- Email: [email protected]
The effect of growth factors on the proliferation of human endometrial stromal cells in culture
The effect of growth factors on the proliferation of human endometrial stromal cells in culture Mary G. Hammond, MD," Sung-Tack Oh, MD,b Joyce Anners, MA," Eric S. Surrey, MD,e and Jouko Halme, MD, PhD"
Chapel Hill, North Carolina, Kwangju, South Korea, and Los Angeles, California OBJECTIVE: Development of ectopic implants of endometriosis is associated with both an inflammatory response by macrophages and endometrial stromal cell proliferation. Macrophages are capable of releasing a variety of inflammatory mediators, including growth factors. To assess the impact of such factors on endometrial tissue, we have studied the effects of recombinant growth factors, fibroblast growth factor, epidermal growth factor, transforming growth factor-a, and inflammation mediators transforming growth factor-J3, and tumor necrosis factor-a on human endometrial stromal cell proliferation. STUDY DESIGN: Increasing concentrations of these compounds were added to cultures of primary, secondary, and long-term stromal cells and the cells were harvested at 24, 48, and 72 hours. RESULTS: Epidermal growth factor, transforming growth factor-a, transforming growth factor-J3, and fibroblast growth factor induced a statistically significant, dose-dependent increase in stromal cell thymidine uptake of 1.5- to fivefold. The cytokine tumor necrosis factor had no effect alone, but the combination of fibroblast growth factor and tumor necrosis factor had a synergistic effect, increasing cell proliferation 25% to 84% over fibroblast growth factor alone. CONCLUSION: The stromal cell response to a wide range of cell growth effectors and the potential of mediators like tumor necrosis factor-a to synergize suggest that such macrophage-secretory products may contribute to proliferation of endometrial implants in vivo. (AM J OBSTET GVNECOL 1993; 168: 1131-8.)
Key words: Endometriosis, cytokines, growth factors, endometrial stromal cells
Both intrauterine and ectopic endometrium are composed of two major cell types: glandular epithelium and stromal cells. The stromal cells form a framework surrounding the glands and blood vessels that are important regulators of growth and differentiation of the epithelial cells. Epidermal growth factor (EGF) binding has been demonstrated in human endometrial stroma and endometriosis implants in rats. I EGF and transforming growth factor-a (TGF-a) production was noted in endometriosis tissue in the rat. 2 These growth factors and receptors have also been recently reported in human endometriosis tissue. 3 The autocrine and paracrine effects of endogenous growth factors or the response of stromal cells to cytokines released by pelvic From the Department of Obstetrics and Gynecology, University of North Carolina School ofMedicine, a the Department of Obstetrics and Gynecology, Chonnam University Medical School,' and the Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, University of California-Los Angeles School of Medicine.' Supported in part by National Institutes of Health grant No. HD21546. Presented by invitation at the Eleventh Annual Meeting of the American Gynecological and Obstetrical Society, Hot Springs, Virginia, September 10-12, 1992. Reprint requests: Mal) G. Hammond, MD, Department of Obstetrics and Gynecology, University of North Carolina School of Medicine, CB # 7570, Chapel Hill, NC 27599. Copyright © 1993 by Mosby-Year Book, Inc. 0002-9378/93 $1.00 + .20 6/6/44824
macrophages may regulate initiation or continuation of the peritoneal endometriosis implant. 4 • 5 For several years this laboratory has investigated the role of macrophages and cytokines in the development and maintenance of endometriosis." Recently our attention has been focused on the use of endometrial stromal cells in culture as a model to study the effects of peritoneal fluid and macrophage-derived cytokines on cell proliferation'" 7-9 This study addresses the effect of a series of growth factors and inflammation mediators, including estradiol, fibroblast growth factor (FGF), EGF, TGF-a, transforming growth factor-(3 (TGF-(3), and tumor necrosis factor (TN F) alone or in combination, on stromal cell proliferation.
Material and methods Tissue preparation. As previously described:' 7 endometrial biopsy specimens were obtained from 18 regularly cycling women (mean cycle day 18.8) immediately before diagnostic laparoscopy or laparoscopic tubal sterilization. The protocol was approved by the University of North Carolina Human Rights Committee, and signed informed consent was obtained from each subject. Endometrium, obtained by aspiration, was placed immediately in Ham's F-I0 culture media (Gibco, 1131
1132
Hammond et al.
April 1993
Am J Obstet Gynecol
en "'0
10
EGF
c
--O.Sng/ml ----.A.--- S.O ng/ml _ _ sOng/ml
I/)
~Controls
CU ::J
o
.c
t:.
5
::E
a..
a
72 HR
48 HR
TIME Fig. 1. Effect of varying concentrations of EGF on tritiated thymidine incorporation by freshly explanted (primary) stromal cells in culture.
::E
a.. a w
>_I/)
~~
15
10
::ll/)
::E5
::l.c
at:. a:
..•
5
J:
C\I ......
•
0
0
First passage Sixth passage Freshly explanted
0.5 5.0 EGF (ng/ml)
50
Fig. 2. Effect of varying concentrations of EGF on tritiated thymidine incorporation by cultured stromal cells from primary, secondary, and sixth passage cultures.
