index in the presence of 10% FBS in a time dependent manner, reaching to approximately 98% nuclear labeling after 48 h of incubation (Fig. 3, A and B). The results of 3H-thymidine incorporation and the labeling index indicated that 48 h of incubation in a serum-free medium induced quiescency in stromal cells which could subsequently be stimulated to resume the cell cycle in the presence of 10% FBS.

OF ENDOMETRIAL

CELLS

2317

A

T

4 04

: N.S

1 0.5

1.0

2.0 5.0 EGF (lu/ml)

10

20

50

B 300-250.. 200-m 1x3-loo--

Effect of EGF and PDGF on DNA synthesis

EGF, PDGF-AB, and PDGF-BB added individually to the serum-free quiescent stromal cells stimulated 3Hthymidine incorporation in a dose-dependent manner (Fig. 4). The stimulation of 3H-thymidine incorporation occurred with EGF at concentration of 1.5 rig/ml (P < 0.001) and ,reached maximal at 5 rig/ml (P < 0.001) as compared to nonstimulated control (Fig. 4A). However, EGF at higher concentrations (over 15 rig/ml) significantly inhibits 3H-thymidine incorporation (P < 0.005) relative to the 5 rig/ml EGF. The dose-response experiments with PDGF-AB indicated a significant increase in 3H-thymidine in co rp oration by stromal cells at 3 rig/ml (P < O.OOl), but not at lower concentrations (Fig. 4B). Maximal stimulation occurred at 10 rig/ml PDGF-AB, which was significantly higher than that at 3 rig/ml (P < 0.001). No further stimulation was observed at higher PDGF-AB concentrations. PDGF-BB as low as 0.5 ng/ ml significantly stimulated 3H-thymidine incorporation (P < O.OOl), and was not statistically different from that induced by PDGF-BB at higher concentrations (Fig. 4C). The effect of PDGF-BB at 0.5 rig/ml was similar to that induced by PDGF-AB at 10 rig/ml (Fig. 4, B and C). The action of EGF and PDGF-AB on 3H-thymidine incorporation occurred in a time-dependent manner and reached maximal after 48 and 36 h of incubation respectively (Fig. 5, A and B). Stromal cells incubated in serumfree medium incorporated a significant amount of 3Hthymidine which reached maximal after 96 h of incuba-

STROMAL

v

50-07

NS

0.2

0.5 PDW

1

2

5

10

be/~)

C 350 300 250 200 150 100 50

i

OJ

NS

I 0.2

0.5 BE-PDGF

1

2

5

10

(&ml)

The dose dependency of EGF (A), PDGF-AB (B), and PDGFBB on 3H-thymidine incorporation by stromal cells after 46 h of incubation. The stromal cells incubated in serum-free media are shown on the &&hand side of the graphs indicated as NS. Note the inhibitory effect of EGF on SH-thymidine incorporation at 50 rig/ml compared with other concentrations (A). Points represent mean + SEM of three separate triplicate experiments from three uteri. Disintegrations per min/lOa cells for controls are A: 560 f 145, B: 497 + 164, and C: 334 f 79. FIG.

4.

tion (Fig. 5, A and B). The 3H-thymidine incorporation by stromal cells grown in the serum-free medium could

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

GROWTH

2378

FACTOR

REGULATION

OF ENDOMETRIAL

STROMAL

CELLS

Endo l 1992 Vol13o*No4

125 100 75 50 25 0

0

24

4s

72

Time

(hours)

Time

(hours)

90

120

rn)+PI

The effect of Ez and P, at 1 pM concentration and their combination (Ez + Pa) on 3H-thymidine incorporation by stromal cells incubated for 48 h in the absence (A) or presence of 15 rig/ml EGF (B) and 10 rig/ml PDGF-AB (C). Control in A is serum-free condition, B is 15 rig/ml EGF, and Fig. C is 10 rig/ml PDGF-AB. Points represent mean f SEM of five (A) and three (B and C) separate triplicate experiments from three to five individual uterus. Disintegrations per min/103 cells in the control experiments are A: 644 + 71, B: 1799 k 437 and C: 953 f 230. FIG.

