Jul .a 8.»..i: .. n ‘ . an. me.....w.m¢ Man. m... 1-Way} 5 4v... 33% .w u.. 5...? O E» .. .. ”93.3 a: .2 1 z... . CL . .11....u........ ,3 :33. z a E 57.2 v:y' q: 3 I. , . . 9W4... : 2.5:... um . , . . . can It -75... 1.39; 1 » , . . . :9. 2. .u...un.:m. 1m... . . . 222?...ufsa... .1335 .21... . . “JP“; 43 r."..:. 9‘31‘1 “Kn"! . . . . . . . , .33.. L7... 2... . is}... , 2.... 2.x...5.z.黣...s. ..... it ...... 3.1 . t .. . 13 . .1. . 2 V .3 t 2. .... .. .5. I... . .3: .1319: 0.... I... . . 3.161,. . . , . . . in»... lii’.‘ ’h 1 .W... R43...» I . Luke. It .1." x. Q:| & 3 i ...H..H.u..¢.mv...q.§n.. ,2... - .33.... ~ {vixhav h‘xn *3; at... .s I... . .. roam.) ': \ . .. . f. x: I a ch .31.. 3.5;; O)...‘ 2 V‘of! .0!- $.57} E :5. . . . 3.2. r. . . . . a . . . , . . H... w 4499.5? . . . h. .. 33?: 3;! :26»; mag...) ‘ 2.13 v . 5:13:35: {It'd—“34:1! .‘l t (43.1». .0. 1 ‘12.... u 0 filth sllv 3 £19):lb\li 3v}! All-t )1!!.€s:!..0. hi... «i I Li) .1! HL/ 599:. .h .6 \vzvfi: u I) .)..!I.I| . 3.1:)... ”tr. 2: . Ed . v )u . L s Intri 11.3.}...7‘. . . 1.. Illil'flmil‘ljrT"?ll‘l’fliflililiiilifllil’iififil 1 93 01020 1725 This is to certify that the dissertation entitled THE m OF MY E‘P CELL MOB IN RELATIW '10 mm: RHIIIATICN BY m AND WIN presented by StfiqingWang has been accepted towards fulfillment of the requirements for Ph-D- degree in mysiology Sandra Z. Haslam, Ph.D. Major professor Date 8124/1994 MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 ,i. LIBRARY M'Chigan State Unlverslty PLACE II RETURN BOXtonmovombehockomm yourncocd. 1’0 AVOID FINES Mum on arm-date”. DATE DUE DATE DUE DATE DUE T [j Hm MSU I: An Namath!- Action/Equal Oppomnky Institution THE ROLE OF MAMMARY EPITHELIAL-STROMAL CELL INTERACTIONS IN RELATION TO GROWTH REGULATION BY ESTROGEN AND PROGESTIN By Shiqing Wang A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for a degree of DOCTOR OF PHILOSOPHY Department of Physiology 1994 ABSTRACT THE ROLE OF MAMMARY EPITHELIAL-STROMAL CELL INTERACTIONS IN RELATION TO GROWTH REGULATION BY ESTROGEN AND PROGESTIN By Shiqing Wang Both m and M2 approaches have been used in this study to investigate how estrogen (E), progestin (P), epidermal growth factor (EGF) and insulin like growth factor-I (IGF-I) interact to promote mammary growth in relation to their interactions with epithelial and stromal components of the mammary gland. The in vivo study focused on local vs. systemic effects of P and E on mammary growth and their relative effects on epithelial and stromal tissue. Different mechanisms of P action in the two types of mammary tissue have been revealed. The growth effects of P in epithelial cells seem to be locally mediated, whereas P stimulation of DNA synthesis in stromal cells appears to be systemically mediated. The growth effects of E on epithelial and stromal cells also appear to be systemically mediated. E+P synergistically stimulated epithelial cell proliferation via E inducible P receptor (PR), and this effect appears to be locally mediated. An m serum-free monolayer culture system was developed for culturing epithelial, fibroblast and mixed cultures, so that the effects of E or P can be examined in vitro free of in vivo systemic influences. P stimulated proliferation of epithelial cells but had no effect on fibroblasts in cell cultures from nulliparous mice. E was able to induce PR in epithelial cells, however, E did not stimulate proliferation of epithelial cells or fibroblasts. These results support the i_r_I_v_iv_o observations, and indicate that P has a direct effect on epithelial cell proliferation. The lack of stimulatory effect of E on epithelial cells and fibroblasts, or P on fibroblasts in serum-free media suggests that additional factors are required. EGF or IGF-I stimulated proliferation of epithelial cells but not fibroblasts, and no interaction with E, P or E+P was observed. From these in vitro studies, the influence of epithelial-stromal cell interactions on hormonal responsiveness has also been identified, such that E+P stimulated proliferatiOn of fibroblasts in mixed cultures whereas no stimulaton effect was observed in fibroblast-enriched cultures. In contrast to cells from nulliparous mice, E or P or E+P did not stimulate proliferation of epithelial, fibroblast or mixed cultures fiom pregnant mice. This latter finding indicates that the developmental state of the gland can modulate hormonal responsiveness. Acknowledgments I would like to thank Dr. Sandra Z. Haslam for her excellent guidance throughout my graduate studies. I also want to express my gratitude to the members in my graduate committee: Dr. Jump, Dr. Kaufman, Dr. Welsch, and Dr. Zipser for their support whenever I needed it. I am indebted to Dr. Welsch for his generosity in letting me use his laboratory. I would also like to give special thanks to Dr. Jump and Dr. Zipser for kindly allowing me to use their instruments. Finally, I want to say that I feel fortunate to have been able to work with graduate students and technicians in Dr. Haslam's laboratory whose knowledge and friendship supported me and accompanied me all along the way. iv Table of Contents List of Tables vii List of Figures viii List of Abbreviations x Chapter 1: Literature review A. Introduction 1 B. Mammary gland development 2 C. Hormonal regulation in mammary gland development 3 D. Epithelial-stromal interactions in mammary gland development 7 E. Hormonal regulation and breast cancer 11 F. Epithelial-stromal interactions in breast cancer 13 G. The objectives of current studies 14 Chapter 2: The efi‘ects of estrogen and progesterone on normal mouse mammary growth and the relative responses of epithelial vs. stromal tissues in vivo. Introduction 16 Materials and Methods 17 Results 18 Discussion 38 Chapter 3: Serum-free primary cultures of mammary epithelial cells, fibroblasts and mixed cultures containing both epithelial cells and fibroblasts. Introduction 41 Materials and Methods 42 Results 47 Discussion 66 Chapter 4: The effects of E, P and growth factors on epithelial, fibroblast and mixed cell cultures in serum-free monolayer culture conditions. Introduction 76 Materials and Methods 77 Results 81 Discussion 98 Summary and Conclusions 106 References 109 Table 1. Table 2. Table 3. Table 4. Table 5. List of Tables Eflect of varying dose estrogen implants on mammary PR concentration. Differential effects of R5020, E+R5020, EGF, and IGF-I on epithelial cells and fibroblasts in mixed cell cultures from nulliparous mice. Differential effects of EGF and IGF-1 on epithelial cells or fibroblasts in mixed cell cultures from pregnant mice. Specific 3H-estradiol and 3H-R5020 binding sites in mixed and epithelial cell cultures from nulliparous and mid-pregnant Mice. Regulation of 3H-R5020 binding sites in mixed and epithelial cell cultures from nulliparous and mid-pregnant mice. vii 27 87 94 95 96 Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16. List of Figures Effect of progestin R5020 and E on mammary gland morphology. 20 Effect of low R5020 dose plus E on mammary gland morphology. 23 Effect of high R5020 dose plus E on mammary gland morphology. 25 Effect of progestins on mammary gland epithelial cell DNA synthesis. 29 Effect of progestins plus E on epithelial cell DNA synthesis. 31 Effect of RU486 on epithelial cell DNA synthesis. 33 Effect of RU486 on stromal cell DNA synthesis. 35 Effect of progestins plus B on stromal cell DNA synthesis. 37 MTT assay standard curves showing the linear relation between cell number and absorbance. 46 Effects of BSA, transferrin or fetuin on growth of mixed cultures from nulliparous mice and mid-pregnant mice. 50 Effects of fetuin dose on mixed culture growth. 52 The effect of fetuin on the growth of mixed cultures on collagen or plastic. 54 The proportional distribution of epithelial cells vs. fibroblasts in mixed cultures. 56 Photomicrographs of cultured mammary cells. 58 Effects of medium composition and substratum on growth of epithelial cultures from nulliparous mice. 61 Effects of medium composition and substratum on the growth viii Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. of epithelial cultures from mid-pregnant mice. 63 Comparison of the growth of mixed cultures and epithelial cultures from nulliparous and mid-pregnant mice in minimally supplemented medium. 65 Effects of BSA, transferrin or fetuin on the growth of fibroblast cultures from nulliparous or mid-pregnant mice. 68 Effects of fetuin dose on fibroblast cultures growth. 70 Effects of hydrocortisone on fibroblast cultures growth. 72 Effects of E, R5020, and EGF alone or in different combinations on cell proliferation of mixed, epithelial cell and fibroblast ulturesfrom nulliparous mice. 84 Effects of IGF-I and EGF alone or in different combinations with E and R5020 on cell proliferation of mixed, epithelial and fibroblast cell cultures from nulliparous mice. 86 Effects of E, R5020, and EGF alone or in different combinations on cell proliferation of mixed, epithelial cell and fibroblast cultures fi'om pregnant mice. 90 Effects of IGF—I and EGF alone or in different combinations with E, R5020 on cell proliferation of mixed and epithelial cells from pregnant mice. 92 ix BM BSA EGF ER HBSS IGF-I PR Prl R5020 RU486 List of Abbreviations basal medium bovine serum albumin estrogen epidermal growth factor estrogen receptor fetuin Hank's balanced salt solution hydrocortisone insulin-like growth factor I insulin labeling index medroxyprogesterone acetate 3-(4,5- dimethylthiazol-Z-yl)-2,5-diphenyltetrazolium bromide progestin progesterone progesterone receptor prolactin R5020 promogestone l l B—(4-dimethylamino-phenyl) l- l 7B-hydroxy- 17a- (prop-lynyl)-estra-4,9-diene-3 -one X transferrin Chapter 1. Literature Review A. Introduction The mammary gland is a suitable system for studying the mechanisms of growth and differentiation, because of its cyclic pattern of morphological and functional development during the estrous cycle, pregnancy and lactation. Furthermore, the mammary gland is an organ where the potential for cancer is very high. In recent years, about 1 out of 9 North American women develop breast cancer (Ries et al., 1991). In this context, the understanding of mammary growth control is highly relevant to prevention and therapy in breast cancer. The mammary gland is composed of epithelium, and adipose and fibrous connective tissue generated by fibroblasts. Adipocytes and fibroblasts are derived from embryonic mesenchyme and are referred to as stromal cells. Epithelial tissue forms ducts which lie in the adipose tissue. Fibroblasts invest the ducts and separate epithelial cells fiom adipocytes. The ovarian hormones estrogen (E) and progesterone (Pg) are required for normal growth and development. Both epithelial and stromal tissues contain estrogen (ER) and progesterone receptors (PR) and are responsive to E and Pg. During the menstrual cycle, pregnancy and lactation, the mammary gland undergoes cyclic changes of epithelial proliferation and degeneration, concomitant with the depletion and replenishment of stromal tissue. Morphological changes occurring during different physiological stages are the result of epithelial-stromal cell interactions and hormonal regulation (Sakakura etal., 1976; Daniel et al., 1984; Haslam, 1991). The abnormal cell proliferation which occurs in breast cancer is likely to involve changes of normal epithelial- stromal cell interactions and loss of normal hormonal control. The purpose of this dissertation research is to understand how mammary epithelial-stromal cell interaction influences E- and Pg- induced cell proliferation. Increasing our 1 2 knowledge of these mechanisms will help us to understand abnormal epithelial- stromal cell interactions and hormonal regulation in breast cancer. The mouse mammary gland is a model system which has been extensively used by researchers in studies of mammary gland biology and breast cancer. This model system will also be used in my dissertation research. Therefore, the main focus of this chapter is a review of studies on mouse mammary development and growth regulation. Growth control in breast cancer will also be discussed to elucidate our recent understanding of abnormal changes in breast cancer in relation to normal development and some critical factors involved in the process of mammary neoplasia. B. Mammy gland development Mammary gland development can be divided into the following stages: embryonic, prepubertal, sexually mature, pregnant, lactating and involution. In the embryonic stage, mammary epithelium derived from epidermis, forms an enlargement on both sides of the trunk. There is no apparent epithelial growth during the first 11 days of fetal life. A rapid mammary epithelial proliferation happens at late fetal life and the increased epithelial tissue penetrates into fat pad precursor stromal tissue. At birth, the mammary gland contains a few short branching ducts. From birth to puberty, development of the mammary gland occurs in two distinct phases. The first phase is isometric growth (the increase of mammary gland is in proportion to the rest of the body) in which connective tissue increases are accompanied by deposition of fat. At 3-4 weeks of age, the mammary gland starts allometric growth (the growth rate is higher than that of the body). At this stage, large club shaped end buds appear at the duct tips and these structures have been demonstrated to be the growing point of the ducts. The 3 growth is characterized by ductal elongation with the epithelium invading the stromal fat pad. At the time of sexual maturity, the mammary ducts extend throughout the fat pad and also give rise to side branches. The pattern of the mammary gland by now is a characteristic tree-shaped branching ductile system. Cyclic variation in mammary epithelial cell proliferation and regression occurs during the estrus cycle. During pregnancy, the mammary gland undergoes extensive side branching and lobuloalveolar development. The interductal spaces are gradually occupied by growing epithelial tissue. As pregnancy approaches term, the epithelium differentiates towards a secretary epithelium. After parturition, active lactation begins. Proliferative activity in lactating mammary gland is very low. After lactation, the epithelium starts to degenerate, fat tissue reappears, and the mammary gland returns to the state of a branched ductal system similar to the mature nulliparous mouse. Thereafter the mammary gland reassurnes the cyclic changes during the estrus cycle. C. Hormonal regulation in mammary gland development 1. Hormonal regglation in vivo The