Grand Island, N.Y.) supplemented with 50 J.Lg/ml gentamicin, 0.4 mg/ml of D-glucose, and 0.03% L-glutamine (Sigma, St. Louis) in ITS plus Premix (Collaborative Research, Bedford, Mass.). This media, with modification, was used throughout. Washed tissue was treated with 0.25% type I collagenase (Worthington) to disperse the cells. Mter centrifugation, the pellet was resuspended in media containing 10% calf serum. The cells were strained through a 38 J.Lm sieve to separate stroma from glands. Cells were allowed to adhere to plastic culture dishes, and the debris was aspirated. The adhered cells were rinsed with fresh media and then mechanically removed from 10-cm plates and counted. The average yield of cells was 9.0 x 106 (range 1.25 to 51 x 106 ). Cells were then diluted and plated in 96-well plates (Costar, Cambridge, Mass.), suspended in 200 J.LI of media containing 10% calf serum, and incubated at 37° C in 95% air and 5% carbon dioxide. Cells for primary and secondary culture conditions
were pipetted 20,000 cells per well, cells for tertiary culture were plated at 10,000 per well. Cells for secondary and tertiary culture were allowed to grow to confluence in 10% calf serum, trypsinized, and harvested and replated. 10 Tertiary cultures had undergone six passes before study. Endometrial stromal cell predominance (95%) was confirmed by phase contrast microscopy. Cell proliferation assays. Mter 24 hours media and dead cells from preincubated plates were aspirated, the cells washed with phosphate-buffered saline solution, and serum-free Ham's F-IO added. The appropriate growth factor was added in quadruplicate to all but 12 control wells. Cells were incubated for 24 to 72 hours. One microcurie per well of tritiated thymidine (New England Nuclear, Boston) was added to the wells at 0, 24, and 48 hours, respectively, and incubated for 24 hours. Mter thymidine incorporation cells were detached with ethylenediaminetetraacetic acid and the cellular material collected on filter disks by means of a
Hammond et al.
Volume 168, Number 4 Am J Obstet Gynecol
1133
.....J
0
II: I-
Z
0
100
U
u.
0
eft. 0
1.25
12.5
TGF~
125
0.5
5.0
50
TGFex (ng/ml)
(ng/ml)
Fig. 3. Effect of TGF-a and TGF-13 on stromal cell proliferation in primary culture as percent of control conditions.
-
20 TGFI3
UJ
- - 1.25 ng/ml --.....- 12.5 ng/ml ____ 125 ng/ml
'0
c:
CU UJ :::J
0 ..r: 10
I:::-
~
0..
U
o
~
24 HR
48 HR
; 72 HR
TIME Fig. 4. Effect of varying concentrations of TGF-13 on tritiated thymidine incorporation by primary endometrial stromal cells in culture.
semiautomatic cell harvester (Skatron, Sterling, Vir.). Radioactive tritiated incorporation was reported in absolute counts per minute. The factors investigated were recombinant human EGF, basic FGF, human recombinant TGF-Q and TGF-f3 (Collaborative Research), and recombinant human TNF-Q (Genzyme, Boston). Statistical analysis. Cumulative counts per minute for cytokine incubation, a summation of total counts per well assayed at 24, 48, and 72 hours, was compared with control wells. Significance was determined by grouped Student t tests.
Results EGF. Results of thymidine incorporation in four separate experiments confirm a statistically significant (p < 0.01) stimulatory effect of 0.5, 5.0, and 50 ng/ml concentrations in primary cell culture (Fig. 1). Maximum stimulation in separate experiments ranged from 1.5- to 4.6-fold. EGF had a stimulatory effect on secondary cells at 0.5 and 5.0 ng/ml and less effect at 50
ng/ml. Maximum stimulation ranged from 104- to 2.6fold in three experiments. Tertiary (sixth passage) cells demonstrated a significant but smaller stimulatory effect (Fig. 2) with a maximum range of 1.2- to lA-fold. Other factors. TGF-Q and TGF-j3 were studied at three concentrations in five separate experiments (Fig. 3). TGF-Q showed maximal stimulation of 1.3- to 1.7fold. TGF-j3 was stimulatory at 1.25, 12.5, and 125 ng/ml in freshly explanted cells (Fig. 4) and in secondary cells as well. Stimulation ranged from 1.7- to 4.3fold over untreated cells. FGF was investigated in five experiments. Concentrations of 0.01 to 100 ng/ml were stimulatory (Fig. 5) with maximal stimulation of 1.9- to 3.1-fold. When TNF (0.05 to 50 ng/ml) was studied, no effect or slight inhibition (71 % to 100% of control) of cell proliferation was noted in five experiments (Fig. 6). Estrogen was studied in concentrations of 1 x 10- 10 to 1 X 10- 8 mol/L in three studies; no consistent response was noted in any system. Combined factors. Attention was then turned to
1134 Hammond et al.
April 1993
Am
J Obstet Gynecol
60,000 ____ 100 ng/ml
50,000
----6-- 10 ng/ml ~
40,000
1.0ng/ml
_O.1ng/ml _O.01ng/ml
30,000
--e--
Control
20,000 10,000
04F~--------~----------~------------U 24 72 o 48 TIME (hrs) Fig. 5. Effect of varying concentrations of FGF on thymidine incorporation by stromal cells in culture.
2.2
-
2.0
TNF 0.5 ng/ml TNF 5.0 ng/ml
c: 1.8 Ql
TNF 50 ng/ml
C
E Ql
u c: 1.6 tIS
.c: c: w 1.4 1.2 1.0
0.1
1.0
10
100
FGF ng/ml Fig. 6. Enhancement of FGF stimulated tritiated thymidine incorporation by varying concentrations ofTNF.