C

0 NS

EGF

PDCF

E+P

5. A and B show the time dependency of EGF (15 rig/ml, A-A) and PDGF-AB (10 rig/ml VV) and nonstimulated serumfree medium (A-A, Vw) on 3H-thymidine incorporation by stromal cells. C indicates the 3H-thymidine incorporation of stromal cells incubated in serum-free medium (NS) or in the presence of EGF (15 rig/ml), PDGF-AB (10 rig/ml), or combination of EGF and PDGFAl3 (E + P) for 48 h. Points represent mean f SEM of three separate triplicate experiments from three uteri. Disintegrations per min/103 cells for control in C is 409 f 88. FIG.

be attributed to a low level of EGF produced by these cells, indicated by immunofluorescence microscopic observations (data not shown). Combination of 15 rig/ml

6.

EGF and 10 rig/ml PDGF-AB to the serum-free medium enhanced the 3H-thymidine incorporation by stromal cells and was significantly higher (P < 0.01) than that induced by EGF or PDGF-AB alone (Fig. 5C). EP and P4 at 1 pM concentration in serum-free medium have no effect on 3H-thymidine incorporation by stromal cells. In combination, however, they significantly increased 3H-thymidine incorporation (Fig. 6A). Moreover, EP and P4 did not further enhance the action of either 15 rig/ml EGF or 10 rig/ml PDGF-AB (Fig. 6, B and C). Immunofluorescence

localization

Endometrial stromal cells in culture express EGF, PDGF-AB, EGF, and PDGF-P receptors, indicated by immunofluorescent microscopic localization (Fig. 7). The immunostaining for EGF was very intense in the cytoplasm, particularly around the nuclear periphery, possi-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

GROWTH

FACTOR

REGULATION

OF ENDOMETRIAL

STROMAL

CELLS

2379

FIG. 7. Immunofluorescence micrographs of stromal cells immunostained for EGF (A), EGF receptor (B), PDGFAB (D), PDGF-j3 receptor (E), and PDGF-(Y receptor (F), anti-human PDGF-a and fl receptor antibodies were used in this presentation. Preabsorption of EGF receptor antibody with human term placental microvilli before application resulted in a considerable reduction in immunostaining (C). The controls for the other antibodies resulted in similar reduction in immunostaining (figures are not shown). Magnification, ~117.

bly associated with endoplasmic reticulum and golgi apparatus regions (Fig. 7A). The stromal cell immunostaining for EGF receptors was localized over both cytoplasmic as well as nuclear regions with more intensity in the cytoplasm (Fig. 7B). The immunofluorescence staining for PDGF-AB indicated that stromal cells also produce PDGF-AB (Fig. 7D), but with less intensity than that seen for EGF. The stromal cells immunostained with anti-human PDGF ,&receptor antibody, and with far less intensity with PDGF a-receptor antibody (Fig. 7, E and F). The stromal cells immunostained with both anti-porcine and anti-human PDGF P-receptor antibodies, with more intensity associated with cells immunostained using anti-human PDGF P-antibody (not shown). In the control experiments either preabsorption of the antibodies with purified proteins, EGF and EGF receptor (Fig. 7C) or replacement with nonimmune normal sera (PDGF-AB, PDGF LY,and PDGF P-receptors) resulted in a considerable reduction in immunostaining intensity (figures not shown).