female pattern of fetal mammary development is a dominant phenotype and does not require hormonal stimulation. Postnatal mammary development depends on the ovarian hormones E and Pg. In female mice, ER can be detected at 3 days of age (Haslam, 1989). The mammary gland starts to grow at 3 weeks of age, when the ovaries begin to function and the mammary gland first becomes responsive to E (Haslam, 1989). The initial growth is end bud epithelial proliferation and ductal elongation. Ovariectomy causes regression of end buds and cessation of ductal growth within a few days, and this process can be reversed 4 by administration of E. Daniel et al. (1987) found that in ovariectomized animals, E stimulated DNA synthesis in the cap cells of the end buds. Using steroid autoradiography, they determined that the binding of E is not in cap cells but in the adjacent epithelial cells and in stromal cells. From these results, they proposed that E stimulated growth involves an indirect mechanism. At 7 weeks of age the mammary gland acquires proliferative responsiveness to Pg (Haslam, 1988). Coincidentally, it is also at this age that E-inducible PR first appear in the epithelium. Under the influence of E and Pg, some side- branching and alveolar development occur. In the adult, the mammary gland undergoes cyclic changes during the estrus cycle as the result of cyclical secretion of E and Pg from the ovaries. Pg was found to stimulate ductal cells as well as end buds (Breciani, 1968). However the real physiological function of Pg is probably to regulate ductal cell proliferation. The effect of E on ductal branching is probably via an indirect effect of stimulation of PR in ductal epithelial tissue. E and Pg are also major hormones involved in mammary gland development during pregnancy. E and Pg not only stimulate lobuloalveolar formation, but also are required for the maintenance of alveoli. Ovariectomy causes alveolar regression during pregnancy (Nandi 1958). Other hormones such as thyroid hormone, prolactin, growth hormone and adrenal steroids are also involved in mammary gland development. Thyroid hormone and prolactin stimulate alveoli formation (Vonderhaar & Greco, 1979). Adrenal steroids are involved in lobule formation (Banerjee, 1976). In triply operated mice (ovariectomy, hypophysectomy and adrenalectomy), lobuloalveolar development can be induced by administration of E, Pg, deoxycorticosterone, prolactin and growth hormone (N andi, 1958; Nandi & Bern, 1960). There is also evidence to suggest that EGF is involved in mammary growth and E and P regulated events. Both endbuds and ductal cells are responsive to 5 EGF. Removal of the major source of EGF by sialodenectomy results in reduced mammary gland growth (Sheffield & Welsch, 1987). Sialadenectomy during pregnancy results in a smaller mammary gland and less milk production after parturition. Consequently, there is a higher number of pup deaths within the first week afier birth. EGF replacement during pregnancy in sialadenectomized mice can restore mammary growth, milk production, and increase pup survival (Okamoto & Oka, 1984; Oka et al., 1988). During pregnancy, when there are sustained elevated levels of E and Pg, EGF concentration in the salivary glands and in the plasma is high, and EGF receptor concentration in the mammary gland increases (Ances, 1973; Kurach & Oka, 1985; Edery et al., 1985). These observations suggest a possible relationship among E, Pg and EGF. Direct evidence that E and Pg interact with EGF came from the finding that the treatment of ovariectomized mice with E and Pg increased EGF content in the salivary gland and EGF receptor concentration in the mammary gland (V onderhaar, 1984). In addition to systemic growth factors which seems to be involved in E and Pg mediated effects in vivo local mammary-derived growth factors have been demonstrated from E and Pg treated mammary glands (Enami et al., 1983; Vonderhaar, 1984). It is not clear, however, whether the production of the locally derived growth factors are the direct effect of E and Pg on the mammary gland or if they are also produced through systemic processes. 2. Hormonal regulation in vitro The in vivo mammary gland developmental cycle can be mimicked in vitro by culturing whole mammary glands in serum-flee hormone supplemented media. However, in vivo priming with E+Pg is required. Mammary glands fi'om prepubertal mice pretreated in vivo with E+Pg for 9 days can exhibit full lobuloalveolar development when cultured in media supplemented with 6 hydrocortisone, insulin, aldosterone and prolactin for 6 days (Ichinose & Nandi 1964; 1966). Withdrawal of all the hormones from the culture media except insulin, results in regression of alveoli similar to mammary gland involution afier lactation (Banerjee et al., 1976). A second round of lobuloalveolar development can be achieved by including EGF in the hormone cocktail capable of inducing the first round of growth (Tonelli & Sorof, 1980). It is possible that EGF carried over fiom in vivo participated in the first round lobuloalveolar formation in vitro since the i_n_viv_o E+Pg pretreatment can increase EGF binding in mammary gland and EGF concentration in submaxillary gland (Vonderhaar, 1984). In cell culture studies, when mammary fibroblasts and epithelial cells have been co-cultured in serum-containing media, E has been shown to stimulate epithelial cell DNA synthesis (McGrath, 1983; Haslam, 1986). No effects of mammogenic hormones on growth have been demonstrated in epithelium-enriched cell cultures in serum supplemented media ( Nandi et al.,-1981). Serum contains unidentified growth stimulatory and inhibitory factors which possibly mask or compensate for hormonal effects. Using a serum-free culture media and growing pure epithelial cells within collagen gels, Nandi and colleagues (1985) found that Pg, prolactin and EGF stimulated proliferation of mammary epithelial cells. Pg plus prolactin had a synergistic effect. EGF at a low dose (1 ng/ml) exhibited an additive effect with Pg plus prolactin. IGF-I has also been found to synergize with EGF in the stimulation of epithelial cell proliferation. Interestingly, no direct growth stimulatory effect has been observed with E. However, E did increase PR concentration. The effects of E, Pg, EGF and IGF-I on mammary fibroblasts and co-cultured epithelial cells and fibroblasts have not been examined. 7 D. Epithelial-stromal cell interactions and mammary gland development Epithelial-stromal cell interactions are important in embryonic organogenesis of the mammary gland and postnatal mammary gland development (Sakakura et al., 1976; Daniel et al., 1984). The embryonic organogenesis is hormone independent. The interactions of the epithelium and embryonic mesenchyme determine the morphological and firnctional organogenesis (Sakakura, 1987). In postnatal mammary gland development, epithelial-stromal cell interactions influence the responses to hormones and growth factors (Haslam & Counterrnan, 1991). Two possible mechanisms have been proposed to describe the effects of stromal cells on epithelial cell hormone responsiveness and growth (Haslam, 1986). One is via substrate effects in which fibroblasts provide epithelium with appropriate substratum for its growth and function. Another is that fibroblasts produce soluble factors which affect epithelial cell growth. The interactions between epithelial cells and fibroblasts are believed to be reciprocal and epithelial cells can also influence stromal cells (Haslam, 1986). 1. In vivo studies When fetal mammary epithelium is combined with fetal mammary mesenchyme or salivary mesenchyme, the subsequent pattern of morphogenesis is determined by the source of mesenchyme (Sakakura et al., 1976). This evidence indicates that embryonic mesenchymal-epithelial cell interactions are critical for normal mammary gland morphogenesis. Mammary mesenchyme includes fibroblast and fat pad precursors. These two types of mesenchyme have been shown to have different influences on mammary epithelium. When transplanted into adult mammary gland, the fibroblast mesenchyme causes ductal hyperplasia, whereas the fat pad mesenchyme induces normal epithelial architecture (Sakakura et al., 1982). Irnaguma et al. (1988) 8 investigated the mechanisms underlying these different effects. They found that fat pad mesenchyme produces larninin and heparin sulfate proteoglycan which are the components of epithelial basement membranes. They hypothesized that the normal basement membrane components may provide a suitable environment for normal epithelial morphogenesis. Fibroblast mesenchyme can produce tenascin. Tenascin is an extracellular matrix glycoprotein containing 13 EGF-like repeats and has been suggested to function as an autocrine or paracrine grth factor (Pearson et al., 1988). So tenascin might be a molecule responsible for the hyperplastic response induced by fibroblast mesenchyme. Postnatal epithelial-stromal cell interactions are also important in mammary gland morphogenesis. During the initial postnatal growth, primary ducts elongate and penetrate the fat pad. The end bud is separated from adipose tissue only by the basement membrane, whereas the subtending duct is also invested by fibroblasts. When ducts reach the limit of the fat pad, elongation ceases, and the tip of the ducts are now surrounded by fibroblasts. Transplantation experiments have demonstrated that adipose stroma is an absolute requirement for ductal growth. Interestingly, adipose stroma does not need to be mammary specific. Upon transplantation to other non mammary adipose tissue, mammary epithelial ducts can also develop. However, transplantation to any other types of stromal matrix failed to support mammary epithelial growth. Daniel et al. (1984) transplanted mammary epithelial tissue embedded in collagen gel to mammary adipose stroma. They found spike-shaped outgrowths of epithelial cells within the gel, however, normal end buds formed when the epithelial spike reached the adipose compartment. Thus the role of adipose stroma seems to be related to normal morphology of ductal elongation. During pregnancy, the epithelial component of the gland increases concomitant with a depletion of the adipose tissue. After lactation, there is a loss of secretary 9 epithelium and lobuloalveolar degeneration. Adipose tissues replenish the resulting space. Thus the adipose tissue undergoes cyclic changes in accordance with epithelial function. In addition, for adequate nutritional supply in rapid growing mammary epithelium during pregnancy and for secretary activity during lactation, the glycogen synthesis and lipogenic rates are increased in adipocytes (Bartley et al., 1981). These effects can only happen in the presence of epithelium indicating interactive phenomena (Bartley et al., 1981). Less is known about fibroblast stroma. It is a minor tissue component compared to adipose stroma. However, it is the tissue that is in direct contact with and invests the epithelial ducts. Its function seems related to epithelial ductal spacing and stabilization. Fibroblast-derived extracellular manix components, particularly glycosarninoglycans, have been shown to play an important role in regulation of ductal branching (Bemfield et al., 1984). Thus fibroblast stroma may participate in the control of the branching ductal morphogenesis of the mammary gland. Hormonal control of mammary growth is also influenced by epithelial-stromal cell interactions. ER are present equally in both the epithelium and stroma of nulliparous mice. ER are decreased in the epithelium during pregnancy, and in the stroma during lactation. Twenty percent of PR are located in stromal cells and are E-independent; eighty percent are in epithelial tissue and can be regulated by E. PR are also decreased in the epithelium during pregnancy. The significance of these changes in receptor levels in epithelium and stroma during different stages of mammary gland development is not known at present. However these differences in receptor distribution suggest different hormonal control programs in the two types of tissues. Recently, it has been demonstrated that mammary epithelium from 3 week old mice prematurely produce E-dependent PR when combined with mature 10 week old stroma (Haslam & Counterman, 1991). This observation 10 indicates that epithelial-stromal cell interactions play an important role in modulating hormonal responsiveness of mammary tissue. EGF receptors have been detected in the outer most cell layer of end buds and in stromal cells adjacent to growing ducts (Coleman et al., 1988). EGF can stimulate end bud cell proliferation. A possible role of EGF on stromal cells may be to modify extracellular components which in turn may influence epithelial proliferation (Haslam, 1991). Recent studies have shown that EGF receptors on epithelial cells can be modulated by E and P (Haslam et al., 1992), and EGF together with E and P synergistically stimulated epithelial DNA synthesis (Haslam et al., 1993). Whether the stromal tissue contributed to these hormone- and growth factor- regulated effects needs further investigation. fin vitro studies m studies have revealed that fibroblasts have profound influences on epithelial responses to hormonal regulation. Using a co—culture system, McGrath has shown that epithelial cells stop proliferating when confronted with fibroblasts (McGrath, 1983). The proliferation can be reinitiated by supplying E to the culture media. Also in a co-culture system, the mechanisms of E-induction of PR and E stimulation of DNA synthesis have been examined (Haslam, 1986). E induction of PR can be achieved by culturing mammary epithelial cells with live or irradiated mammary fibroblasts or on collagen type I but not by culturing epithelial cells alone on plastic. These results indicate that E induction of PR in epithelial cells depends on an appropriate substratum provided by fibroblasts. In contrast, E stimulation of DNA synthesis in epithelial cells can only be achieved by culturing epithelial cells with metabolically active fibroblasts indicating that a paracrine mechanism might be operating in this hormone regulated pathway. In these studies epithelial cells also promoted E-induced fibroblast proliferation. ll Adipose stroma has also been suggested to influence mammary epithelial proliferation and differentiation by both substrate (Sakakura, 1987) and soluble factors (Levine & Stockdale, 1984; 1985; Stockdale et al., 1986; Wiens et al., 1987). Co-culturing with 3T3-Ll cell line, which can exist as preadipocytes or fully differentiated