factors in combination. The effect of FGF and TNF-o: on stromal cells was evaluated. A synergistic effect of the two factors was noted. Examples of individual experiments are presented in Fig. 6. Although no clear TNF dose-response relationship was noted, TNF stimulated FGF-dependent thymidine incorporation maximally at the lower FGF concentrations (0.1 to 1.0 ng/ml). The stimulation varied from l.25 ± 0.08- to l.84 ± 0.17fold. The increase was statistically significant in all but FGF 100 ng/ml. Comment
Our study supports the concept that endometrial stromal cells in serum-free culture can respond appropriately to many growth factors and cytokines known to be released by macrophages and demonstrated to be present in endometrial tissues. The well defined growth factors EGF, TGF-o:, TGF-/3, and FGF stimulate cell proliferation in our system by a factor of 1.5-fold to fivefold. This sensitivity persists throughout six cell passages, eliminating tissue or se-
rum contaminants as a factor. Stromal cells from both the proliferative and secretory phase responded with stimulation of cell proliferation. Other recombinant cytokines and growth factors have been studied in in vitro systems. Platelet-derived growth factor, a well characterized secretory product of activated macrophages that is a potent mitogen for fibroblasts, also stimulates stromal cell growth. 7 Interleukin1, a pleiotropic cytokine with multiple inflammatory and growth regulatory activities, was inhibitory in our model. 9 Other growth factors synthesized by macrophages have been evaluated in an endometrial cancer cell line. l l EGF and TGF-o:, homologous mitogens for a variety of cells, also stimulated these cells. In addition to cell proliferative effects, these factors have other effects on endometrial stromal cells. TNF appears to augment the adhesion of stromal cells to mesothelial cells in culture!! and to stimulate the synthesis of the cytokine interleukin-6 messenger ribonucleic acid and protein and its release into the medium.!2 TGF-/3, in addition to stimulating cell proliferation in
Volume 168, Number 4 Am J Obstet Gynecol
our system, also increased the propro-endothelin-l messenger ribonucleic acid levels in stromal cells. I3 Endothelin-l is postulated to regulate blood flow to the endometrium and might affect implant vasculature as well. Matthews et al. 14 have characterized cultured human epithelial and stromal cells from normal endometrium and endometriosis and report close structural similarity. This supports the suitability of our system as a model for endometriosis derived stromal cells. Two areas of impact should be considered regarding the effect of growth factors and cytokines on endometriosis. The first is the effect of peritoneal macrophages and peritoneal fluid on implantation and growth of endometrial tissue arriving in the pelvis during retrograde menstruation, and the second is autocrine and paracrine effects of the growth factors produced by stromal cells themselves. The role of peritoneal fluid components and macrophage-derived secretory products in the development of endometriosis has been studied for several years. Peritoneal fluid from women with endometriosis is increased in volume and contains a higher concentration of macrophages than it does in normal women. 6 • 15 Macrophages from this fluid secrete increased quantities of cytokines, in particular TNF. 16. 17 Peritoneal fluid from endometriosis patients has been demonstrated to stimulate stromal cell proliferation in vitro" Macrophage-conditioned media has been demonstrated to stimulate proliferation of an endometrial cancer cell line" and mouse endometrial stroma cells in vitro.5 A potential direct effect of macrophages usually present in endometrial tissue has also been shown by coculture. 18 FGF acts as a competence factor in the fibroblast. This confers on the cell the ability to respond to other mitogens such as EGF. Modulation of the ability of j3-FGF to stimulate keratinocyte cell proliferation has already been demonstrated. A synergistic effect of j3-FGF and EGF was noted with threefold to fourfold higher cell numbers.19 In our system FGF acts as a mitogen stimulating cell proliferation. The production of TNF -IX, a cytokine that has similar biologic activity to interleukin-l, is increased in activated macrophages isolated from the peritoneal cavity of women with endometriosis." It had no effect on cell proliferation in our system, but increased synergistically the stimulatory effect of lower concentrations of FGF. TNF would seem to act as a competence factor in increasing the cells' response to FGF. This interaction of a stromal cell growth factor and a macrophage inflammation mediator may have significance for understanding the interaction of these systems in women with endometriosis. In addition to peritoneal fluid growth factors, endometrial implants are also exposed to endogenous
Hammond et al.
1135
factors. Immunohistochemical studies have revealed the presence ofEGF, platelet-derived growth factor, TGF-IX, and EGF and platelet-derived growth factor receptor in human endometrium and induced endometrial implants in the rat. I. 2. 20 EGF stimulates the proliferation of many cell types, including smooth muscle cells and human uterus. Activity is mediated through a specific high-affinity surface receptor. TGF-IX binds to the same receptor with similar effects. We conclude that growth factors and cytokines released by peritoneal macrophages or endometrial stromal cells in endometrial tissue may promote implantation and proliferation of ectopic endometrium in the peritoneal cavity. REFERENCES 1. Chegini N, Rao Cv, Wakim N, Sanfilippo J. Binding of 125I-epidermal growth factor in human uterus. Cell Tissue Res 1986;246:543-8. 2. SimmsJS, Chegini N, Williams RS, Rossi AM, Dunn WAJr. Identification of epidermal growth factor, transforming growth factor-alpha and epidermal growth factor receptor in surgically induced endometriosis in the rat. Obstet Gynecol 1991;78:850-7. 3. Melega C, Balducci M, Bulletti C, Galassi A, Jasonni VM, Flamigni C. Tissue factors influencing growth and maintenance of endometriosis. Ann N Y Acad Sci 1991;662: 256-68. 4. Surrey ES, Halme J. Effect of peritoneal fluid from endometriosis patients on endometrial stromal cell proliferation in vitro. Obstet Gynecol 1990;76:792-7. 5. Olive DL, Montoya I, Riehl RM, Shenken RS. Macrophage-conditioned media enhance endometrial stromal cell proliferation in vitro. AM J OBSTET GYNECOL 1991; 164: 953-8. 6. Halme J, Becher S, Hammond MG, Raj S. Pelvic macrophages in normal and infertile women. AM J OBSTET GYNECOL 1982; 142:890-5. 7. Surrey ES, Halme J. Effect of platelet derived growth factor on endometrial stromal cells proliferation in vitro: a model for endometriosis. Fertil Steril 1991 ;56:672-9. 8. Halme J. Release of tumor necrosis factor by human peritoneal macrophages in vivo and in vitro. AM J OBSTET GYNECOL 1989;161:1718-25. 9. Van Le L, Oh ST, Anners J, Rinehart CA, Halme J. Interleukin-I inhibits growth of normal human endometrial stromal cells. Obstet Gynecol 1992 [In press). 10. Hayflick L, Moorhead PS. Serial cultivation of human diploid cell strains. Exp Cell Res 1961;25:585-621. 11. Zhang R, Wild RA, Medders D, Gunupudi SR. Effects of peritoneal macrophages from patients with endometriosis on the proliferation of endometrial carcinoma cell line ECC-1. AM J OBSTET GYNECOL 1991;165:1842-6. 12. Semer D, Reisler K, MacDonald PC, Casey ML. Responsiveness of human endometrial stromal cells to cytokines. Ann NY Acad Sci 1991;622:99-110. 13. Economos K, MacDonald PC, Casey ML. Endothelin-l gene expression and protein biosynthesis in human endometrium: potential modulator of endometrial blood flow. J Clin Endocrinol Metab 1992;74:14-9. 14. Matthews CJ, Redfern CPF, Hirst BH, Thomas EJ. Characterization of human purified epithelial and stromal cells from endometrium and endometriosis in tissue culture. Fertil Steril 1992;57:990-7. 15. Haney AF, Muscato JJ, Weinberg BJ. Peritoneal fluid cell populations in infertility patients. Ferti! Steril 1981 ;35: 696-8. 16. Halme J, Becher S, Haskill S. Altered maturation and function of peritoneal macrophages: a possible role in
1136 Hammond et al.
17. 18.