Immunohistochemical

localization

Immunohistochemical observations indicated that endometrial tissue immunostained for EGF and PDGF-AB as well as EGF and PDGF-P receptors (Figs. 8 and 9). The immunostaining for EGF was associated with luminal and glandular epithelial as well as stromal cells (Fig. 8, A and B). The immunostaining was not uniform as some subpopulation of cells in all the cell types showed more intensity. This was more often observed with stromal than luminal or glandular epithelial cells. The inflammatory cells, namely monocytes/macrophages, infiltrated among the stromal cells showed intense immunostaining for EGF (Fig. 8A). These cells are not always present in this tissue. The endothelial and smooth muscle cells of the arterioles also immunostained for EGF (figure not shown). The immunostaining for EGF receptors showed immunoreaction associated with all the endometrial cell types including stromal cells (Fig. 8, D and E). Although there was some variation in immuno-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

COWTH

FACTOR

REGULATION

OF ENDOMETRIAL

STROMAL

CELLS

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

En& l 1992 Vol13O*No4

GROWTH

FACTOR

REGULATION

staining among the specimens examined, there were not noticeable differences in immunostaining intensity for EGF or EGF receptor in tissue from proliferative or secretory phases of the menstrual cycle. In the control experiments preabsorption of the EGF antibody with recombinant EGF (Fig. SC), or EGF receptor antibody with placental microvilli preparation (Fig. 8F) resulted in a considerable reduction in the immunostaining of endometrial tissues. The immunostaining for PDGF-AB was only observed in stromal cell compartment but not luminal and glandular epithelial cells with no differences in proliferative or secretory phases of the cycle (Fig. 9A). The immunostaining of endometrial tissue for PDGF-/3 receptors indicated that luminal and glandular epithelial as well as stromal cells immunostained with anti-human PDGF ,&receptor antibody (Fig. 9, C, D, and E). During the proliferative phase the immunostaining intensity for the PDGF P-receptor was higher in stromal cells than in luminal and glandular epithelial cells (Fig. 9, C and D). The intensity was considerably reduced in secretory phase particularly in stromal and luminal epithelial cells (Fig. 9E). The anti-porcine PDGF P-receptor antibody used on frozen endometrial sections only immunostained the stromal cells. The immunostaining intensity was similar in both phases, with lower intensity than that observed with anti-human PDGF-P receptor antibody (figure not shown). The anti-human PDGF a-receptor antibody was found not to be suitable for immunohistochemical experiments either in frozen or fixed tissue. Deletion of primary antibodies or their replacement with normal serum showed a lack of immunostaining of endometrial tissue for PDGF-AB and PDGF-P receptor (Fig. 9, B and F). Discussion

The results of the present study have indicated that human uterine endometrial tissue from either phase of the menstrual cycle contain immunoreactive EGF, PDGF-AB, as well as EGF and PDGF-/3 receptors. In addition to endometrial tissue, the isolated stromal cells in primary cell culture produce EGF and PDGF-AB and respond to their mitogenic actions, through the specific EGF and possibly PDGF-P receptors. Immunocytochemical observations in this study using antibody specific to human EGF indicated that all the major cell types in human endometrial tissue contain immunoreactive EGF. These results are in contrast with

OF ENDOMETRIAL

STROMAL

CELLS

2381

studies carried out in mouse demonstrating the presence of immunoreactive EGF only in luminal epithelial cell layer (9, 10). We have previously reported the presence of immunoreactive TGF-a in human uteri with a similar pattern of distribution as observed for EGF in this study (14). Furthermore, normal cycling endometrium and endometrial implants in surgically induced endometriosis in rat demonstrated a pattern of immunostaining for EGF and TGF-(x similar to that seen in human endometrium (11). The immunohistochemical observations of the endometrial tissues, coupled with immunofluorescence localization of EGF in primary stromal cell cultures suggests that in addition to luminal and glandular epithelial cells, stromal cells are also the site of EGF production in human. These observations are supported by recent reports that human endometrial tissue expresses EGF (12) and TGF-a (12, 13) mRNA, although no correlation with the phases of the menstrual cycle was reported (12). The lack of correlation in EGF immunoreactivity with the phases of menstrual cycle in our study suggests that either EGF production in stromal cells is regulated equally by Ez and Pq or is an independent process. In fact, the presence of immunoreactive EGF and a low but significant 3H-thymidine incorporation by quiescent stromal cells grown in serum-free media suggests local production of growth factors independent of sex steroid stimulation, at least in uitro. Uterine tissue from different species have been shown to be a source of EGF (9-11) and Ez administration significantly increased the immunoreactive EGF concentration in this tissue (9, 10). In mice uteri the immunoreactive EGF was present in late proestrus and on day 1 of pregnancy, but not in ovariectomized uteri, and EP, but not Pq treatment of ovariectomized mice stimulated the occurrence of both EGF mRNA and mature EGF rather than prepro-EGF (10). This suggests Ez-dependent de noua synthesis of the mature form of EGF not the processing of preexisting prepro-EGF (10). Recent reports also indicate the ability of EGF to induce an estrogenic effect in ovariectomized rodent genital tract growth and differentiation (28, 29). The presence of EGF receptors in all the major uterine cell types in human and rodent have previously been demonstrated (15-22) with highest concentrations associated with stromal cells (15, 18, 22). Immunocytochemical observations in this study indicated the presence of EGF receptors in luminal and glandular epithelial and stromal cells with similar distribution to previous autoradiographic and immunohistochemical observations in