adipocytes, resulted in maximal DNA synthesis in mammary epithelial cells from mid-pregnant mice. The stimulatory factor was present in both conditioned media and in substratum materials generated by adipocytes. Adipocyte metabolic products, such as unsaturated fatty acids, stimulated mammary epithelial cell proliferation in cell culture (W icha et al., 1979; Bandyopadhyay et al., 1987). Mammary epithelium from non-pregnant, pregnant, lactating and involuting mice can all be induced to synthesize rrrilk protein by mammogenic hormones when co-cultured with preadipocytes or adipocytes and soluble factors (Levine & Stockdale, 1984; 1985; Wiens et al., 1987). The efi‘ects of adipose stroma on hormonal regulation of mammary epithelial proliferation is not known. B. hormonal regulation in breast cancer Studies on hormonal regulation in breast cancer have been done mostly in jg vitro using breast cancer cell lines or in vivo by injecting or transplanting human cells into nude mice. While evidence supporting E as direct mitogen on normal mammary cells is lacking, several studies suggest direct proliferative responsiveness of breast cancer cells to E (Lippman et al., 1976; Chablos et al., 1982; Darbre et al., 1983; Berthois etal., 1986). E can regulate tumor growth by stimulating a large number of enzymes and proteases involved in DNA synthesis (Edwards et al., 1980; Aitken & Lippman, 1983; 1985; Dubik et al., 1987). E can also stimulate several proteins which mediate tumor cell movement across the basement membrane, release 12 growth factors and growth inhibitors from ECM components or from their carrier proteins (Kaufman et al., 1977; Huff & Lippman, 1984; Liotta et. al., 1986). Regulation of neoplastic growth by E is probably via modification of the normal regulatory pathway. In MCF-7 cells, a human breast carcinoma cell line, E induces cell surface receptors for laminin, suggesting that E can mediate cell attachment to basement membranes and increase MCF-7 cell invasiveness (Thompson et al., 1988). Without E, MCF-7 cells cannot form tumors in athymic mice (Soule & McGrath, 1980). These results indicate that E is involved in both tumor cell proliferation and metastasis. Similar to normal cells, the interactions of E with P and peptide growth factors have been identified in breast cancer cells. E regulates ER and PR expression (Saceda et al., 1988; Wei et al., 1988) and induces several growth factors (IGF-H, TGF-a, PDGF) (Dickson et al., 1986; Rozengurt et al., 1985; Yee et al., 1988; Bates et al., 1988) and growth factor receptors (EGF, IGF-I, IGF-II receptors) (Mukku & Stance], 1985; Stewart et al., 1990; Mathieu et al., 1991). In E-fi'ee conditions, the anti-estrogen tamoxifen can totally block cell proliferation stimulated by IGF-I and EGF (Vignon et al., 1987) when ER is present. Clinically, tamoxifen has been used for breast cancer treatment. P are growth inhibitory for human breast cancer cells (Chablos & Rochefort, 1984). P has been shown to modulate several growth factors and their receptors (Murphy et al., 1988; Murphy et al., 1988, Papa et al., 1991). Conversely, growth factors can also stimulate PR expression (Wei et al., 1988, Sumida & Pasqualini 1989) 13 F. Epithelial-stromal interactions in breast cancer Stromal components of breast cancer are generally considered to be normal. However, stromal tissue can undergo morphological and chemical changes under the influence of neoplastic epithelium. In the desmoplastic reaction, a common host response to tumor invasion, fibroblasts undergo morphological changes to smooth muscle phenotype and express a-actin and extensive collagen (Sappino et al., 1988). In DMBA-induced mammary gland tumors, tumor cells are surrounded by collagen and an abnormally thickened basement membrane (Orrnerod et al., 1985). It has been suggested that the abnormal changes of basement membrane and extracellular matrix may alter cell attachment characteristics increasing invasive growth (Orrnerod et al., 1985). Interestingly, the abnormal changes seem not to be restricted to breast fibroblasts. Skin fibroblasts from breast cancer patients have also been demonstrated to exhibit several abnormal behaviors: anchorage independence, overgrowth in culture and increased migratory activities (Duming et al., 1984). These abnormalities in non-malignant stromal tissue may be related to genetic predisposition. Skin fibroblasts from normal relatives of breast cancer patients also show abnormalities (Haggie et al., 1987). It is possible that an abnormal fibroblast environment offers a greater chance for mammary epithelial cells to develop malignancy. Identification of these characteristics has been shown to be helpful for the diagnosis of the disease (Schor et al., 1986). Fibroblasts can also influence cancerous epithelial cells by substratum and soluble factors. Abnormal soluble factors secreted by tumor-derived fibroblasts have been detected in cell cultures (Adams et al., 1988). Conditioned media fi'om fibroblasts of either benign or malignant tumors stimulated the growth of MCF-7 cells. By contrast, conditioned media from normal mammary fibroblasts inhibited MCF—7 cell growth (Adams et al., 1988). This observation suggests that a growth l4 inhibitory factor from normal fibroblasts is lost, or an abnormal growth factor is generated by fibroblasts from benign and malignant tumors. Secreted stimulatory factors from breast cancer cells (malignant epithelial cells) have been identified in conditioned media from E treated cells (Butler WB, Kirkland et al., 1979; Dickson et al., 1986; Yee et al., 1988). Some of these factors might act as autocrine regulatory factors on epithelial cells (DeLarco & Todaro 1978; Osborne & Arteaga, 1990). Others might act on stromal cells and induce paracrine factors which affeet epithelial cells (Lippman et al., 1986; Vignon et al., 1987). Thus identification of possible paracrine and autocrine factors produced by normal epithelial cells and fibroblasts and their effects could advance the understanding of the abnormal changes which happen in malignant . cells. G. The Objectives of Current Studies The ovarian hormones E and Pg have been shown to be importantly involved in mammary gland development. The mammary ductal tree fails to develop when E and Pg are withdrawn by ovariectomy. However, these in vivo observations could be the net result of systemic activation of other growth factors and inhibitors by E and Pg. Alternatively, E and Pg may act on the mammary gland locally to mediate epithelial and stromal cell proliferation. In order to further understand the ,. mechanisms of E and P action, the first specific aim of this dissertation is to determine if E and P act locally to promote mammary gland growth and to determine the relative effects on epithelial vs. stromal tissue in vivo. As discussed above, tissues can interact to modify their hormonal responsiveness. An in vitro cell culture system is a possible approach to address the question about the direct effects of E , P and growth factors which might mediate the effects of E and P on their target tissues. So my second objective is to 15 establish serum-fi'ee media, monolayer culture conditions which would maintain primary cultures of mammary epithelial cells, fibroblasts or mixed cultures containing both epithelial cells and fibroblasts. Next, using the above cell culture conditions, the effects of E and P on different cell types will be studied and the interactions between epithelial cells and fibroblasts in relation to the growth regulation by E, P and growth factors will be investigated. These studies will facilitate our understanding of the effects of E and P, and the role of epithelial-stromal interactions in normal marmnary growth. Advancing our knowledge of these mechanisms will be helpful for the potential design of new therapeutic methods for breast cancer. Chapter 2: The local action of E and P on mammary growth and the relative effects on epithelial vs. stromal tissue in vivo Introduction It has been shown that E and P play an important role in mammary growth and differentiation in rodents (Bresciani 1968; Edery et al., 1984) and in humans (Masters et al., 1977; Brown, 1981). E can induce PR in mammary epithelial tissue. In stromal tissue there are E-independent PR which are constitutively present. The contribution of the two types of P binding sites to mammary growth and development is not known. Even though both ER and PR are present in mammary epithelial and stromal cells, whether E and P mediated mammary growth is a local effect or mediated by systemic factors, and the relative effects of E and P on epithelial and stromal tissues are not clear. Furthermore it is quite likely that E and P synergize to promote mammary growth via E-regulated PR (Haslam & Shyamala, 1979; Haslam & Shyamala, 1980; Haslam,]986). The purpose of the present investigation was to elucidate further how P mediates mammary growth and interacts with the epithelial and stromal components of the gland as well as to assess the role of E in these interactions. Using Elvax 40P implants it is possible to confine bioactive molecules directly to the mammary gland (Silberstein & Daniel, 1982). Local vs. systemic effects of the bioactive compounds have been shown to be distinguished by using this methodology (Haslam, 1988). Therefore E and P were combined with Elvax 40F and implanted into the mammary gland. The effects of E and/or P on mammary gland morphology and DNA synthesis in both hormone implanted and control implanted mammary gland and in epithelial vs. stromal cells were investigated. PR concentration in epithelial cells is subject to E-regulation (Haslam & Shyamala, 1979), therefore E effects on PR were also assessed. l6 17 Materials and Methods Reagents and chemicals: [l7B-methyl-3H] promegestone (R5020; SA, 71 Ci/mmol) and radioinert R5020 were purchased from New England Nuclear Corp. (Boston, MA). 113—(4- dimethylamino-phenyl) 1-1 7 B-hydroxy- l 7a-(prop-lynyl)-e stra-4,9-diene-3 -one (RU486) was a gift from Roussel Uclaf (Romainville, France). All other hormones were purchased from Sigma Chemical Co. (St. Lousis, MO). [Methyl- 3H]-thymidine (50 Ci/mmol) was purchased from ICN Radiochemicals (Irvine, CA). All other chemicals were reagent grade. Hormone implantation: Ten-week-old BALB/c mice were ovariectomized 1 week before Elvax implantation. Elvax pellets containing either Pg, R5020, medroxyprogesterone acetate (MPA), or the anti-progestin, RU486 (0.1 or 1 pg /pellet) alone or combined with varying arnounts of 17B-estradiol (5, 50, or 500 ng/pellet ) were prepared as previously described (Silberstein & Daniel, 1982; Haslam, 1988). Hormone-containing pellets were implanted into the right inguinal gland, and control pellets were implanted in the contralateral left inguinal gland. DNA histoautoradiography: Two or 4 days after implantation, mice received a single ip injection of 3H- thymidine (2 uCi/g body wt.) 1 hour before death. The mammary glands were removed and processed for whole mount analysis as well as autoradiographic analysis as previously described (Banerjee et al., 1976; Haslam, 1988). Determination of mammary epithelial and stromal cell labeling indices ( LI, 18 percent labeled nuclei) was accomplished with the use of a computer-interfaced morphometric digitizing system (Haslam, 1988). Steroid hormone binding assay: Hormone- or control-implanted mammary glands were removed, homogenized separately, and prepared as cytoplasmic extracts, as previously described (Haslam & Shyamala, 1979). Extracts were incubated with 1-20 nM [3H ]R5020 with or without a lOO-fold excess of radioinert R5020 in the presence of 100-fold excess radioinert dexamethasone (to suppress P binding to glucocorticoid binding sites ). Specific binding was determined using a dextran-coated charcoal assay procedure (Haslam & Shyamala, 1979), and binding data were analyzed according to Scatchard (Scatchard, 1947). Tissue DNA was quantitated as previously described (Ceriotti, 1952). Statistics: All results were analyzed by Analysis of Variance (AN OVA) and the Newman-Keuls Multiple Range Test. Results 1. Mammgy gland momhology Fig. 1 shows that implantation of Elvax pellets containing either R5020, a synthetic progestin, or E alone had no effect on mammary gland morphology (Fig. 1a, 1b). However, when R5020 and E were combined and implanted together, ductal sidebranching similar to that which occurs in early pregnancy was observed. The sidebranching occurred directly adjacent to the implant (Fig. 1c) and no sidebranching was observed in the 19 Figure 1. Effect of R5020 and E on mammary gland morphology. Mice were ovariectomized 1 week before implantation with Elvax pellets containing either 5ng E (a), 1 ug R5020 (b), or 1 ug R5020 + 5ng E (c) in the right inguinal gland and control pellets in the contralateral left inguinal gland (d). Mammary gland wholemounts were analyzed 4 days after implantation. 1n = lymph node. It = pellet implant. Magnification, x 64. 21 contralateral control gland (Fig. 1d), indicating the localized nature of the response. Mammary gland wholemounts were analyzed 2, 4, and 6 days after implantation with the maximal morphological response observed on day 4 (data not shown). Pg and medroxyprogesterone acetate (MPA, a synthetic P), when combined with B, were both able to produce morphological responses similar to R5020; the anti-progestin RU486 either alone or combined with E had no stimulatory effect on marmnary gland morphology (data not shown). Dose response experiments with R5020 and E revealed that the dose of E appears to be a limiting factor in determining the extent of mammary gland responsiveness to P. When a low dose of R5020 (0.1 pg) was used, increasing doses of E (5, 50, and 500 ng) caused the area and density of ductal sidebranching surrounding the implant to increase (Fig. 2). At 5 or 50 ng E no ductal sidebranching was observed in the contralateral control gland, indicating that the P response was localized to the hormone-implanted gland. At 500 ng E, plumping of the duct ends without sidebranching was observed in the contralateral gland, indicating that at this dose the effect was due to E (Fig 2), whereas the R5020 efl‘ect still was confined to the hormone-implanted gland. When a higher dose of R5020 (1 pg) was used, increasing doses of E caused much greater increases in the area of sidebranching in the implanted gland (Fig. 3). Furthermore, the sidebranching now also extended to the contralateral gland, indicating that the R5020 effect was no longer limited to the implanted gland. These results demonstrate that the extent of the effect of R5020 was only partially dose dependent and that the major factor determining the extent of the P response was the amount of E. 22 Figure 2. Effect of low R5020 dose plus B on mammary gland morphology. Mice were ovariectomized 1 week before implantation with Elvax pellets containing 0.1 pg R5020 + 5 ng E (a), 50 ng E (c), or 500 ng E (e) respectively in the right inguinal gland and control pellets in the contralateral left inguinal gland (b, d, and f). Mammary gland wholemounts were analyzed 4 days after implantation. t = pellet implant. Magnification, x 64. 