19.
20.
pathogenesis of endometriosis. AM j OBSTET GYNECOL 1987;156:783-7. Madtes DK, Malden LT, Raines EW, Ross P. Induction of transcription and secretion of TGF-alpha in activated human monocytes. Chest 1991;99:795. Rossi Mj, Tang X, Masterson Bj, Chegini N. The direct proliferative action of macrophages in human endometrial and myometrial tissue [Abstract 577]. In: Proceedings of the thirty-ninth annual meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. San Antonio: Society for Gynecologic Investigation, 1992. Gospodarowicz D, Plouet j, Malerstein B. Comparison of the ability of basic and' acidic fibroblast growth factor to stimulate proliferation of an established keratinocyte cell line: modulation of their biologic effects by heparin, TGF beta and EGF. j Cell Physiol 1990;142:325-33. 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: 2373-85.
Discussion
A. ZACUR, Baltimore, Maryland. These data are consistent with the supposition that cytokines stimulate endometrial cell growth and that macrophages and endometrial stromal cells might be responsible for releasing these growth factors. To place these results in perspective I would like to introduce this discussion with a brief review. In 1981 Haney et ai. I reported that the peritoneal fluid of infertile women with or without endometriosis contained more macrophages than the peritoneal fluid of fertile controls. On the basis of this observation it was theorized that the presence of macrophages could exert an adverse effect on fertility. This hypothesis gained credibility when it was reported that macrophages derived from peritoneal fluid could phagocytize sperm, a process that could even be enhanced when macrophages were obtained from the peritoneal fluid of women with endometriosis. 2 Also consistent with this thesis were results from a previous study that reported increased activation of macrophages in the peritoneal fluid of women with endometriosis. 3 Activation of macrophages could result in the release of substances into the peritoneal cavity, which is potentially deterimental to fertility. 4, 5 Could endometriosis serve as the attractant for these macrophages? Recent data would indicate that this question may not be answered in the affirmative. With a rabbit model neither peritoneal macrophage number nor macrophage activation could be altered by the presence of ectopic endometrium. 6 Increased numbers of macrophages in peritoneal fluid of infertile women without endometriosis, reported previously by Haney et ai., I has now been confirmed by others.7, 8 With these results in mind it is possible to hypothesize that macrophages might not only be responsible for infertility but also for the presence of endometriosis. This provocative theory, proposed by Halme et al." in 1987, suggests that the release of growth factors from macrophages could stimulate ectopic endometrial growth. Data presented in the current study extends this thesis by showing that endometrial stromal cell growth DR. HOWARD
April 1993
Am J Obstet Gynecol
may be influenced by cytokines, which could possibly be released from macrophages or endometrial stromal cells themselves. With thymidine uptake as a marker for deoxyribonucleic acid synthesis, stimulatory effects on endometrial cell growth were noted for some, but not all, factors tested but were not observed for estrogen. These results raise several questions. First, why were these particular growth factors selected for use in this study? Have concentrations of these factors been measured in peritoneal fluid of women with endometriosis? If so, is there a correlation between growth factor concentration and stage of endometriosis? Does a bal-· ance exist between factors released by stromal cells and those released by macrophages? Second, because eutopic rather than ectopic endometrial tissue was used, how can we be persuaded that endometriosis tissue would respond in a similar manner? Third, how do the authors explain a lack of response to estrogen? Increased thymidine uptake by endometrial stromal tissue after estrogen exposure was reported previously by this same research groUp.1O Finally, if time permits, I would like to ask Dr. Hammond to speculate for us on what attracts the macrophages initially and why all infertile women with peritoneal macrophages don't develop endometriosis. REFERENCES
n,
1. Haney AF, Muscato Weinbert lB. Peritoneal fluid cell populations in infertility patients. Fertil Steril 1981;35: 696-8. 2. Muscato Haney AF, Weinberg lB. Sperm phagocytosis by human peritoneal macrophages: a possible cause of infertility in endometriosis. AM j OBSTET GYNECOL 1982; 144:503-10. 3. Halme j, Becker S, Hammond MG, Raj MHG, Raj S. Increased activation of pelvic macrophages in infertile women with mild endometriosis. AM j OBSTET GYNECOL 1983;145:333-7. 4. Halme J. Release of tumor necrosis factor-a by human peritoneal macrophages in vivo and in vitro. AM j OBSTET GYNECOL 1989;161:1718-25. 5. Umesaki N, Uda S, Kawabata M, Ogita S. Significance of peritoneal macrophages on fertility in mice. AM j OBSTET GYNECOL 1992;167:261-4. 6. johnsonJV, Rozek MM, Moreno AC, Olive DL, Schenken RS. Surgically induced endometriosis does not alter peritoneal factors in the rabbit model. Fertil Steril 1991;56: 343-8. 7. Olive DL, Weinberg jB, Haney AF. Peritoneal macrophages and infertility: the association between cell number and pelvic pathology. Fertil Steril 1985;44:772-7. 8. Hill jA, Faris HMP, Schiff I, Anderson DJ. Characterization of leukocyte subpopulations in the peritoneal fluid of women with endometriosis. Fertil Steril 1988;50:216-22. 9. Halme j, Becker S, Haskill S. Altered maturation and function of peritoneal macrophages: possible role in pathogenesis of endometriosis. AM j OBSTET GYNECOL 1987;156:783-9. 10. Surrey ES, Halme J. Effect of platelet-derived growth factor on endometrial stromal cell proliferation in vitro: a model for endometriosis? Fertil Steril 1991;56:672-9.