FIG. 8. Immunohistochemical localization of EGF (A-C) and EGF receptor (D-F) in human endometrial tissue. Note the presence of immunostaining associated with luminal (A) and glandular epithelial (B) and stromal cells (A and D). Preabsorption of EGF antibody with recombinant human EGF (C) and EGF receptor antibody with purified human placental microvilli (F) resulted in considerable reduction in immunostaining of the tissues, shown in Nomarski interference contrast. Magnification, A, B; D; and E; ~300 C and F, X151.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

2382

GROWTH

FACTOR

REGULATION

OF ENDOMETRIAL

STROMAL

CELLS

./

.- ./

/

;-’ ,

.,’,.

_: ,+ :,

i ,. “-I i

, I.

.p’

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

Endo l 1992 Voll2O’No4

GROWTH

FACTOR REGULATION

human and rodent uteri (11, 15, 17, 18, 22). The data indicate a lack of variation with the phases of the menstrual cycle, which is in agreement with previous immunohistochemical, autoradiographic, and binding studies (15-17); however, it differs with the binding studies reported in rat during the estrous cycle (19). Rat uterine EGF receptor expression has been shown to be regulated by EB, In cycling rats the EGF receptor levels have been shown to reach maximal at proestrus and parallel the alteration in plasma Ez and their occupied nuclear receptors (19). In the mature ovariectomized rat, Ez administration induced an increase in expression of EGF receptor mRNA and EGF receptor level (21). In addition to EGF and EGF receptors our observations indicated for the first time that human endometrial tissue contains immunoreactive PDGF-AB and PDGF-B receptor. A recent report indicates the expression of PDGF-BB and low levels of PDGF-AA in human uterine tissue (23). The presence of immunoreactive PDGF-AB in this study and transcripts for PDGF-A and B chains (23), suggest that human endometrium produces the three different isoforms of PDGF. Human uterus has previously been shown to contain PDGF receptors (27), however the cellular distribution and the receptor types have not been demonstrated. Our immunohistochemical and immunofluorescence microscopic observations indicate that the predominant PDGF receptor in human endometrium is the P-type with low levels of PDGF-a! receptors. Porcine uterine tissue has been shown to be a major source of PDGF-/3 receptor, associated with endometrial stromal cells (24, 26). In situ hybridization using complementary DNA to the PDGF-fi receptor demonstrated the presence of its mRNA preferentially localized in stromal cells (26). Using anti-porcine PDGF-6 antibody in frozen sections, we observed similar cellular localization in human endometrial stromal cells as reported in porcine (24, 26). In contrast, using the antihuman PDGF-P receptor antibody a different pattern of cellular distributions were observed, both in frozen and fixed tissue. The reason for these differences is not clear but it could be related to the antibody’s specificity applied in tissue sections, although both antibodies immunostain the stromal cells in culture. There is no data available regarding PDGF and PDGF receptor regulation by steroids. However, it has been shown that transcripts for PDGF-BB chain in human endometrium appear to be expressed equally in all phases of the menstrual cycle