23 24 Figure 3. Effect of high R5020 dose plus B on mammary gland morphology. Mice were ovariectomized 1 week before implantation with Elvax pellets containing 1 pg R5020 + 5 ng E (a), 50 ng E (c), or 500 ng E (e) respectively in the right inguinal gland and control pellets in the contralateral lefi inguinal gland (b, d, and f). Marmnary gland wholemounts were analyzed 4 days after implantation. at: = pellet implant. Magnification, x 64. d 26 2. Regulation of PR concentration The effect of E on P-mediated effects as observed above could be associated with E's regulation of PR concentration. To determine if this was responsible for the limiting effect of E on P-mediated growth and morphology, the effect of varying doses of E on PR concentration in hormone-implanted and control contralateral glands was analyzed. The results shown in Table 1 demonstrate that at low dose of E (5 ng) only PR in the implanted gland is increased. However at the 50 and 500 ng doses of E, both contralateral control and E-implanted glands had elevated PR levels. These results demonstrate that the dose dependent effects of E on the degree and pattern of increased ductal sidebranching induced by R5020 + E at various doses could be determined by the amount of E-induced PR. 3. Mammy gland DNA mthesis The effects of hormone implants on DNA synthesis of mammary epithelial cell are shown in Fig. 4. In initial experiments, DNA synthesis was analyzed at both 2 and 4 days after implantation. The greatest increase in L1 was consistently observed at 2 days (data not shown), therefore all subsequent experiments were analyzed at day 2. In contrast to the effect on mammary gland morphology, P alone significantly increased DNA synthesis (Fig. 4). R5020 had a greater effect than MPA, and Pg had the least effect (p < 0.05). This order and magnitude of responses in the marmnary gland is compatible with the P potency of these compounds and relative binding affinity to PR observed in other target organs (F eil , et al., 1972, Philbert & Raynaud, 1974). When P were combined with E, greater increases in DNA synthesis were observed (Fig. 5). As shown above, R5020 and MPA had similar effects and greater effects than Pg (p < 0.05). The anti- progestin, RU486, alone or in combination with E, did not stimulate DNA 27 Table 1. Effect of varying dose estrogen implants on mammary PR concentration Specific 3H-R5020 binding (fmol/mg DNA) Hormone Contralateral E dose implanted control Control 400 : 35a 408 i 39 5 ng 600 i 49* 425 i 50 50 ng 786 i 57 712 j; 39 500 ng 787 i 155 735 j; 163 a Each value represents mean j; SEM from three experiments of sixteen animals for each treatment group. . * p < 0.05 that in the hormone-implanted gland, specific 3H—R5020 binding is greater than that of contralateral control gland. synthesis. However, the E + R5020-induced increase in DNA synthesis was blocked in the presence of RU486 (Fig. 6). The inhibitory effect of RU486 indicates that the R5020 effect on DNA synthesis was due to its P activity and was mediated via E-induced PR. In all cases the stimulation of epithelial cell DNA synthesis was much greater in hormone-implanted glands than in the contralateral control gland, implying a local mode of P action. The results of P effects on stromal cell DNA synthesis are shown in Figs. 7 and 8. P alone increased stromal cell DNA synthesis, and the biological potency 28 Figure 4. Effect of P on mammary gland epithelial cell DNA synthesis. Mice were ovariectomized 1 week before implantation with Elvax pellets containing 1 pg R5020, MPA, Pg, or RU486, respectively in the right inguinal gland (D) and control pellets in the contralateral left inguinal gland (I). Two days after implantation, LI was determined in ductal epithelium by DNA histoautoradiography. Each value represents the mean 1 SEM of two or three experiments; each experimental group contained three or four animals. * p < 0.05 that hormone treated groups are greater than controls; ** p < 0.05 that control-implanted gland are greater than control group but less than hormone-implanted glands. Labefinglndex (% labeled Nuclei) mm M a 29 fi :- tt * n72 control R5020 :72 MPA Pg RU486 Hormone Treatment 30 Figure 5. Efiect of P plus B on epithelial cell DNA synthesis. Mice were ovariectomized 1 week before implantation with Elvax pellets containing either 5 ng E alone or in combination with lpg R5020, 1 pg MPA, lpg Pg or 1 pg RU486, respectively in the right inguinal gland ( ) and control pellets in the contralateral left inguinal gland ( ). LI was determined in ductal epithelium by DNA histoautoradiography. Each value represents the mean : SEM of two or three experiments; each experimental group contained three or four animals. “ p < 0.05 that P plus E-implanted glands are greater than E alone or contralateral control-implanted glands. Labefinglndex (% labeled Nuclei) E 31 , o i ll / a mi R5020 MPA Pg RU486 +E +E +E +E Hormone Treatment 32 Figure 6. Effect of RU486 on epithelial cell DNA synthesis. Mice were ovariectomized 1 week before implantation with Elvax pellets containing 5 ng E, 0.1 pg R5020, or 1 pg RU486, in indicated combinations in the right inguinal gland (D) and control pellets in the contralateral left inguinal gland (I). LI was determined in ductal epithelium by DNA histoautoradiography. Each value represents the mean : SEM of two or three experiments; each experimental group contained three or four animals. * p < 0.05 that R5020-implanted glands are greater than controls; ** p < 0.05 that R5020 + E-implanted glands are greater than E-irnplanted glands. Labefinglndex (% labeled Nuclei) 33 l—to |-—ri . 7 a “I \\\\\\\\‘ \\\\\\\\V control E R5020 R5020 R5020 R5020 + -+ + RU486 E RU486+E Hormone Treatment 34 Figure 7. Effect of RU486 on stromal cell DNA synthesis. Mice were ovariectomized 1 week before implantation with Elvax pellets containing either lpg R5020, 1 pg MPA, lpg Pg-orlpg RU486, respectively in the right inguinal gland (I:I)and control pellets in the contralateral left inguinal gland (I). Two days after implantation, LI was determined in stromal cells by DNA histoautoradiography. Each value represents the mean : SEM of two or three experiments; each experimental group contained three or four animals. "' p < 0.05 that hormone and control-implanted glands are greater than control group. Labefinglndex (% labeled Nuclei) 35 7‘ J—H 35 Ha M control R5020 EPA Pg RU486 Hormone Treatment 36 Figure 8. Effect of P plus E on stromal cell DNA synthesis. Mice were ovariectonrized 1 week before implantation with Elvax pellets containing either 5 ng E alone or in combination with lpg R5020, 1 pg MPA, lpg Pg orl pg RU486, respectively in the right inguinal gland (a) and control pellets in the contralateral left inguinal gland (I). LI was determined in stromal cells by DNA histoautoradiography. Each value represents the mean j; SEM of two or three experiments; each experimental group contained three or four animals. * p < 0.05 that hormone- and control-implanted glands in P plus E-treated groups are greater than B alone group. at RU486 +E kg Du .éa P N .1 M m n O #1 L% i l "uuuuuuu. 987654321 9282 8.82 3 xonc_mc=onom _ I _ 0 +E +E +E Hormone Treatment 38 of the P followed a similar pattern as that observed for epithelial cells except that MPA and R5020 were equally potent (Fig. 7). In the case of MPA or Pg, DNA synthesis in hormone-implanted mammary glands was higher than that of the contralateral control mammary glands. However, R5020, the most stable compound, caused similar increase in DNA synthesis in both sides of the mammary glands (Fig. 7). One interpretation of the lower DNA synthesis in MPA or Pg treated groups in their respective contralateral control mammary glands may be due to the metabolic inactivation of MPA or Pg. Thus the local mode of P effect in stromal tissue was not indicated. RU486 by itself increased stromal cell DNA synthesis. In contrast to its effect in epithelial cells, E did not increase the effects of R5020, MPA, or Pg in stromal cells (Fig. 8). However, the effect of ' RU486 was increased by estrogen. These results further demonstrate that the regulatory effect of E on P effects observed in marmnary epithelium is not operative in stromal cells. Discussion The ability of P to promote mammary gland growth and differentiation iuxi_vu was assessed in two ways by their ability to: 1) stimulate DNA synthesis in epithelial and stromal cells; and 2) produce ductal sidebranching that gives rise to lobuloalveolar morphogenesis that occurs during pregnancy. Furthermore, the role of E in these events was also assessed. The results demonstrate that P tested herein produce their stimulatory effects on epithelial DNA syntheses and ductal sidebranching by virtue of intrinsic P activity. The high activity of R5020 and MPA, which have reduced sensitivity to metabolic inactivation, and the lesser activity of Pg, which is rapidly metabolized, indicates that most likely the native compounds rather than P metabolites are responsible for the activity. 39 It is of interest to note that whereas RU486 inhibited the increase in epithelial cell DNA synthesis resulting from E + R5020 treatment, it did not reduce the DNA synthesis increase observed with R5020 alone. One possible explanation for this observation is that RU486 interacts preferentially with newly synthesized E-induced PR. It has been previously reported that PR, newly synthesized in response to E appears to be different from preexisting old PR with regard to molecular size and/or form (Haslam 1987). Thus, these differences may be the basis for the observed differential effect of RU486. In combination with E, P can produce a localized effect on mammary gland morphology and epithelial DNA synthesis, implying direct action of P in the mammary gland. E enhances the' stimulatory effect of P on epithelial cell DNA synthesis. Analysis of the effect of E dose on PR concentration in conjunction with the extent and patterns of ductal sidebranching indicate that the limiting factor for P-mediated effects in epithelial cells is E-induced PR concentration. Although P alone increased DNA synthesis, this does not appear to be sufficient to cause sidebranching. It has been also observed that there is significant heterogeneity in the distribution of epithelial ER (Haslam & Nurnmy, 1992). This suggests that topographic differences also may exist with regard to localization of DNA synthesis, and this also may be a determining factor for the occurrence of sidebranching. Thus, these results extend the previous observations and further support the concept that in normal mammary epithelial cells P has major mitogenic activity. Furthermore, the P-induced response is the result of direct/local action in the mammary gland and that it is mediated via E-induced PR. P also stimulates mammary stromal cell DNA synthesis. However, the mechanisms operative in stromal cells appear to be different from those occurring 40 in epithelial cells. Firstly, even at low R5020 doses the hormone effects are not confined to the implanted gland but are expressed equally in the control contralateral gland. Secondly, E does not synergize to enhance the proliferative effect of P. Thus it can be concluded that the effects of P on stromal cells are not mediated via E-induced PR. This finding is compatible with the previous observations that stromal cell PR are E-independent (Haslam & Shyarnala 1981). Unexpectedly, DNA synthesis was also stimulated by the anti-progestin, RU486. Whereas E did not synergize with P, it did increase the effect of RU486. Others recently have shown that RU486 also can stimulate proliferation of certain breast cancer cell lines in vitro (Bowden et al., 1989; Hissom et al., 1989); however, the basis for this response has not been determined. RU486 also has potent anti- glucocorticoid activity (Moguilewsky & Philibert 1984; Philibert 1984), and it remains to be established whether the effect of this compound in stromal cells is due to its anti-glucocorticoid effects. Chapter 3: Serum-free primary culture of normal mouse mammary epithelial and stromal cells Introduction Mammary gland growth and development are known to be dependent upon systemic hormones (Nandi, 1958). However, increasing evidence indicates that local epithelial-stromal cell interactions also play an important role in hormonal control of the mammary gland (Haslam, 1991; Haslam & Counterman, 1991). Possible mechanisms include paracrine effects caused by soluble factors produced locally by various cell types in the gland (Haslam, 1986). Identification of the specific mechanisms underlying epithelial-stromal cell interactions are difiicult to approach in vivo in the whole animal and in vitro cell culture systems have provided important insights (McGrath 1983; Haslam & Levely, 1985; Haslam, 1986). However, most previous m studies of mammary epithelial-stromal cell interactions have been carried out in serum containing medimn (McGrath, 1983; Haslam & Levely, 1985; Haslam, 1986). Because of the presence of numerous growth factors and other undefined potential growth promoting components in serum our ability to define specific mechanisms is limited. In one study using serum-free media (Beck & Hosick, 1988), adipocyte effects on collagen-embedded epithelial cell cultures were studied; it was found that adipocyte conditioned medium promoted epithelial cell proliferation. However, the mechanisms underlying epithelial and stromal interactions remain to be identified. In the majority of studies where serum-free media have been used to study mammary cells, only the behavior of mammary epithelial cells has been investigated (Irnagawa et al., 1982; 1985). Furthermore, the composition of the media was designed to support maximal epithelial growth. However, when using maximal growth conditions it has been difficult to demonstrate in vitro growth 41 42 promoting effects of mammogenic hormones comparable to their growth promoting effects