n,
DR. MARVIN S. AMSTEY, Rochester, New York. I'm wondering if there are any controls for which other fibroblasts were used. Histologically, even with phasecontrast microscopy, these cells look similar to many other fibroblasts. Are they just generally responsive to
Chapel Hill, North Carolina, Kwangju, South Korea, and Los Angeles, California OBJECTIVE: Development of ectopic implants of endometriosis is associated with both an inflammatory response by macrophages and endometrial stromal cell proliferation. Macrophages are capable of releasing a variety of inflammatory mediators, including growth factors. To assess the impact of such factors on endometrial tissue, we have studied the effects of recombinant growth factors, fibroblast growth factor, epidermal growth factor, transforming growth factor-a, and inflammation mediators transforming growth factor-J3, and tumor necrosis factor-a on human endometrial stromal cell proliferation. STUDY DESIGN: Increasing concentrations of these compounds were added to cultures of primary, secondary, and long-term stromal cells and the cells were harvested at 24, 48, and 72 hours. RESULTS: Epidermal growth factor, transforming growth factor-a, transforming growth factor-J3, and fibroblast growth factor induced a statistically significant, dose-dependent increase in stromal cell thymidine uptake of 1.5- to fivefold. The cytokine tumor necrosis factor had no effect alone, but the combination of fibroblast growth factor and tumor necrosis factor had a synergistic effect, increasing cell proliferation 25% to 84% over fibroblast growth factor alone. CONCLUSION: The stromal cell response to a wide range of cell growth effectors and the potential of mediators like tumor necrosis factor-a to synergize suggest that such macrophage-secretory products may contribute to proliferation of endometrial implants in vivo. (AM J OBSTET GVNECOL 1993; 168: 1131-8.)
Key words: Endometriosis, cytokines, growth factors, endometrial stromal cells
Both intrauterine and ectopic endometrium are composed of two major cell types: glandular epithelium and stromal cells. The stromal cells form a framework surrounding the glands and blood vessels that are important regulators of growth and differentiation of the epithelial cells. Epidermal growth factor (EGF) binding has been demonstrated in human endometrial stroma and endometriosis implants in rats. I EGF and transforming growth factor-a (TGF-a) production was noted in endometriosis tissue in the rat. 2 These growth factors and receptors have also been recently reported in human endometriosis tissue. 3 The autocrine and paracrine effects of endogenous growth factors or the response of stromal cells to cytokines released by pelvic From the Department of Obstetrics and Gynecology, University of North Carolina School ofMedicine, a the Department of Obstetrics and Gynecology, Chonnam University Medical School,' and the Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, University of California-Los Angeles School of Medicine.' Supported in part by National Institutes of Health grant No. HD21546. Presented by invitation at the Eleventh Annual Meeting of the American Gynecological and Obstetrical Society, Hot Springs, Virginia, September 10-12, 1992. Reprint requests: Mal) G. Hammond, MD, Department of Obstetrics and Gynecology, University of North Carolina School of Medicine, CB # 7570, Chapel Hill, NC 27599. Copyright © 1993 by Mosby-Year Book, Inc. 0002-9378/93 $1.00 + .20 6/6/44824
macrophages may regulate initiation or continuation of the peritoneal endometriosis implant. 4 • 5 For several years this laboratory has investigated the role of macrophages and cytokines in the development and maintenance of endometriosis." Recently our attention has been focused on the use of endometrial stromal cells in culture as a model to study the effects of peritoneal fluid and macrophage-derived cytokines on cell proliferation'" 7-9 This study addresses the effect of a series of growth factors and inflammation mediators, including estradiol, fibroblast growth factor (FGF), EGF, TGF-a, transforming growth factor-(3 (TGF-(3), and tumor necrosis factor (TN F) alone or in combination, on stromal cell proliferation.
Material and methods Tissue preparation. As previously described:' 7 endometrial biopsy specimens were obtained from 18 regularly cycling women (mean cycle day 18.8) immediately before diagnostic laparoscopy or laparoscopic tubal sterilization. The protocol was approved by the University of North Carolina Human Rights Committee, and signed informed consent was obtained from each subject. Endometrium, obtained by aspiration, was placed immediately in Ham's F-I0 culture media (Gibco, 1131
1132
Hammond et al.
April 1993
Am J Obstet Gynecol
en "'0
10
EGF
c
--O.Sng/ml ----.A.--- S.O ng/ml _ _ sOng/ml
I/)
~Controls
CU ::J
o
.c
t:.
5
::E
a..
a
72 HR
48 HR
TIME Fig. 1. Effect of varying concentrations of EGF on tritiated thymidine incorporation by freshly explanted (primary) stromal cells in culture.
::E
a.. a w
>_I/)
~~
15
10
::ll/)
::E5
::l.c
at:. a:
..•
5
J:
C\I ......
•
0
0
First passage Sixth passage Freshly explanted
0.5 5.0 EGF (ng/ml)
50
Fig. 2. Effect of varying concentrations of EGF on tritiated thymidine incorporation by cultured stromal cells from primary, secondary, and sixth passage cultures.
Grand Island, N.Y.) supplemented with 50 J.Lg/ml gentamicin, 0.4 mg/ml of D-glucose, and 0.03% L-glutamine (Sigma, St. Louis) in ITS plus Premix (Collaborative Research, Bedford, Mass.). This media, with modification, was used throughout. Washed tissue was treated with 0.25% type I collagenase (Worthington) to disperse the cells. Mter centrifugation, the pellet was resuspended in media containing 10% calf serum. The cells were strained through a 38 J.Lm sieve to separate stroma from glands. Cells were allowed to adhere to plastic culture dishes, and the debris was aspirated. The adhered cells were rinsed with fresh media and then mechanically removed from 10-cm plates and counted. The average yield of cells was 9.0 x 106 (range 1.25 to 51 x 106 ). Cells were then diluted and plated in 96-well plates (Costar, Cambridge, Mass.), suspended in 200 J.LI of media containing 10% calf serum, and incubated at 37° C in 95% air and 5% carbon dioxide. Cells for primary and secondary culture conditions
were pipetted 20,000 cells per well, cells for tertiary culture were plated at 10,000 per well. Cells for secondary and tertiary culture were allowed to grow to confluence in 10% calf serum, trypsinized, and harvested and replated. 10 Tertiary cultures had undergone six passes before study. Endometrial stromal cell predominance (95%) was confirmed by phase contrast microscopy. Cell proliferation assays. Mter 24 hours media and dead cells from preincubated plates were aspirated, the cells washed with phosphate-buffered saline solution, and serum-free Ham's F-IO added. The appropriate growth factor was added in quadruplicate to all but 12 control wells. Cells were incubated for 24 to 72 hours. One microcurie per well of tritiated thymidine (New England Nuclear, Boston) was added to the wells at 0, 24, and 48 hours, respectively, and incubated for 24 hours. Mter thymidine incorporation cells were detached with ethylenediaminetetraacetic acid and the cellular material collected on filter disks by means of a
Hammond et al.