OF ENDOMETRIAL

STROMAL

CELLS

2333

(23). Additionally, our immunohistochemical observation suggest that PDGF-AB is expressed equally in both proliferative and secretory phases of the cycle. In contrast, we observed a considerable reduction in PDGF-P receptor immunostaining in secretory phase primarily in luminal epithelial and stromal cells, suggesting a possible modulation by steroids. These data and previous reports in human uteri indicate that of EGF, TGF-a, TGF-/I, PDGF-AB, PDGF-BB, EGF receptor, and PDGF-P receptor, only PDGF-8 receptors show a variable pattern related to the phases of the cycle suggesting their possible regulation by sex steroids in ho. The present study indicates that both EGF and PDGF stimulate 3H-thymidine incorporation by human endometrial stromal cells in a dose- and time-dependent manner. The rate of 3H-thymidine incorporation due to EGF and PDGF-AB treatment was similar and after 48 h of incubation roughly doubled, relative to the serumfree (control) condition for each growth factor. In the serum-free condition, 3H-thymidine incorporation continues to increase with time. However, the 3H-thymidine incorporation (-100-300 dpm/103 cells) did not increase significantly even after 120 h of incubation. These variations in 3H-thymidine incorporation were due to differences in individual stromal cell cultures. EGF at concentrations of 50 rig/ml inhibited the 3Hthymidine incorporation of stromal cells relative to lower EGF concentrations. Such inhibitory action of EGF has previously been reported in A-431 cells and other cell lines (33, 34). Furthermore, the interaction of EGF and PDGF in stimulating DNA synthesis shows a significant enhancement in the action of either EGF or PDGF alone. PDGF-BB at 0.5 rig/ml concentration is as effective as higher concentration of PDGF-AB. We are currently examining the mitogenic action of PDGF-AA in these cells; however, considering the low level of PDGF (Y receptors in stromal cells and the lack of interaction of PDGF-AA with /? receptors, it appears that the mitogenic effect of PDGF in human endometrial tissue is predominantly through the PDGF-/3 receptors. Furthermore it has been shown that in diploid human fibroblast PDGF(Y receptor makes up less than 10% of the total PDGF receptors and PDGF-BB is the major mitogen in PDGF action of these cells (35). Both Ez and Pq were ineffective in stimulation of 3Hthymidine incorporation by stromal cells in serum-free conditions, or in the presence of EGF, PDGF, or their

FIG. 9. Immunohistochemical localization of PDGF-AB (A) and PDGF-8 receptor (C-E) in human endometrial tissue. A indicates the presence of immunostaining of PDGF-AB in association with stromal ceils, without any staining of the glandular epithelial cells (A). Replacement of primary antibody with non-immune normal serum resulted in considerable reduction in immunostaining of PDGF-AB (B). C-E show immunostaining of endometrial tissue with anti-human PDGF @ receptor antibody associated with luminal and glandular epithelial and stromal cells in proliferative (C and D) and secretory phase of the menstrual cycle (E). Note a considerable reduction in immunostaining intensity of luminal epithelial and stromal cells in E. Replacement of primary antibody with nonimmune normal serum resulted in considerable reduction in immunostaining of PDGF B receptor (F) shown in Nomarski interference contrast. Magnification, X151.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