in vivo (Irnagawa et al., 1982; 1985). In vivo studies to investigate the mitogenic effects of hormones and growth factors have always been carried out in ovariectomized mice to reduce cell proliferation to basal levels and to remove endogenous hormones (Nandi, 1958; Haslam, 1988; 1988; Wang et al., 1990). Thus the creation of serum-free cell culture conditions which support cell maintenance rather than maximal proliferation resemble more closely the m environment for the study of hormonal control of cell proliferation. The purpose of this study was to define serum-free media and monolayer culture conditions that would support the maintenance (rather than extensive growth) of primary cultures containing only mammary epithelial cells, only mammary fibroblasts or mixed cultures containing both epithelial cells and fibroblasts. Monolayer conditions were chosen to facilitate future experimental manipulations of interactions between epithelial and stromal cells in co-culture conditions. Comparative studies were carried out with mammary cells from nulliparous and mid-pregnant mice to determine the influence of the state of mammary gland differentiation on growth in serum-free medium. The identification of suitable culture conditions should provide an approach to studies of growth factor interactions with mammogenic hormones and the investigation of potential paracrine interactions between mammary epithelial cells and fibroblasts. Materials and Methods Reagents and Chemicals: Collagenase HI and pronase were purchased from Cooper Biomedical (Freehold, NJ) and Calbiochem (La Jolla, CA) respectively. Culture media phenol red-free Dulbecco's modified Eagle's/ Ham's nutrient mixture F12 (DME/F12 ) (1:1), Hank's balanced salt solution (HBSS), bovine serum albumin (BSA) 43 (fi'action V), penicillin and streptomycin were purchased from GIBCO (Grand Island, NY). All other culture ingredients and hormones were obtained from Sigma Chem. Co. (St. Louis, MO). All other chemicals were reagent grade. Cell Culture: Balb/c 12-15 week-old, nulliparous or mid-pregnant (12-14 days) mice from our own colony were the source of mammary tissues. Animals were fed mouse chow (Teklad, Madison, WI) ad libitum and had free access to water. Animal facilities were temperature and humidity controlled with a light cycle of 12h on 12h ofi‘. Cell cultures containing both mammary epithelial cells and fibroblasts (mixed cultures), only epithelial cells, or only fibroblasts were obtained as previously described (Haslam & Levely, 1985). Briefly, mammary cell suspensions were obtained by enzymatic dissociation of minced mammary tissue in 0.1% collagenase. To obtain cultures enriched for mammary fibroblasts, the cell suspension was differentially centrifuged twice at 80 x g for 30 sec and the resulting pooled supematants were the source of the fibroblasts. To obtain cultures enriched for mammary epithelial cells, the pellet remaining after the removal of fibroblasts was fruther purified by percoll density gradient separation. Cell numbers were determined by counting nuclei using a hemocytometer as previously described (Haslam & Levely, 1985) and were plated at a density of 4.2 x 105 viable cells/cmz, or 1.4 x 105 cells per well of a 96-well culture dish for mixed and epithelial cultures or 9.6 x 105 cells per well‘of a 4-well culture dish for fibroblast cultures. Cell viability averaged 90% as determined by trypan blue exclusion. Both epithelial cells and mixed cells were plated on rat-tail collagen- coated culture plastic dishes to enhance cell attachment. Undiluted rat-tail collagen prepared as previously described (Emerrnan et al., 1977) was aliquoted (0.15 ml/cm2 ) into each culture dish, air dried and washed once with HBSS prior 44 to plating cells. Pure fibroblast cultures were plated on tissue culture plastic since poor cell attachment was obtained on collagen-coated dishes. The basal culture medium for all cell types contained phenol red-free DME/F 12 (1:1), Hepes (lSmM), penicillin (0.1 mg/ml), and streptomycin (0.05 mg/ml), hydrocortisone (H, 0.2pM), and insulin (1, 0.] pg/ml). To this medium were added indicated combinations of fetuin (F, l, 0.5, 0.1 mg/ml), BSA (5mg/ml) and transferrin (T, 5 pg/ml). Media was changed every day. Quantitation of Cell Growth: Cell number was quantified using a modified 3-(4,5- dimethylthiazol-Z-yl)- 2,5-diphyltetrazolium bromide (MTT) assay (Carmichael et al., 1987; Dennizo & Lang, 1986; Hahm & IP, 1990). The assay is based upon the conversion of tetrazolium into blue formazan by the mitochondrial enzyme succinate- dehydrogenase. The color production is directly related to viable cell number (Dennizo & Lang, 1986). Briefly, phenol red-free DME/F12 media containing 1 mg/ml MTT was added to the cultures for 4 h at 37 ° C. Cells were then fixed with formal-saline (10% formaldehyde in 0.85% NaCl) for 30 min to prevent cell detachment and loss and then the MTT-formal-saline supernatant was removed. Isopropanol (100%) was then added to each well to dissolve the formazan crystals and absorbance at 540 nm was measured using a multiwell spectrophotometer Bio- tek EL 310 reader (Biotek Instruments, Inc., Burlington, VT). A standard curve relating cell number to absorbance for mixed cells, epithelial cells and fibroblasts from nulliparous and mid-pregnant mice (Fig. 9) was generated by counting nuclei of viable, fi'eshly dissociated cells with the aid of a hemocytometer as previously described (Hahm & IP 1990). No significant difference in absorbance between epithelial cells and fibroblasts was observed. In all experiments, absorbance has been converted to cell number based upon the standard curves. 45 Figure 9. MTT assay standard curves showing the linear relation between cell number and absorbance. Dissociated mammary cells from nulliparous or mid-pregnant rrrice were quantified for cell number by the MTT assay and by cell counting as described in the Method section. Each point represents the mean 1 SEM of triplicate determinations. Three representative experiments are plotted for each cell type. Abs 540 (n m) 2.0 - 1.5 '- 1.0 0.5 - ‘ 0.0 - ,_ B pregnant 2.0 46 A nulliparous O fibroblasts O epithelial cells A mixed cells 1- 0 2 4- 6 8 i 10 12 14 16 Cell Number (x10‘5) 47 Irnmunocytochemistry: To establish the purity of epithelial cell and fibroblast cultures and the ratio of epithelial cells vs. fibroblasts in mixed cultures, fibroblasts were identified immunocytochemically using goat antivimentin antibody (a gift of Dr. B. Asch, Roswell Park Cancer Institute, Buffalo, NY). Fibroblasts were cultured on glass coverslips; epithelial cells and mixed cells were cultured on collagen-coated coverslips. Cells were fixed at -20°C in absolute methanol for 5 min and then in acetone for 5 min. Endogenous peroxidase activity was blocked with 0.1% hydrogen peroxide in methanol. Nonspecific antibody binding was blocked with normal rabbit serum. Antivimentin antibody binding was detected with a peroxidase labeled rabbit antigoat IgG. Peroxidase activity was detected by incubating with 3,3'-diaminobenzidine and hydrogen peroxide. Statististics: All results were analyzed by Analysis of Variance (AN OVA) and the Newman- Keuls Multiple Range Test. am 1. Optimal serum-free culture conditions for the maintenance of mixed cell m Enzymatic dissociation of the mammary gland resulted in a suspension of cells comprised of multicellular epithelial organoids, single epithelial cells and fibroblasts (Haslam & Levely, 1985). The mixed cell suspensions containing both epithelial cells and fibroblasts from either nulliparous or mid-pregnant mice were plated on collagen-coated plates to facilitate attachment in the absence of serum. Serum-free medium containing hydrocortisone, transferrin, BSA and fetuin has previously been used for the successful culture of mammary epithelial cells 48 (Imagawa et al., 1982) and was used as a starting point for the present studies. Fetuin, a known attachment factor, is required for cell attachment in serum-free medium (Fisher et al., 1958; Hayman et al., 1982). However, fetuin along with BSA and transferrin can also contribute to epithelial cell grth in serum-free medium (Imagawa et al., 1982). The relative contributions of these components to growth were examined for cells plated on collagen. As can be seen in Fig. 10, fetuin, transferrin or BSA each contributed significantly to mixed culture growth. Thus tranferrin and BSA were deleted from future media. A fetuin dose response study carried out in medium without transferrin and BSA (Fig. 11) showed that reducing fetuin concentration to 0.1 mg/ml resulted in reduced growth stimulation while sustaining good maintenance of mixed cultures. This was true for mixed cultures derived from both nulliparous and mid-pregnant mice. Comparing the effect of fetuin on collagen vs. plastic (Fig. 12) showed that optimal maintenance was obtained on the collagen-coated substratum. Hydrocortisone did not appear to be required for mixed culture maintenance since its removal from the medium did not have a negative effect on cell maintenance (data not shown). From these experiments, the optimal conditions and minimal medium requirements for maintaining mixed cultures were determined to be basal medium containing fetuin (0.1 mg/ml), and a substratum coated with collagen 1. Under these conditions good cell attachment and maintenance were obtained without high levels of proliferation. The proportion of fibroblasts in mixed cultures was determined based upon a quantitative morphometric analysis as previously described (Haslam, 1988). Fibroblasts were identified by staining with antivimentin antibody. The cultures obtained from nulliparous mice were composed of a ratio of epithelial cells to fibroblasts of 60:40 on days 1 and 70:30 by day 7 of culture (Figs. 13, 14A). The 49 Figure 10. Effects of BSA, transferrin or fetuin on grth of mixed cultures from nulliparous mice and mid-pregnant mice. Dissociated mixed cells from nulliparous (A) or pregnant (B) mice were plated (4.2 x 105 cells/cm2 ) on collagen in basal medium (BM) (0), BM + T (O), BM + BSA (A), BM + F (1 mg/ml) (A), or BM + BSA, T, F (a). Each value represents the mean 1 SEM of triplicate determinations from a representative experiment. * p = 0.05 that all treatment groups had higher total cell numbers than the BM group. Cell Number (x10‘5) 50 MIXED CULTURE 1o-- A nulliparous O—O BM C—. T a__ A—A BSA A—A F El—El T+BSA+F 5... 4-- * .. * - 2" '% % o i i _‘ 10.. B pregnant a... 6-- * Ill . at: 4-- ééz :6? —'= C 2' 29 5 A T .. Y O .T 1 1 3 5 7 Days 51 Figure 11. Effects of fetuin dose on mixed culture growth. Dissociated mixed cells from nulliparous (A) and mid-pregnant (B) mice were plated (4.2 x 105 cells/cmz) on collagen in BM containing 1(5), 0.5 (A), 0.1 (e), or 0 (0) mg/ml F. Each point represents the mean : SEM of triplicate determinations from representative experiments. "' p = 0.01 that higher total cell numbers were obtained at all dosages of F than in BM without F. Cell Number (trio-5) 52 MIXED CULTURE 10-- A nulliparous O—O 0mg O—O 0.1mg . 5.. A-—,A 0.5mg A—A 1.0mg 5.. 4-- * * * w- #5 2-- '/§ 5 . v%t/’* * ‘8 c 5 5 o ‘ a. " A 10" B pregnant 5.. 5“ at: al- '1' 4" /éi _§ § - *0 O 2“ £/ . .. éé e e_ .. o r *0 1 3 5 7 53 Figure 12. The effect of fetuin on the growth of mixed cultures on collagen or plastic. Dissociated mixed cells from nulliparous (A) or mid-pregnant (B) mice were plated (4.2 x 105 cells/cm2 ) on collagen-coated (0) or plastic (0) substratum in BM + F (0.1 mg/ml). Each point represents the mean i SEM of triplicate determinations from 2 experiments. * p = 0.01 that higher cell numbers were obtained on collagen. Cell Number (x10‘5) 54 MIXED CULTURE 10-- ' A nulliparous O—O collagen a-- 0—0 plastlc 5-.. 4... at: * at: 2; ' 4.5 .. §§r 4‘9? Y *9 o -T i=3 3 i 10-- B pregnant 5.1.. 5.. 4 at :l: * .. * - - .. 2"/9 \" T - T 0T L‘ 4 4 4__ 1 3 5 7 55 Figure 13. The proportional distribution of epithelial cells vs. fibroblasts in mixed cultures. Dissociated cells from nulliparous (A) or mid-pregnant (B) mice were plated (4.2 x 105 cells/cmz) on collagen in BM + F, 0.1 mg/ml. Morphometric quantification of epithelial cells and fibroblasts was carried out at indicated times. Each value represents the mean 1 SEM of 2 separate experiments. 96 of Total Cell Number 100- 01 o r A nulliparous 56 [:1 epithelial cells - fibroblasts a. 50- B pregnant Days 57 Figure 14. Photomicrographs of cultured mammary cells. Cells from nulliparous mice cultured as mixed cultures (A), epithelial cultures (B), or fibroblast cultures (C) were photographed under phase contrast after 3 days for mixed cultures and fibroblast cultures and 5 days for epithelial cultures. Mag. X 200. 59 mixed cultures obtained from mid-pregnant mice contained a higher ratio of epithelial cells to fibroblasts ranging from 65:35 on day l to 75:25 on day 7 of culture. 2. Optimal serum-free conditions for the maintenance of epithelial cell cultures Cell maintenance and growth of epithelial cultures from nulliparous and mid- pregnant mice were also analyzed in maximally supplemented medium (BM plus tranfenin, BSA, fetuin, 1 mg/ml) and in media lacking transferrin, BSA and with high (1 mg/ml) or low (0.1 mg/ml) fetuin concentrations on plastic or collagen (Fig. 15, 16). Good maintenance was obtained with 0.1 mg/ml fetuin on collagen (Fig. 15A, 16A). Epithelial cultures from nulliparous mice showed similar maintenance on collegen or on plastic (Fig. 153). Epithelial cultures from mid- pregnant mice were not maintained as well on plastic, particularly at low fetuin dose (Fig. 16B). As in the case of mixed cultures, hydrocortisone was not required for epithelial culture maintenance since its removal from medium did not negatively affect cell maintenance (data not shown). In all cases epithelial cells comprised more than 99% of the total cells (Fig. 14B). Based upon the above experiments the minimally supplemented medium and substratum conditions identified for mixed cultures can also be used for the maintenance of epithelial cultures from mammary glands of nulliparous or mid- pregnant mice. A comparison of epithelial culture and mixed culture growth on collagen in the minimally supplemented medium is shown in Fig. 17. Under these conditions, mixed cultures and epithelial cultures from nulliparous mice exhibited identical growth behavior. In contrast, mixed cultures from mid-pregnant mice exhibited significantly better maintenance between days 3 and 7 than epithelial cultures. These results suggest that interactions between epithelial cells and 60 Figure 15. Effects of medium composition and substratum on growth of epithelial cultures from nulliparous mice. Dissociated epithelial cells fi'om nulliparous mice were plated (4.2 x 105 cells/cmz) on collagen (A) or plastic (B) in BM + BSA, T, F, 1 mg/ml (O ,O), BM + F, l mg/ml(l,c1), or BM + F, 0.1 mg/ml (A,A). Each point represents the mean : SEM of triplicate determinations from two experiments. Cell Number (mo-5) 61 NULLIPAROUS EPITHELIAL CULTURE A collagen o—o F 1mgg+BSA+T I—I F 1mg A—A r 0.1mg) * Ill '%.