Volume 168, Number 4 Am J Obstet Gynecol
1133
.....J
0
II: I-
Z
0
100
U
u.
0
eft. 0
1.25
12.5
TGF~
125
0.5
5.0
50
TGFex (ng/ml)
(ng/ml)
Fig. 3. Effect of TGF-a and TGF-13 on stromal cell proliferation in primary culture as percent of control conditions.
-
20 TGFI3
UJ
- - 1.25 ng/ml --.....- 12.5 ng/ml ____ 125 ng/ml
'0
c:
CU UJ :::J
0 ..r: 10
I:::-
~
0..
U
o
~
24 HR
48 HR
; 72 HR
TIME Fig. 4. Effect of varying concentrations of TGF-13 on tritiated thymidine incorporation by primary endometrial stromal cells in culture.
semiautomatic cell harvester (Skatron, Sterling, Vir.). Radioactive tritiated incorporation was reported in absolute counts per minute. The factors investigated were recombinant human EGF, basic FGF, human recombinant TGF-Q and TGF-f3 (Collaborative Research), and recombinant human TNF-Q (Genzyme, Boston). Statistical analysis. Cumulative counts per minute for cytokine incubation, a summation of total counts per well assayed at 24, 48, and 72 hours, was compared with control wells. Significance was determined by grouped Student t tests.
Results EGF. Results of thymidine incorporation in four separate experiments confirm a statistically significant (p < 0.01) stimulatory effect of 0.5, 5.0, and 50 ng/ml concentrations in primary cell culture (Fig. 1). Maximum stimulation in separate experiments ranged from 1.5- to 4.6-fold. EGF had a stimulatory effect on secondary cells at 0.5 and 5.0 ng/ml and less effect at 50
ng/ml. Maximum stimulation ranged from 104- to 2.6fold in three experiments. Tertiary (sixth passage) cells demonstrated a significant but smaller stimulatory effect (Fig. 2) with a maximum range of 1.2- to lA-fold. Other factors. TGF-Q and TGF-j3 were studied at three concentrations in five separate experiments (Fig. 3). TGF-Q showed maximal stimulation of 1.3- to 1.7fold. TGF-j3 was stimulatory at 1.25, 12.5, and 125 ng/ml in freshly explanted cells (Fig. 4) and in secondary cells as well. Stimulation ranged from 1.7- to 4.3fold over untreated cells. FGF was investigated in five experiments. Concentrations of 0.01 to 100 ng/ml were stimulatory (Fig. 5) with maximal stimulation of 1.9- to 3.1-fold. When TNF (0.05 to 50 ng/ml) was studied, no effect or slight inhibition (71 % to 100% of control) of cell proliferation was noted in five experiments (Fig. 6). Estrogen was studied in concentrations of 1 x 10- 10 to 1 X 10- 8 mol/L in three studies; no consistent response was noted in any system. Combined factors. Attention was then turned to
1134 Hammond et al.
April 1993
Am
J Obstet Gynecol
60,000 ____ 100 ng/ml
50,000
----6-- 10 ng/ml ~
40,000
1.0ng/ml
_O.1ng/ml _O.01ng/ml
30,000
--e--
Control
20,000 10,000
04F~--------~----------~------------U 24 72 o 48 TIME (hrs) Fig. 5. Effect of varying concentrations of FGF on thymidine incorporation by stromal cells in culture.
2.2
-
2.0
TNF 0.5 ng/ml TNF 5.0 ng/ml
c: 1.8 Ql
TNF 50 ng/ml
C
E Ql
u c: 1.6 tIS
.c: c: w 1.4 1.2 1.0
0.1
1.0
10
100
FGF ng/ml Fig. 6. Enhancement of FGF stimulated tritiated thymidine incorporation by varying concentrations ofTNF.