2384

GROWTH

FACTOR REGULATION

combinations. Similar results describing the lack of mitogenic effect of EP and Pq in rodent uterine epithelial and stromal cell, and vaginal epithelial cell in vitro have been previously reported (36, 37). Even in uterine organ cultures, which contain all the cell types, with possible cell-cell interactions, Ez was ineffective in stimulating DNA synthesis (38). Insulin-like growth factor-I does not have a significant effect on DNA synthesis in rat uteri, however in a dose-dependent manner increased DNA synthesis in the presence of 1 PM E, (38). Our results indicate an equal stimulation of stromal cells derived from endometria at either phase of the menstrual cycle and after several passages (up to 10) by EGF and PDGF. This correlates well with the presence of immunoreactive EGF, PDGF, and their receptors in this study and previously reported data on EGF receptors (15, 16) and expression of PDGF-BB (23), indicating, at least for these growth factors, a lack of variation in regard to the phases of menstrual cycle. The additive or synergetic mechanisms of growth factors on cell proliferation is complex and not well understood. Growth-arrested fibroblasts have been shown to require more than a single mitogen to stimulate DNA synthesis. PDGF is regarded as a competency factor, required to make the Go to G1 transition, whereas, EGF andIGF-I are considered progression factors, stimulating cells in G1 to complete the cell cycle after becoming competent (39). EGF and PDGF appear to be mitogenic in quiescent human endometrial stromal cells. However, initiation and maximal stimulation of 3H-thymidine incorporation by these cells due to EGF or PDGF appeared later than in FBS-stimulated cells. In addition EGF and PDGF treatment of quiescent stromal cells under the conditions used in this study did not complete the cell cycle, reflected by lack of dilution in the disintegration per min/103 cells as seen in FBS-treated cells. Under serum-free conditions the stromal cells did not become completely quiescent, suggesting the possibility of minimal production of growth factors which may, to an extent, allow the cells to become competent for further stimulation by progression factors such as EGF. This indicates that while stromal cells respond to the mitogenie action of EGF and PDGF, they clearly require interaction with more than a single growth factor for maximal stimulation. EGF and PDGF trigger several events in their target cells which are considered to be early signals for mitogenie response, Na*+/H+ exchange, phosphoinositol turnover, Ca2+ fluxes, and an increase in c-fos and c-nyc genes expression (40-42). In human fibroblasts PDGF induces the expression of TGF-P mRNA (43) and the mitogenic action of all three PDGF isoforms in these cells is enhanced by TGF-& through the induction of PDGF-a receptors (35). These observations (35,43) and

OF ENDOMETRIAL

STROMAL

CELLS

Endo. 1992 Voll30. No 4

the presence of immunoreactive TGF-p in human endometrial tissue (14) suggest that the action of PDGF in uterine tissue may involve the interaction among several growth factors with multifunctional properties. In summary these observations show for the first time the presence and cellular distribution of EGF, PDGFAB, and PDGF-P receptor as well as confirming our previous observations regarding the EGF receptor in human endometrial tissue. Moreover, these data indicate that EGF, PDGF-AB, and PDGF-BB are mitogenic in human endometrial tissue and the major mitogenic action of PDGF is mediated through the PDGF-/3 receptor. Acknowledgments The authors would like to thank the Department of OB/GYN Gynecologists for their help in the collection of the tissues.

References 1. Clarke CL, Sutherland RL 1990 Progestin regulation of cellular proliferation. Endocr Rev 11:266-301 2. Soto AM, Sonnenschein C 1987 Cell proliferation of estrogensensitive cells: the case for negative control. Endocr Rev 844-52 3. Fritz MA, Westfahl PK, Graham RL 1987 The effect of luteal phase estrogen antagonism on endometrial development and luteal function in women. J Clin Endocrinol Metab 65:10061013 4. Knabbe C, Lippman ME, Wakefield LM, Flanders KC, Kasid RD, Dickson RB 1987 Evidence that transformins arowth factor-beta is a hormonally regulated negative growth factor in human breast cancer cells. Cell 48:417-428 5. Murphy LJ, Ghahary A 1990 Uterine insulin-like growth factor-I: regulation of expression and its role in estrogen-induced uterine proliferation. Endocr Rev 11:443453 6. Fisher DA, Lakshmanan J 1990 Metabolism and effects of epiderma1 growth factor and related growth factors in mammals. Endocr Rev 11:418-442 7. Gill GN 1990 Regulation of EGF receptor expression and function. Mol Reprod Dev 27~46-53 8. Westermark B 1990 The molecular and cellular biology of plateletderived growth factor. Acta Endocrinol (Copenh) 123:131-142 9. DiAugustine RP, Petrusz P, Bell GI, Brown CF, Korach KS, McLachlan JA, Tseng CT 1988 Influence of estrogen on mouse epidermal growth factor precursor protein and messenger ribonucleic acid. Endocrinology 122~2355-2363 10. Huet-Hudson YM, Chakraborty C, De SK, Suzuki Y, Andrews GK, Dey SK 1990 Estrogen regulates the synthesis of epidermal growth factor in mouse uterine epithelial cells. Mol Endocrinol4:510-523 11. Simms JS, Chegini N, Williams RS, Rossi AM, Dunn Jr WA 1991 Identification of epidermal growth factor, transforming growth factor (TGF-(r) and epidermal growth factor receptors in surgically induced endometriosis in rat. Obstet Gynecol 78850-857 12. Haining REB, Schofield JP, Jones DSC, Rajput-Williams J, Smith SK 1991 Identification of mRNA for epidermal growth factor and transforming growth factor-a present in low copy number in human endometrium and decidua using reverse transcriptase-polymerase chain reaction. J Mol Endocrinol6:207-214 13. Williams RS, Schultz GS, Geoghegan TE, Steffan MC, Yussman MA 1989 Variation in the level of transforming growth factor alpha (TGF-ol) mRNA during the menstrual cycle. In: Hirshfield AN (ed) Growth Factors and the Ovary. Plenum Publishing, New York, pp 205-208 14. Williams RS, Chegini N Localization of TGF-(U and TGF-P in human uterine and ovarian tissue. Program of 46th Annual Meeting of the American Fertility Society, Washington DC, 1990, p S5 (Abstract) 15. Chegini N, Rao ChV, Wakim N, Sanfilippo J 1986 Binding of iz61-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.