~ ‘1 “I B plastic 62 Figure 16. Effects of medium composition and substratum on the growth of epithelial cultures from mid-pregnant mice. Dissociated epithelial cells were plated (4.2 x 105 cells/cmz) on collagen (A) or plastic (B) in BM + BSA, T, F, l mg/rrrl (O, O), BM + F, 1 mg/ml (I,D), or BM + F, 0.1 mg/ml (A, A). Each point represents the mean 1 SEM of triplicate determinations from two experiments. Cell Number (mo-5) 63 PREGNANT EPITHELIAL CULTURE 10-- A collagen o— _0F{1mg)+BSA+T l— IFlmg 8-- A— AFO1mg) 5.. 4" . .. -—— ’ I. 2-_/‘_ 9 5 f 0 10.. B plastic 5-. 5.. 4» 6 , 2"/¢/: : \lfl \-//A A\- A " ‘ .. o ‘ A— 1 3 5 7 Days 64 Figure 17 . Comparison of the growth of rrrixed cultures and epithelial cultures from nulliparous and mid-pregnant mice in minimally supplemented medium. Dissociated mixed cells (0) or epithelial cells (0) from nulliparous (A) or mid-pregnant (B) mice were plated (4.2 x 105 cells/cm2 ) on collagen in BM + F, 0.1 mg/ml. Each point represents the mean 1 SEM of triplicate determinations from representative experiments. * p = 0.05 that mixed cultures total cell numbers are greater than epithelial cultures. Cell Number (x10-5) 65 10-- A nulliparous O—OE ithelial Culture . O—O lxed Culture 5.. 5-- 4-. 2'”. '/§ 6 (3‘ o & 1o-- B pregnant 3.. 5... a: * * 4.. ' - - zu/Q ‘9‘ —Q Q ,1 1 3 5 7 66 fibroblasts in mixed cultures from mid-pregnant mice may be facilitating this effect. 3. Optimal serum-free conditions for the muintenunce of mammaryfifibroblrfl m Purified mammary fibroblasts attached but failed to spread and grow on collagen (data not shown). Thus all studies with fibroblasts were carried out on tissue culture plastic. Maximally supplemented serum-free media (BM + tranferrin + BSA + fetuin 1.0 mg/ml ) caused a significant increase in cell proliferation (Fig. 18). The effects of fetuin, transferrin and BSA on fibroblast maintenance and growth were tested individually (Fig. 18). Since basal medium plus fetuin sustained the same degree of proliferation as basal medium plus fetuin, tranferrin and, BSA, a requirement of transferrin and BSA was not indicated. In fact, addition of BSA appeared to have an inhibitory effect on fibroblast maintenance. A dose response study with fetuin revealed that optimal fetuin dose for fibroblast culture maintenance was 0.5 and 1.0 mg/ml for cultures fi'om mid- pregrrant and nulliparous mice, respectively (Fig. 19). In contrast to mixed cultures and epithelial cultures, the presence of hydrocortisone was necessary for optimal maintenance of fibroblast cultures from mid-pregnant mice (Fig. 20). Under these conditions fibroblast cultures were comprised of greater than 95% fibroblasts (Fig. 14C). Discussion In this study we have determined suitable compositions of serum-flee media and substratum conditions that allow for maintenance without extensive proliferation of primary, monolayer cultures containing mouse 67 Figure 18. Effects of BSA, transferrin or fetuin on the growth of fibroblast cultures from nulliparous or mid-pregnant mice. Dissociated fibroblasts from nulliparous (A) or mid-pregnant (B) mice were plated (4.2 x 105 cells/cmz) on plastic in BM (0), BM + T (A), BM + BSA (A), BM + F, 1 mg/ml (a), or BM + BSA, T, F, 1 mg/ml) (0). Each point represents the mean 1 SEM of 2 experiments. * p = 0.05 that total numbers of cells obtained in the presence of F, or T, or T+BSA+F were greater than in BSA or BM. 68 FIBROBLAST CULTURE w a..omen.fl.;:xu * _ .o.:fl B mrwrn 3.34 OAADO . t .3“. t. .A. .C‘. * 4J3. .Otf _ I I _ _ _ 7 _ i I : TGITDAO:A.1 .86. .A. w t m m p n n L m. L n A p can v‘. A Y a ‘\ .a u u u. in u u n “ T .n... u 9 6 3 0.5.. u 9 6 3 o Amlo. to 53:52 __oo Days 69 Figure 19. Effects of fetuin dose on fibroblast cultures growth. Dissociated cells from nulliparous (A) or mid-pregnant (B) mice were plated (4.2 x 105 cells/cmz) on plastic in BM containing 1 (A), 0.5 (A), 0.1 (O), or 0 (0) mg/ml F. Each point represents the mean : SEM of triplicate determinations from representative experiments. "‘ p = 0.05 that F dose of 0.5 and 1mg resulted in higher cell numbers than a 0 or 0.1mg dose. Cell Number (trio-5) 70 FIBROBLAST CULTURE 15 .. A nulliparous O O 0mg 0 O 0.1mg 12" A A 0.5mg A A 1.0mg 9.. * al: 6-- A g . 1‘ 1' sfi§i/:t/9 4‘ .A , . o I i 15“ B pregnant 12-- 9.. 6" * g '7’ 3 \ée‘SVg—r #— 0 1 3 5 7 71 Figure 20. Effects of hydrocortisone on fibroblast cultures growth. Dissociated fibroblasts from nulliparous (A) or mid-pregnant (B) mice were plated (4.2 x 105 cells/cmz) on plastic BM F, (0.5 mg/ml) with (o), or without hydrocortisone (0). Each point represents the mean : SEM of triplicate determinations from a representative experiment. * p = 0.05 that the total numbers of cells grown in the presence of hydrocortisone are greater than those grown in its absence. Cell Number (trio-5) 72 FIBROBLAST CULTURE 15- T . A nulllparous o—O +Hydrocortisone 12“ 0—0 —Hydrocortisone 9... 5- i a 15" B pregnant 12-- 9... 5.. at: A 5 O 7 . ‘\2 0 1 3 5 7 73 mammary epithelial cells, mammary fibroblasts or mixed cultures containing both epithelial cells and fibroblasts. In the past, serum-free media have been developed for the primary culture of mammary epithelial cells (Imagawa et al.,1982). Serum factors such as fetuin, BSA and transferrin and the hormone hydrocortisone have been added to serum- fiee media to enhance cell attachment and maintenance. However, these serum factors and hydrocortisone can also cause such high baseline cell proliferation that the growth stimulatory effects of mammogenic hormones could not be detected using these media (Imagawa et al., 1982). In our study we have therefore examined these serum factors and hormones for their effects on cell growth. Fetuin, a required attachment factor, also promoted mammary cell growth herein. Several studies have shown that fetuin prepared according to Pederson's method contains high molecular weight species which caused mouse embryonal carcinoma cell growth in serum-free medium (Salomon et al., 1982; 1984). We have been able to reduce the mitogenic effect of fetuin on mixed cultures and epithelial cultures without losing its attachment activity by reducing its concentration to 0.1 mg/ml. In our studies BSA promoted cell proliferation possibly due to its lipid contaminants as has been previously reported (Imagawa et al., 1982). Transferrin has also been reported to replace serum for the stimulation of mouse 3T6 cell proliferation (Rudland et al., 1977). However, in the present studies neither BSA nor transferrin were required for mixed culture, epithelial culture or fibroblast culture maintenance. Hydrocortisone is frequently included in media for the culture of mammary cells and may play an important role in the expression of differentiated functions such as casein synthesis (Emerrnan et al., 1977). However, the presence of hydrocortisone was required only for the maintenance of fibroblast cultures from mid-pregnant mice. 74 In the case of both mixed cultures and epithelial cultures, optimal maintenance was observed in basal media supplemented with fetuin (0.1 mg/ml). Although cell attachment did take place on plastic, Optimal attachment and maintenance was observed on a collagen-coated substratmn. The importance of an appropriate substratum for normal function of mammary epithelial cells has been previously observed and collagen has been reported to support more normal function than tissue culture plastic (Emerman et al., 1977). Comparison of behavior of mixed cultures vs. epithelial cultures revealed that mixed cultures fiom mid-pregnant mice exhibited better maintenance than epithelial cultures. We have also recently observed that nulliparous epithelial cell responsiveness to ovarian hormones is altered in the presence of fibroblasts under mixed culture conditions (unpublished observations), indicating that our culture system is capable of exhibiting paracrine interactions. Taken together, these results indicate not only that interactions between epithelial cells and fibroblasts can modify growth behavior but also that the developmental state and degree of mammary gland differentiation can affect cell-cell interactions and ensuing proliferative behavior in serum-free medium. Media composition and substratum conditions required for attachment and maintenance of fibroblast cultures were significantly different from those required for mixed or epithelial cell cultures. Optimal maintenance required a higher concentration of fetuin. This fetuin requirement for fibroblasts does not appear to be due to its function for attachment, since fibroblasts initially attached well in minimal media established for mixed cultures and epithelial cultures, but failed to be maintained. Thus as is the case for epithelial cultures and mixed cultures, fibroblast cultures also require the growth promoting activity of fetuin for maintenance. Hydrocortisone was shown to be required for maintenance of fibroblast cultures from mid-pregnant mice. Others have also reported that 75 glucocorticoids are essential for pregnant mammary tissue survival and secretory activity development in organ culture (Topper & Oka, 1974). With the advancing of pregnancy there was an increased tissue uptake of corticosteroid (Pearlrnan et al., 1981). The mechanism of this hydrocortisone requirement is not clear. Interestingly, fibroblasts were well maintained in a less supplemented medium when cultured in the presence of epithelial cells in mixed cultures. These observations indicate that epithelial cells may be involved in the regulation of fibroblast firnction. Comparison of attachment and growth of fibroblasts on collagen vs. plastic revealed another major difference. Fibroblasts failed to spread out and proliferate on collagen. However, when a mixture of epithelial cells and fibroblasts were plated together, the fibroblasts did spread out and proliferate. These observations suggest another level of cell-cell interaction and indicate that in the presence of epithelial cells, fibroblast spreading and proliferative behavior is modified. In summary, we have identified suitable serum-free conditions for the maintenance of mammary epithelial cells, fibroblasts and mixed cell cultures. We have also observed interactions between epithelial cells and fibroblasts which support previous reports that such interactions can potentially modify hormonal responsiveness and proliferative behaviors of mammary cells (Enami et al., 1983; Haslam 1986; Haslam, 1991; Haslam & Counterman 1991). The results provided herein should facilitate future studies to investigate the roles of growth factors and their interactions with mammogenic hormones and the potential autocrine and/or paracrine interactions between the various mammary cell types. Chapter 4: The effects of estrogen, progesterone and growth factors on epithelial, fibroblast, and mixed primary cultures in serum-free monolayer culture conditions Introduction The media components and culture conditions for supporting the maintenance of epithelial, fibroblast, and mixed cell cultures have been carefully defined in Chapter Three. The occurrence of direct effects of E and P on mammary epithelial or fibroblast cells can now be investigated under these serum-free conditions. Identification of the effects of these hormones on epithelial cells vs. fibroblasts and on mixed cell cultures will help to elucidate the role of epithelial-stromal cell interactions in these hormone regulated effects. E and P have been shown to interact with epidermal growth factor (EGF) in the stimulation of epithelial DNA synthesis in vivo (Haslam et al., 1993) and the regulation of lobuloalveolar development in vitro organ culture system (Tonelli & Sorof, 1980). IGF—I has been shown to synergize with EGF in stimulating mammary epithelial cell proliferation in cells cultured in collagen gels (Decks et al., 1988). However, the relationship between IGF-I, E and P and their effects on mammary fibroblasts are not clear at present. Therefore the effects of EGF and IGF-1 on mammary epithelial, fibroblast, and mixed cell cultures and their interactions with E and P were examined in our serum-free monolayer culture system. In these studies, the steroid hormones E and R5020 (a synthetic progestin) and the growth factors EGF and IGF-1 were added to culture media separately or in various combinations. The effects of these hormones and growth factors on cell proliferation in epithelial, fibroblast and mixed cell cultures from nulliparous or mid-pregnant mice was compared. The maintenance of ER and the regulation of PR were also examined. These studies helped to define: a) the interactions of 76 77 steroid hormones and grth factors in regulating cell proliferation; b) the proliferative effects of steroid hormones and growth factors on different mammary cell types and the role of epithelial-stromal cell interactions and c) the effects on mammary cells at different developmental stages. Material and Methods Cell culture: BALB/c 12-15 week-old, nulliparous or mid-pregnant (12-14 days) mice were the source of mammary tissues. Mixed cells were obtained by enzymatic dissociation of finely minced mammary tissue in 0.1% collagenase for 30 min at 37 C. To obtain cultures enriched for mammary fibroblasts and epithelial cells, frnely minced mammary tissue was subjected to 2 hours enzymatic dissociation with 0.1% collagenase at 37 C. The cell suspension was centrifirged twice at 80 x g for 30 sec and the pooled supematants were the source of the fibroblasts. To obtain cultures enriched for mammary epithelial cells, the pellet remaining after the removal of fibroblasts was further dissociated with 0.05% pronase for 20 min at 37 C. The cell suspension was then purified by percoll density gradient centrifugation. All cell types were plated at a density of 4.2 x 105 viable cells/ cm2. Epithelial cells and mixed cultures were plated at 1.4 x 105 viable cells/well in a 96-well dish, and fibroblasts were plated at 9.5 x105 viable cells lwell in a 4- well dish. Epithelial cells and mixed cells were plated on rat-tail collagen-coated (collagen