factors in combination. The effect of FGF and TNF-o: on stromal cells was evaluated. A synergistic effect of the two factors was noted. Examples of individual experiments are presented in Fig. 6. Although no clear TNF dose-response relationship was noted, TNF stimulated FGF-dependent thymidine incorporation maximally at the lower FGF concentrations (0.1 to 1.0 ng/ml). The stimulation varied from l.25 ± 0.08- to l.84 ± 0.17fold. The increase was statistically significant in all but FGF 100 ng/ml. Comment
Our study supports the concept that endometrial stromal cells in serum-free culture can respond appropriately to many growth factors and cytokines known to be released by macrophages and demonstrated to be present in endometrial tissues. The well defined growth factors EGF, TGF-o:, TGF-/3, and FGF stimulate cell proliferation in our system by a factor of 1.5-fold to fivefold. This sensitivity persists throughout six cell passages, eliminating tissue or se-
rum contaminants as a factor. Stromal cells from both the proliferative and secretory phase responded with stimulation of cell proliferation. Other recombinant cytokines and growth factors have been studied in in vitro systems. Platelet-derived growth factor, a well characterized secretory product of activated macrophages that is a potent mitogen for fibroblasts, also stimulates stromal cell growth. 7 Interleukin1, a pleiotropic cytokine with multiple inflammatory and growth regulatory activities, was inhibitory in our model. 9 Other growth factors synthesized by macrophages have been evaluated in an endometrial cancer cell line. l l EGF and TGF-o:, homologous mitogens for a variety of cells, also stimulated these cells. In addition to cell proliferative effects, these factors have other effects on endometrial stromal cells. TNF appears to augment the adhesion of stromal cells to mesothelial cells in culture!! and to stimulate the synthesis of the cytokine interleukin-6 messenger ribonucleic acid and protein and its release into the medium.!2 TGF-/3, in addition to stimulating cell proliferation in
Volume 168, Number 4 Am J Obstet Gynecol
our system, also increased the propro-endothelin-l messenger ribonucleic acid levels in stromal cells. I3 Endothelin-l is postulated to regulate blood flow to the endometrium and might affect implant vasculature as well. Matthews et al. 14 have characterized cultured human epithelial and stromal cells from normal endometrium and endometriosis and report close structural similarity. This supports the suitability of our system as a model for endometriosis derived stromal cells. Two areas of impact should be considered regarding the effect of growth factors and cytokines on endometriosis. The first is the effect of peritoneal macrophages and peritoneal fluid on implantation and growth of endometrial tissue arriving in the pelvis during retrograde menstruation, and the second is autocrine and paracrine effects of the growth factors produced by stromal cells themselves. The role of peritoneal fluid components and macrophage-derived secretory products in the development of endometriosis has been studied for several years. Peritoneal fluid from women with endometriosis is increased in volume and contains a higher concentration of macrophages than it does in normal women. 6 • 15 Macrophages from this fluid secrete increased quantities of cytokines, in particular TNF. 16. 17 Peritoneal fluid from endometriosis patients has been demonstrated to stimulate stromal cell proliferation in vitro" Macrophage-conditioned media has been demonstrated to stimulate proliferation of an endometrial cancer cell line" and mouse endometrial stroma cells in vitro.5 A potential direct effect of macrophages usually present in endometrial tissue has also been shown by coculture. 18 FGF acts as a competence factor in the fibroblast. This confers on the cell the ability to respond to other mitogens such as EGF. Modulation of the ability of j3-FGF to stimulate keratinocyte cell proliferation has already been demonstrated. A synergistic effect of j3-FGF and EGF was noted with threefold to fourfold higher cell numbers.19 In our system FGF acts as a mitogen stimulating cell proliferation. The production of TNF -IX, a cytokine that has similar biologic activity to interleukin-l, is increased in activated macrophages isolated from the peritoneal cavity of women with endometriosis." It had no effect on cell proliferation in our system, but increased synergistically the stimulatory effect of lower concentrations of FGF. TNF would seem to act as a competence factor in increasing the cells' response to FGF. This interaction of a stromal cell growth factor and a macrophage inflammation mediator may have significance for understanding the interaction of these systems in women with endometriosis. In addition to peritoneal fluid growth factors, endometrial implants are also exposed to endogenous
Hammond et al.
1135
factors. Immunohistochemical studies have revealed the presence ofEGF, platelet-derived growth factor, TGF-IX, and EGF and platelet-derived growth factor receptor in human endometrium and induced endometrial implants in the rat. I. 2. 20 EGF stimulates the proliferation of many cell types, including smooth muscle cells and human uterus. Activity is mediated through a specific high-affinity surface receptor. TGF-IX binds to the same receptor with similar effects. We conclude that growth factors and cytokines released by peritoneal macrophages or endometrial stromal cells in endometrial tissue may promote implantation and proliferation of ectopic endometrium in the peritoneal cavity. REFERENCES 1. Chegini N, Rao Cv, Wakim N, Sanfilippo J. Binding of 125I-epidermal growth factor in human uterus. Cell Tissue Res 1986;246:543-8. 2. SimmsJS, Chegini N, Williams RS, Rossi AM, Dunn WAJr. Identification of epidermal growth factor, transforming growth factor-alpha and epidermal growth factor receptor in surgically induced endometriosis in the rat. Obstet Gynecol 1991;78:850-7. 3. Melega C, Balducci M, Bulletti C, Galassi A, Jasonni VM, Flamigni C. Tissue factors influencing growth and maintenance of endometriosis. Ann N Y Acad Sci 1991;662: 256-68. 4. Surrey ES, Halme J. Effect of peritoneal fluid from endometriosis patients on endometrial stromal cell proliferation in vitro. Obstet Gynecol 1990;76:792-7. 5. Olive DL, Montoya I, Riehl RM, Shenken RS. Macrophage-conditioned media enhance endometrial stromal cell proliferation in vitro. AM J OBSTET GYNECOL 1991; 164: 953-8. 6. Halme J, Becher S, Hammond MG, Raj S. Pelvic macrophages in normal and infertile women. AM J OBSTET GYNECOL 1982; 142:890-5. 7. Surrey ES, Halme J. Effect of platelet derived growth factor on endometrial stromal cells proliferation in vitro: a model for endometriosis. Fertil Steril 1991 ;56:672-9. 8. Halme J. Release of tumor necrosis factor by human peritoneal macrophages in vivo and in vitro. AM J OBSTET GYNECOL 1989;161:1718-25. 9. Van Le L, Oh ST, Anners J, Rinehart CA, Halme J. Interleukin-I inhibits growth of normal human endometrial stromal cells. Obstet Gynecol 1992 [In press). 10. Hayflick L, Moorhead PS. Serial cultivation of human diploid cell strains. Exp Cell Res 1961;25:585-621. 11. Zhang R, Wild RA, Medders D, Gunupudi SR. Effects of peritoneal macrophages from patients with endometriosis on the proliferation of endometrial carcinoma cell line ECC-1. AM J OBSTET GYNECOL 1991;165:1842-6. 12. Semer D, Reisler K, MacDonald PC, Casey ML. Responsiveness of human endometrial stromal cells to cytokines. Ann NY Acad Sci 1991;622:99-110. 13. Economos K, MacDonald PC, Casey ML. Endothelin-l gene expression and protein biosynthesis in human endometrium: potential modulator of endometrial blood flow. J Clin Endocrinol Metab 1992;74:14-9. 14. Matthews CJ, Redfern CPF, Hirst BH, Thomas EJ. Characterization of human purified epithelial and stromal cells from endometrium and endometriosis in tissue culture. Fertil Steril 1992;57:990-7. 15. Haney AF, Muscato JJ, Weinberg BJ. Peritoneal fluid cell populations in infertility patients. Ferti! Steril 1981 ;35: 696-8. 16. Halme J, Becher S, Haskill S. Altered maturation and function of peritoneal macrophages: a possible role in
1136 Hammond et al.
17. 18.