GROWTH

16. 17.

18. 19.

20. 21. 22.

23.

24.

25. 26.

27.

28.

FACTOR

REGULATION

epidermal growth factor in human uterus. Cell Tissue Res 246:543548 Hoffmann GE, Rao ChV, Barrows GH, Schultz GS, Sanfilippo JS 1984 Binding sites for epidermal growth factor in human uterine tissues and leiomyomas. J Clin Endocrinol Metab 58880884 Berchuck A, Soisson AP, Olt GJ, Soper JT, Clarke-Pearson DL, Bast Jr RB, McCarty Jr KS 1989 Epidermal growth factor receptor expression in normal and malignant endometrium. Am J Obstet Gynecol161:1247-1252 Lin T-H, Mukku VR, Verner G, Kirkland JL, Stance1 GM 1988 Autoradiographic localization of epidermal growth factor receptors to all major uterine cell types. Biol Reprod 38403-411 Gardner RM, Verner G, Kirkland JL, Stance1 GM 1989 Regulation of uterine epidermal growth factor (EGF) receptors by estrogen in the mature rat and during the estrous cycle. J Steroid Biochem 32339343 Chakraborty C, Tawfik 0, Dey SK 1988 Epidermal growth factor binding in rat uterus during peri-implantation period. Biochem Biophys Res Commun 154564-569 Lingham R, Stance1 GM, Loose-Mitchell DS 1988 Estrogen induction of the epidermal growth factor receptor mRNA. Mol Endocrino1 2:230-235 Bossert NL, Nelson KC, Ross KA, Takahashi T, McLachlan JA 1990 Epidermal growth factor binding and receptor distribution in the mouse reproductive tract during development. Dev Biol142:7585 Boehm KD, Daimon M, Gorodeski IG, Sheean LA, Utian WH, Ilan J 1990 Expression of the insulin-like and platelet-derived growth factor genes in human uterine tissues. Mol Reprod Dev 27:93-101 Heldin CH, Backstrom G, Ostman A, Hammacher A, Ronnstrand L. Ruhin K. Nister M. Westermark B 1988 Bindine of different dimeric forms of PDGF to human fibroblasts: evi&nce for two separate receptor types. EMBO J 21387-1393 Ronnstrand L, Beckmann MP, Faulders B, Ostman A, Ek B, Heldin CH 1987 Purification of the receptor for platelet-derived growth factor from porcine uterus. J Biol Chem 262:2929-2932 Terracio L, Ronnstrand L, Tingstrom A, Rubin K, Claesson-Welsh L, Funa K, Heldin CH 1988 Induction of platelet-derived growth factor receptor expression in smooth muscle cells and fibroblasts upon tissue culturing. J Cell Biol 107:1947-1957 Fayed YM, Tsibsis JCM, Langenberg PW, Robertson AL 1989 Human uterine leiomyoma cells: binding and growth responses to epidermal growth factor, platelet-derived growth factor, and insulin. Lab Invest 60:30-37 Nelson KJ, Takahashi T, Bossert NL, Walmer DK, McLachlan JA 1991 Epidermal growth factor replaces estrogen in the stimulation of female genital-tract growth and differentiation. Proc Nat1 Acad Sci USA 88:21-25