I) dishes to enhance cell attachment. Fibroblasts were plated on tissue- culture plastic. After 2 hours of plating, fibroblast cultures were washed 3 times with HBSS to remove any residual epithelial cells. The basal medium for epithelial and mixed cultures contained phenol red-free DME/F 12 (1:1), Hepes (15 mM), penicillin (0.1 mg/ml), streptomycin (0.05 mg/ml), insulin (0.] pg/ml) and 78 fetuin (0.1 mg/ml). The basal medium for fibroblast cultures contained phenol red-free DME/F 12 (1:1), Hepes (15 mM), penicillin (0.1 mg/ml), streptomycin (0.05 mg/ml), insulin (1, 0.1 pg/ml), hydrocortisone (0.2 pM) and fetuin (0.5 mg/ml). Media was changed every day. Cells were kept at 37 0C in 5% C02 for 6 days. Hormones 17 B-estradiol (E, 23nM), R5020 (promegestone; 7,21- dimethyl-19-nor-4,9-pregna-diene-3,20-dione; 10nM) and grth factors EGF (lng/ml) and IGF-1 (25ng/ml) were added to the media individually or in combinations 24 hours after plating. The concentrations of 23nM, 10nM and 25ng/ml were used for 17 B-estradiol, R5020 and IGF-1 respectively, based on their ability to produce a maximal proliferative effect in a dose response study (data not shown). A low dose (lng/ml) of EGF was used because it has been previously reported that a higher dose of EGF obscures interactions with other hormones (Imagawa et al., 1985). Quantitation of Cell Growth: Cell number was quantified using a modified 3-(4,5- dimethylthiazol-Z-yl)- 2,5-diphenyltetrazolium bromide (MTT) assay. The assay is based upon the conversion of tetrazoliurn into blue formazan by the mitochondrial enzyme succinate-dehydrogenase. The color production is directly related to viable cell number (Hahm & IP, 1990). Briefly, 5 days after addition of hormone- and growth factor- supplemented media, phenol red-free DME/F12 media containing 1 mg/ml MTT was added to the cultures for 4 h at 37 ° C. Cells were then fixed with formal-saline (10% formaldehyde in 0.85% NaCl) for 30 min to prevent cell detachment and loss and then the MTT-formol-saline supernatant was removed. Isopropanol (100%) was then added to each well to dissolve the formazan crystals and absorbance at 540 nm was measured using a multiwell spectrophotometer Bio- tek EL 310 reader. 79 Immunocytochenristry: Cells of mesenchymal origin express the intermediate filament, vimentin (Miettinen et al., 1984). Anti-vimentin antibody has been used previously to identify mammary fibroblasts in cell culture and in tissue sections (Asch et al., 1981; Haslam & Levely, 1985). Therefore to establish the purity of epithelial cell and fibroblast cultures and the ratio of epithelial cells to fibroblasts in mixed cultures, the fibroblasts were identified immunocytochemically using goat anti- vimentin antibody (a gift of Dr. B. B. Asch, Roswell Park Memorial Institute). Fibroblasts were cultured on glass coverslips; epithelial cells and mixed cells were cultured on collagen-coated coverslips. Cells were fixed at -20°C in absolute methanol for 5 min and then in acetone for 5 min. Endogenous peroxidase activity was blocked with 0.1% hydrogen peroxide in methanol. Nonspecific antibody binding was blocked with normal rabbit serum (1:100 dilution). Cells were then incubated with anti-vimentin antisertun (1:50 dilution) for 1 hour at room temperature. Control coverslips were incubated with normal goat serum (1:50 dilution). After rinsing three times with PBS, cells were incubated with a secondary antibody, peroxidase labeled rabbit antigoat IgG (1:3200 dilution) (Cappel Research Products, Durham, NC) for 2 hours at room temperature. Peroxidase activity was detected by incubating 3 minutes with 3,3'- diaminobenzidine (25 pg/ml) in hydrogen peroxide (0.01%). Quantitation of epithelial cells and fibroblasts in mixed cell cultures: Epithelial cells and fibroblasts in the mixed cell cultures which had been treated with ovarian steroid hormones or growth factors were quantitated, to determine their relative efi'ects on proliferation of different cell types. After MTT assay, isopropanol was removed from culture dishes and cells were stained with methylene blue (1%) for five minutes. The percentage of epithelial cells and 8O fibroblasts in mixed cell cultures was determined by counting cell nuclei of cells identified to be epithelial cells or fibroblasts based upon morphological criteria under an inverted microscope. Three samples of 5000 cells from two experiments were counted for each treatment. Steroid hormone binding assay: The methods for measuring specific [3 H]estradiol binding in intact cells has been described previously (Haslam & Levely, 1985). Cells grown in culture dishes were washed three times with 37°C HBSS (Hanks' balanced salt solution). Basal culture medium containing various concentrations of [3H]estradiol (1-10 nM) (142.0 Ci/mmol., New England Nuclear) with or without a 100-fold excess of unlabeled estradiol was added to duplicate culture wells and incubated at 37°C for 1 h. The binding reaction was terminated by the addition of ice-cold HBSS, followed by three additional washes. Cells were then extracted with 100% ethanol at room temperature for 10 minutes, and radioactivity was determined by counting dupliCate aliquots of the ethanol extracts in scintillation fluid consisting of toluene, 30% Triton X-100, and 6 g/ liter Omnifluor (New England Nuclear). All samples were counted in a Beckrnan LCS-7000 liquid scintillation counter (Beckrnan, Palo Alto, CA) with a 50% counting efficiency. The amount of specifically bound estradiol was determined by subtracting the amount of [3H]-estradiol bound in the presence of excess unlabeled estradiol from that bound in its absence. To measure specific progestin-binding sites, the same procedure was used, except that cells were incubated with [3H]R5020 (1-13 nM) (84.8 Ci/mmol., New England Nuclear) with or without a 100-fold excess of unlabeled R5020. Since R5020 can also bind to glucocorticoid-binding sites in mammary gland, total binding was measured in the presence of a 100-fold excess of unlabeled 81 dexamethasone. All binding data was analyzed by the method of Scatchard (Scatchard G, 1947). BLults 1. Cell culture morphology Epithelial cells were plated as small organoids of aggregated cells. The cell clumps grew out into colonies containing polygonal-shaped cells and cultures reached confluence about four days after plating. Epithelial cells from nulliparous and mid-pregnant mice were morphologically similar. After cell dissociation, fibroblasts existed as large round cells and had a granular surface. Two hours after plating, fibroblasts attached to the culture dish and had two to four cytoplasmic processes. Fibroblast cultures reached confluence by three or four days after plating. No morphological differences were observed between fibroblast cultures of nulliparous and mid-pregnant mice. Anti-vimentin antibody binding was observed only in fibroblast cultures (results not shown) indicating that epithelial cultures contained only epithelial cells. Nulliparous mixed cultures contained two types of epithelial cell organoids. One type was a round clump of cell aggregates; the second type resembled small pieces of mammary ducts. The duct-like epithelial organoids usually attached but failed to grow and form colonies. In contrast, mid-pregnant mixed cultures contained mostly the round type of epithelial organoids which grew out into colonies. Mixed cultures also contained fibroblasts (epithelial cells : fibroblasts = 70 : 30; see Chapter Three). Epithelial cells grew in colonies and fibroblasts as single cells which surrounded the epithelial colonies. In confluent mixed cultures, epithelial cells were in contact with fibroblasts. 82 Addition of the hormones E and R5020 as well as the growth factors EGF and IGF-1 to all the three types of cell cultures from nulliparous and mid- pregnant mice did not change cell morphology. 2. The effects of E, R5020, EGF and IGF-1 on cell proliferation of cell cultures of mammm cells obtained from nulliparous mice E, R5020 and EGF were added alone or in combination to fibroblast, epithelial and mixed cultures from nulliparous mice as shown in Fig. 21. The growth of fibroblasts was not stimulated by any of these hormones or growth factors either alone or in different combinations (Fig. 21C). Epithelial cells were stimulated by the progestin R5020, and EGF; there was no finther stimulatory effect when E, R5020 and EGF were combined (Fig. 21B). Mixed cells were stimulated by E + R5020, EGF, but not R5020 alone (Fig. 21A). Thus in mixed culture there was a loss of the R5020 stimulatory effect. It is possible that this loss of a R5020 effect was due to the dilution of the epithelial cell population by fibroblasts which did not respond to R5020; Alternatively, the presence of fibroblasts may have altered epithelial cell hormonal responses. Fig. 22 shows the effects of IGF-I and its interactions with E, R5020 and EGF. Fibroblasts were not stimulated by IGF-I alone or in any combination with E, R5020 or EGF (Fig. 22C). In epithelial cultures either IGF-I Cr EGF stimulated epithelial cell proliferation. When IGF-I, E, R5020 and EGF were combined there was no further stimulatory effect (Fig. 22B). In the mixed cultures IGF-I and EGF also stimulated cell growth (Fig. 22A). The stimulatory effect of IGF-I plus EGF in mixed cultures, albeit higher than IGF-I or EGF alone, was not statistically significantly (Fig. 22A). In mixed cell cultures, the stimulatory effects of E+R5020, EGF and IGF-1 on epithelial cells vs. fibroblasts have been analyzed to determine which cell types 83 Figure 21. Effects of E, R5020, and EGF alone or in different combinations on cell proliferation of mixed, epithelial cell and fibroblast cultures from nulliparous mice. Mixed and epithelial cells were plated on collagen, and fibroblasts were plated on plastic in basal media (C), or plus E 23nM, R5020 (R) 10nM, EGF lng/ml. Each point represents the mean 1: SEM of triplicate determinations from two to four experiments. * p = 0.05 that cell numbers in hormone or growth factor treated group are greater than in control groups. 84 Effects of E, R5020, EGF on Cell Cultures of Nulliparous Mice 5 .. A Mixed ., . t 4 .. H H H H Ell-Q. 3.. 0M 0" 2.I. 1.. O B Epithelial 4" o . c ‘ 0“. HI. 3.. DU" .01 OM D". 2.. Cell Number (trio-'5) C Fibroblast 3.. 24.. m “I' NH 'l‘ O". I'M C E R ER EGF EGFE EGF'R , EGFER Hormone Treatment 85 Figure 22. Effects of IGF-I and EGF alone or in different combinations with E and R5020 on cell proliferation of mixed, epithelial and fibroblast cell cultures from nulliparous mice. Mixed and epithelial cells were plated on collagen, and fibroblasts were plated on plastic in basal media (C) or, plus IGF-I 25ng/ml, E 23nM, R 10nM, or EGF lug/ml. Each point represents the mean 1- SEM of triplicate determinations from two to three experiments. "' p = 0.05 that cell numbers in hormone or growth factor treated groups are greater than in control group. 86 Effects of E, R5020. EGF and IGF—I on Mammary Cells of Nulliparous Mice 5" A Mixed ‘1 I. . a 4" . t . . . I I I 3-- r I z I ‘ 2.. I 1... 0 t6" BEpithelial l o 4" 3 t " e . I. " " . e r. 3" . r j r r I m L L ‘ L e - a E 1.. ‘55 0 o CFibroblost 4.. 3... r 1- f L 2., a 3 .L L I J. L I 5 1d- c IGF' IGFE IGFR IGFER IGF IGFE IGFR IGFER EGF EGF EGF EGF EGF Hormone Treatment 87 Table 2. Differential effects of R5020, E+R5020, EGF, and IGF-1 on epithelial cells and fibroblasts in mixed cell cultures from nulliparous mice. Cell Number (x 105) Treatmenta Ratio of the % Total Epithelial Fibroblast Epithelial Cells: Cellsb Cells Fibroblasts Control 1.28 i 0.10 0.83 fig 0.04 0.45 i 0.04 65:35 R5020 1.43 1; 0.10 1.00 i 0.03* 0.43 i 0.05 69:31 E+R5020 1.66 p: 0.05* 1.10 i 006* 0.56 i 003* 66:34 EGF 1.82 j; 009* 1.46 i 0.09”“ 0.36 i 0.08 80:20 IGF-I 1.66 i 0.11"“ 1.29 i 0.09* 0.37 i 0.08 78:22 a Cells were cultured as described in Materials and Methods for 5 days in the presence of basal medium (control), R5020 (10 nM), E+R5020 (23 nM+10 nM), EGF (1 ng/ml) or IGF-I (25ng/ml). 9 Cells were counted with the aid of ocular grid and a microscope. Each point represents the mean : SEM of triplicate determinations from three experiments. * p < 0.05 that there is a significant difference from the control group. were affected. Although R5020 did not significantly stimulate mixed culture proliferation overall, the epithelial cell and fibroblast composition was also analyzed to determine if R5020 stimulated epithelial cells or if the presence of fibroblasts abrogated the epithelial cell response. As shown in Table 2, R5020 increased epithelial cell proliferation 1.2-fold but had no effect on fibroblasts. This is the same degree of stimulation that was observed in epithelial cultures 88 (Fig. 21). Thus, it appears that the presence of non-responsive fibroblasts diluted the epithelial cell response to R5020. E+R5020 increased both epithelial cell and fibroblast numbers, but did not change the ratio of epithelial cells and fibroblasts compared to the control. Therefore, E+R5020 appears to stimulate the proliferation of both epithelial cells and fibroblasts in the mixed cultures. In contrast, EGF and IGF-1 preferentially increased only the numbers of epithelial cells, and the total cell number increases were accompanied by increases in epithelial cell number and slight decreases in fibroblasts. Therefore the growth effect of EGF and IGF-1 in mixed cultures was most likely due to increased epithelial cell proliferation. 3. The effects of E, R5020, EGF and IGF-1 on cell proliferation of cell cultures of mammruy cells obtained from mid-preguant mice E, R5020 and EGF were added alone or in various combinations to fibroblast, epithelial and mixed cultures fiom mid-pregnant mice as shown in Fig. 23. The growth of fibroblasts was not stimulated by any of these hormones, growth factors alone or in their different combinations (Fig. 23C). In contrast to nulliparous cell cultures, epithelial cells were not stimulated by R5020 and mixed cells were not stimulated by E + R5020. Epithelial and mixed cells were only stimulated by EGF, and there was no further stimulatory effect when E, R5020 and EGF were combined (Fig. 23B, 23A). These results indicate that at different stages of mammary gland development, the hormones responsible for growth control may change. Fig. 24 shows the effects of IGF-I alone and in combination with E, R5020 and EGF. As can be seen in Fig. 24C, fibroblasts did not respond to any of the treatments. Epithelial cells were stimulated by IGF-I or EGF, and there was no increased stimulatory effect when IGF-I and EGF were combined (Fig. 24B). E 89 Figure 23. Effects of E, R5020, and EGF alone or in different combinations on cell proliferation of mixed, epithelial cell and fibroblast cultures from pregnant mice. Mixed and epithelial cells were plated on collagen, and fibroblasts were plated on plastic in basal media (C),or plus E 23nM, R5020 (R)10nM, EGF lng/ml. Each point represents the mean 1 SEM of triplicate determinations from two to three experiments. * p = 0.05 that cell numbers in hormone or growth factor treated groups are greater than in control group. 