19.
20.
pathogenesis of endometriosis. AM j OBSTET GYNECOL 1987;156:783-7. Madtes DK, Malden LT, Raines EW, Ross P. Induction of transcription and secretion of TGF-alpha in activated human monocytes. Chest 1991;99:795. Rossi Mj, Tang X, Masterson Bj, Chegini N. The direct proliferative action of macrophages in human endometrial and myometrial tissue [Abstract 577]. In: Proceedings of the thirty-ninth annual meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. San Antonio: Society for Gynecologic Investigation, 1992. Gospodarowicz D, Plouet j, Malerstein B. Comparison of the ability of basic and' acidic fibroblast growth factor to stimulate proliferation of an established keratinocyte cell line: modulation of their biologic effects by heparin, TGF beta and EGF. j Cell Physiol 1990;142:325-33. 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: 2373-85.
Discussion
A. ZACUR, Baltimore, Maryland. These data are consistent with the supposition that cytokines stimulate endometrial cell growth and that macrophages and endometrial stromal cells might be responsible for releasing these growth factors. To place these results in perspective I would like to introduce this discussion with a brief review. In 1981 Haney et ai. I reported that the peritoneal fluid of infertile women with or without endometriosis contained more macrophages than the peritoneal fluid of fertile controls. On the basis of this observation it was theorized that the presence of macrophages could exert an adverse effect on fertility. This hypothesis gained credibility when it was reported that macrophages derived from peritoneal fluid could phagocytize sperm, a process that could even be enhanced when macrophages were obtained from the peritoneal fluid of women with endometriosis. 2 Also consistent with this thesis were results from a previous study that reported increased activation of macrophages in the peritoneal fluid of women with endometriosis. 3 Activation of macrophages could result in the release of substances into the peritoneal cavity, which is potentially deterimental to fertility. 4, 5 Could endometriosis serve as the attractant for these macrophages? Recent data would indicate that this question may not be answered in the affirmative. With a rabbit model neither peritoneal macrophage number nor macrophage activation could be altered by the presence of ectopic endometrium. 6 Increased numbers of macrophages in peritoneal fluid of infertile women without endometriosis, reported previously by Haney et ai., I has now been confirmed by others.7, 8 With these results in mind it is possible to hypothesize that macrophages might not only be responsible for infertility but also for the presence of endometriosis. This provocative theory, proposed by Halme et al." in 1987, suggests that the release of growth factors from macrophages could stimulate ectopic endometrial growth. Data presented in the current study extends this thesis by showing that endometrial stromal cell growth DR. HOWARD
April 1993
Am J Obstet Gynecol
may be influenced by cytokines, which could possibly be released from macrophages or endometrial stromal cells themselves. With thymidine uptake as a marker for deoxyribonucleic acid synthesis, stimulatory effects on endometrial cell growth were noted for some, but not all, factors tested but were not observed for estrogen. These results raise several questions. First, why were these particular growth factors selected for use in this study? Have concentrations of these factors been measured in peritoneal fluid of women with endometriosis? If so, is there a correlation between growth factor concentration and stage of endometriosis? Does a bal-· ance exist between factors released by stromal cells and those released by macrophages? Second, because eutopic rather than ectopic endometrial tissue was used, how can we be persuaded that endometriosis tissue would respond in a similar manner? Third, how do the authors explain a lack of response to estrogen? Increased thymidine uptake by endometrial stromal tissue after estrogen exposure was reported previously by this same research groUp.1O Finally, if time permits, I would like to ask Dr. Hammond to speculate for us on what attracts the macrophages initially and why all infertile women with peritoneal macrophages don't develop endometriosis. REFERENCES
n,
1. Haney AF, Muscato Weinbert lB. Peritoneal fluid cell populations in infertility patients. Fertil Steril 1981;35: 696-8. 2. Muscato Haney AF, Weinberg lB. Sperm phagocytosis by human peritoneal macrophages: a possible cause of infertility in endometriosis. AM j OBSTET GYNECOL 1982; 144:503-10. 3. Halme j, Becker S, Hammond MG, Raj MHG, Raj S. Increased activation of pelvic macrophages in infertile women with mild endometriosis. AM j OBSTET GYNECOL 1983;145:333-7. 4. Halme J. Release of tumor necrosis factor-a by human peritoneal macrophages in vivo and in vitro. AM j OBSTET GYNECOL 1989;161:1718-25. 5. Umesaki N, Uda S, Kawabata M, Ogita S. Significance of peritoneal macrophages on fertility in mice. AM j OBSTET GYNECOL 1992;167:261-4. 6. johnsonJV, Rozek MM, Moreno AC, Olive DL, Schenken RS. Surgically induced endometriosis does not alter peritoneal factors in the rabbit model. Fertil Steril 1991;56: 343-8. 7. Olive DL, Weinberg jB, Haney AF. Peritoneal macrophages and infertility: the association between cell number and pelvic pathology. Fertil Steril 1985;44:772-7. 8. Hill jA, Faris HMP, Schiff I, Anderson DJ. Characterization of leukocyte subpopulations in the peritoneal fluid of women with endometriosis. Fertil Steril 1988;50:216-22. 9. Halme j, Becker S, Haskill S. Altered maturation and function of peritoneal macrophages: possible role in pathogenesis of endometriosis. AM j OBSTET GYNECOL 1987;156:783-9. 10. Surrey ES, Halme J. Effect of platelet-derived growth factor on endometrial stromal cell proliferation in vitro: a model for endometriosis? Fertil Steril 1991;56:672-9.
n,
DR. MARVIN S. AMSTEY, Rochester, New York. I'm wondering if there are any controls for which other fibroblasts were used. Histologically, even with phasecontrast microscopy, these cells look similar to many other fibroblasts. Are they just generally responsive to