OF ENDOMETRIAL

STROMAL

CELLS

2385

29. Bass BE, Williams RS, Rossi AM, Chegini N, Simms J, Epidermal growth factor (EGF) infusion in rat uterine horns stimulates an estrogenic effect. Program of 47th Annual Meeting of the American Fertilitv Societv. Orlando FL. 1991.0-082 S36 (Abstract.1 30. Noyes RW, He&g AT, Rock J 195i Dating of ehdometrial biopsy. Fertil Steril 1:3-25 31. Chegini N, Safa AR 1987 Influence of mitoxantrone on nucleolar function in MDA-MB-231 human breast tumor cell line. Cancer Lett 32327-336 32. Ronnstrand L, Terracio L, Claesson-Welsh L, Heldin C-H, Rubin K 1988 Characterization of two monoclonal antibodies reactive with the external domain of the platelet-derived growth factor receptor. J Biol Chem 263:10429-10435 33. Polet H 1990 Epidermal growth factor stimulates DNA synthesis while inhibiting cell multiplication of A-431 carcinoma cells. Exp Cell Res 186390-393 34. Hirai M, Gamou S, Minoshima S, Shimizu N 1988 Two independent mechanisms for escaping growth factor-mediated growth inhibition in epidermal growth factor receptor-hyperproducing human tumor cells. J Cell Biol 107:791-799 35. Ishikawa 0, LeRoy EC, Trojanowska M 1990 Mitogenic effect of transforming growth factor 81 on human fibroblasts involves the induction of platelet-derived growth factor (Y receptors. J Cell Physiol145:181-186 36. Tomooka Y, DiAugustine RP, McLaachian JA 1986 Proliferation of mouse uterine epithelial cells in vitro. Endocrinology 118:10111018 37. Iguchi T, Uchima F-DA, Ostrander PL, Bern HA 1983 Growth of normal mouse vaginal epithelial cells in and on collagen gels. Proc Nat1 Acad Sci USA 80:3743-3747 38. Ghahary A, Chakrabarti S, Murphy W 1990 Localization of the sites of synthesis and action of insulin-like growth factor-I in the rat uterus. Mol Endocrinol4:191-195 39. Pardee AB, Coppock DL, Yang HC 1986 Regulation of cell proliferation and the onset of DNA synthesis. J Cell Sci [Suppl] 4:171180 40. DeLaat SW, Moolenaar WH, Defize LH, Boonstra J, Van Der Saag PT 1985 Ionic signal transduction in growth factor action. Biochem Sot Symp 50:205-220 41. Freter CE, Lippman ME, Cheville A, Zinn S, Gelmann EP 1988 Alteration in phosphoinositide metabolism associated with 17j3estradiol and growth factor treatment of MCF-7 breast cancer cells. Mol Endocrinol2:159-166 42. Bravo R, Buckhardt J, Curran T, Muller R 1985 Stimulation and inhibition of growth by EGF in different A431 cell clones is accompanied by the rapid induction of c-fos and c-myc protooncogenes. EMBO J 4:1193-1197 43. Paulsson Y, Hammacher A, Heldin C-H, Westermark B 1987 Possible positive autocrine feedback in the prereplicative phase of human fibroblasts. Nature 328715-717

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 09:45 For personal use only. No other uses without permission. . All rights reserved.