90 Effects of E, R5020, EGF on Mammary Cells of Pregnant Mice 5 "- A Mixed 4, .. 3T- -I—‘ 2" D”. it. 1.. 0 B Epithelial 4 .. 3'1” 0 ' ‘ fill-0. H". 2:.- T T NC 0‘ Cell Number (trio—5) C Fibroblast 31b 2.. l-rl H we F-I N we C E R ER EGF EGFE EGFR EGFER Hormone Treatment 91 Figure 24. Effects of IGF-I and EGF alone or in different combinations with E, R5020 on cell proliferation of mixed and epithelial cells from pregnant mice. Mixed and epithelial cells were plated on collagen and fibroblasts were plated on plastic in basal media (C) or plus IGF-I 25ng/ml, E 23nM, R 10nM, EGF lng/ml. Each point represents the mean : SEM of triplicate determinations from two to three experiments. * p = 0.05 that cell numbers in hormone or growth factor treated groups are greater than in confiol group. 92 Effects of E, R5020, EGF, lGF-l on Mammary Cells of Pregnant Mice Cell Number (x10-5) 4..- 3.. 2.. 5 T A Mixed Oil-I H 4.. 3m- 2dr- B Epithelial H fl“ 4.. 3.. C Fibroblast I"! C IGF IGF IGFE IGFR IGFER IGFE IGFR IGFER EGF EGF EGF EGF EGF Hormone Treatment 93 and R5020 appeared to be inhibitory to the effect of IGF-I in epithelial cell cultures. In the mixed cultures only EGF stimulated cell growth (Fig. 24A). IGF-I failed to stimulate mixed cell cultures. It is possible that the presence of fibroblasts in the mixed cultures either diluted or abrogated the epithelial cell response. The proliferative effect of EGF and IGF-1 on epithelial cells vs. fibroblasts in the mixed cell culture was examined further. As shown in Table 3, the presence of IGF-I slightly increased epithelial cell number. Thus, it is possible that the epithelial cell response was not detectable due to dilution by non- responsive fibroblasts, and therefore IGF-I did not show an overall stimulatory effect in mixed cultures. In contrast, EGF caused a significant increase in the ratio of epithelial cells to fibroblasts. Thus, increase in total cell number was accompanied by an increase in epithelial cells and a decrease in fibroblasts, but the decrease of fibroblasts was not significantly different from the control group. Thus these results indicate that the growth effect of EGF in mixed cultures was due to its ability to specifically increase epithelial cell proliferation. 4. Steroid hormone receptor studies The biologic effects of steroid hormones are believed to be mediated by steroid receptors. ER and PR have been detected in mammary cells jump (Haslam & Shyamala, 1981) and iumrp (Edery et al., 1985; Haslam & Levely, 1985). In the present studies, in serum-free media, R5020 stimulated proliferation of nulliparous epithelial and mixed cell cultures. E alone did not show any cell proliferative effects, and E or R5020 or E+R5020 failed to stimulate the growth of cell cultures from mid-pregnant mice. To investigate whether the observed effects of E and R5020 are directly related to the presence 94 Table 3. Differential effects of EGF and IGF-1 on epithelial cells or fibroblasts in mixed cell cultures from pregnant mice. Cell Number (x 105) Treatmenta Ratio of the % Total Epithelial Fibroblast Epithelial Cells: Cellsb Cells Fibroblasts Control 1.8 i 0.15 1.26 i 0.07 0.54 i 0.06 70:30 IGF-I 1.9 i 0.22 1.52 i 0.06“ 0.38 i 0.06 80:20 EGF 2.3 i 018* 1.90 i 010* 0.39 i 0.05 83:17 a Cells were cultured as described in Materials and Methods for 5 days in the presence of basal medium (control), IGF-I (25ng/ml) or EGF (1 ng/ml). b Cells were counted with the aid of ocular grid and a microscope. Each point represents the mean : SEM of triplicate determinations from three experiments. "' p < 0.05 that there is a significant difference from the control group. or absence of steroid hormone receptors, ER and PR status was examined in cultured mixed and epithelial cells from nulliparous nd mid-pregnant mice. The results obtained demonstrated that there were single classes of high affinity specific E and P binding sites in mixed and epithelial cell cultures fi'om both nulliparous and mid-pregnant mice after 5 days of culture in serum-free basal media (Table 4). Both epithelial and mixed cultures from nulliparous mice contained about 2-fold more ER than cells from mid-pregnant mice. PR content was similar in cultured cells from both nulliparous and mid-pregnant mice. Since ER and PR were present in epithelial and nrixed cultures from both nulliparous and mid-pregnant mice, the lack of cell proliferative response of nulliparous cells to E and mid-pregnant cells to E or R5020 does not appear to 95 Table 4. Specific 3H-estradiol and 3H-R5020 binding sites in mixed and epithelial cell cultures from nulliparous and mid-pregnant mice 3H-Estradiol Binding 3H-R5020 Binding frnol/mg DNA Kd (nM) pmol/mg DNA Kd (nM) Nulliparous: Mixeda 815 1.5 3.56 5.3 Epithelialb 800 1.02 3.42 5.3 Mid-Pregnant: Mixed 417 2.5 3.22 4.6 Epithelial 428 2.8 3.12 4.1 a Mixed: Epithelial cells and fibroblasts mixed cultured together. b Epithelial: Epithelial cell culture. be due to the lack of the ER or PR in the cell cultures. In the In vivo studies described in Chapter Two, E (5 ng) induced a 1.5-fold increase of .PR in nulliparous mice, and this increase of PR led to the synergistic stimulatory effects on E+R5020 in epithelial cell proliferation. To determine whether E was able to increase PR in the present in_vi;r_p serum-free conditions, E inducible PR were examined in nulliparous epithelial and mixed cell cultures. As shown in Table 5, E increased PR about 1.5-fold in both epithelial and mixed cell cultures of nulliparous mice. In nulliparous epithelial cultures, 96 Table 5. Regulation of 3H-R5020 binding sites in mixed and epithelial cell cultures from nulliparous and mid-pregnant mice. Specific 3H-R5020 Binding (pmol/mg DNA) Nulliparous Mid-Pregnant Treatmenta Mixedb Epithelialc Mixed Epithelial Control 3.26:0.11 2821031 33035091 3.42:0.63 E 5.17:0.15“ 3.95:0.18* 4.58:0.80 4.13:0.68 IGF-I 6.01:0.52“ 4.10:0.09“ 4.83_+_0.50 5.26:0.88 IGF-I+E 53510.50“ 4.27:0.56" 52810.78 53950.52 H+Prl+I 4.91:0.37“ _ 4.24_«_r_0.08* 5.14:0.66 5.09:0.63 H+Prl+I+E 65510.61" 5.01:0.28“ 4.74:0.13 4.81:0.59 a E 23nM, IGF-I 25ng/ml, I 6pg/ml, H 10nM, Prl lpg/ml were added in indicated combinations to the cultures afier 24 hours platting. The receptor binding assay was performed on day five of the culture. b Mixed: Epithelial cells and fibroblasts in mixed cultures. 9 Epithelial: Epithelial cell culture. * p < 0.05 that the treatment groups are higher than the control group. B did not show a proliferative effect, while R5020 stimulated cell proliferation. The effect of E+R5020 was not higher than R5020 alone. Thus the observed E- induced increase in PR did not enhance epithelial cell proliferation. In the mixed cultures, E alone had no effect, and R5020 alone stimulated only epithelial cell proliferation. E+R5020 stimulated cell proliferation of both 97 epithelial and fibroblasts. The stimulatory effect of E+R5020 on epithelial cells was about the same as R5020 alone. Thus, no synergistic effect was observed. This implies that E regulated PR may not play any role in E+R5020 mediated stimulatory effect on epithelial cells in the mixed culture. In mid-pregnant epithelial and mixed cell cultures, E regulation of PR concentration was also examined. As shown in table 5, E did not significantly increase PR. Previously, in studies of mixed and epithelial cells of mid- pregrrant mice cultured in media supplemented with serum, PR were increased 1.8 to 20- fold by E (Haslam & Levely, 1985; Haslam, 1986). However in those studies the culture medium was also supplemented with hydrocortisone (H), prolactin (Prl) and a high level insulin (1) (6 pg/ml) (Haslam & Levely, 1985; Haslam, 1986). To determine if the lack of H+Prl+I in the media was the cause of the lack of the E induction of PR in the present serum-free conditions, PR concentration was examined in epithelial and mixed cells cultured in presence of H+Prl+I or H+Prl+I+E. The supplementation with H+Prl+I without E enhanced PR concentration in mixed and epithelial cell cultures from nulliparous but not fi'om mid-pregnant mice (Table 5). The magnitude (1.4 to 1.6-fold) of the increase in PR was about the same as the E-mediated increase in nulliparous cultures, and E did not show any further increase in PR when combine with H+Prl+I (Table 5). Thus, it does not appear that the lack of H+Prl+I was responsible for the lack of E induction of PR in mid-pregnant cells cultured in serum-free media herein. IGF-I has been shown to increase PR levels in the MCF—7 human breast carcinoma cell line (Cho etal., 1994). To investigate if IGF-I can regulate PR in our cell culture system, PR concentration was determined in epithelial and mixed cells cultured in media supplemented with IGF-I. As can be seen in Table 5, IGF-I increased PR levels in nulliparous, but not mid-pregnant 98 epithelial and mixed cell cultures. There was no further increase in PR when IGF-I combined with E. In this regard, since a high concentration of I (6 pg/ml) was used in H+Prl+I treatment group, it is possible that the increase in PR is being mediated by insulin acting through the IGF-I receptor pathway (Imagawa et al., 1986). In summary, in epithelial and mixed cell cultures from nulliparous mice, PR can be regulated by E, H+Prl+I and IGF-1. PR in cell cultures from mid- pregnant mice were not increased by any of the hormones and growth factors tested. These results indicate that there are differences in hormonal regulation of PR concentration in nulliparous vs. mid-pregnant mammary cells. Discussion 1. Hormonal regulation of cell proliferation: 1. studies in nulliparous mammary gland: a. my studies: E and P are known to be involved in mammary growth and differentiation. Ovariectomy of adult nulliparous mice, which removes the major source of E and P, caused regression of the mammary ductal tree (N andi 1958). Using a DNA historadiography technique, it has been shown that injection of E can stimulate DNA synthesis in epithelial cells at the end of mammary ducts (Bresciani, 1968). Injection of P stimulated DNA synthesis in the epithelial cells of the duct proper. Injection of both E and P increased DNA synthesis in ductal epithelial cells and led to the formation of side-branching. Previously, in immature mice, by implanting hormones directly into mammary gland E was demonstrated to increase epithelial DNA synthesis via a local effect in mammary gland (Haslam, 1988). In contrast, the ability of E to increase 99 epithelial DNA synthesis in adult nulliparous mice appears to be systemically mediated. In the present study, by implanting a small dose of P alone or combined with E directly into mammary gland, local vs. systemic effects of P have been distinguished. This is based on the observation that in the P-implanted mammary gland, epithelial DNA synthesis was much higher than the contralateral control-implanted mammary gland. E+P synergistically stimulated epithelial cell proliferation, and this effect was also locally mediated via an increase in E-induced PR. In stromal tissue, DNA synthesis was stimulated by E, P and E+P. However, these stimulatory effects appeared to be systemically mediated. These results indicate that different mechanisms of P action may be operating in the two types of mammary tissue. b. m studies in serum-supplemented media: Much effort has been devoted to study the mechanisms of action of E and P on nulliparous mammary gland i_u_\gitr_o, so that the m systemic influences and the direct hormone effects can be separated. Previously in organ culture, both E and P have been demonstrated to stimulate rat mammary epithelial tissue growth (Koyama et. al., 1972). E, P and Prl together stimulated proliferation of rat mammary epithelial cells cultured in collagen gels (Pasco et al., 1982). However, in that study, the effects of individual hormones were not evaluated. Since those studies were done in serum-containing media, the interpretation of hormonal effects is complicated by the presence of serum factors. c. In vitro studies in serum-free media: By studying the effects of E and P in serum-free conditions, the complications of serum factors are eliminated, and the direct influence of E and 100 P on target cells can be characterized. In the case of cells from nulliparous mice, P or Prl were shown to stimulate mammary epithelial cells cell growth in collagen gels, and P+Prl had a much higher effect than P or Prl alone. E had no effect either alone or in combination with P and Prl (Imagawa et al., 1985), Since in vivo epithelial cells exist in association with fibroblasts, fibroblasts may also play a role in some of the hormonal effects on epithelial cells (McGrath, 1983; Haslam, 1986). Thus it is necessary to examine the effects of E and P in epithelial cell cultures and also in mixed cell cultures containing both cell types in serum-free conditions. In the present studies, a serum-free monolayer cell culture condition has been developed to culture mammary epithelial, fibroblast and mixed cultures containing both epithelial cell and fibroblasts. Under this culture condition, P stimulated proliferation of epithelial cells in both epithelium-enriched and mixed cell cultures. E alone did not exhibit any cell proliferative effect. These results are in agreement with the previous in vivo observation that E-mediated increase of epithelial DNA synthesis in adult nulliparous mice appears to be systemically mediated, and involves unknown factors. P alone stimulated epithelial cell proliferation iu vitro in serum-free media. This result supports the in vivo observation that P can act directly to stimulate mammary epithelial cell growth. Neither E nor P or E+P stimulated growth in fibroblast-enriched cultures. However, in the mixed cultures, fibroblast growth was stimulated by E+P. This result implies that the presence of epithelial cells is probably necessary for E+P-induced fibroblast proliferation. As mentioned above, an indirect growth stimulatory effect of E on both epithelial and stromal cells and P in stromal cells has been indicated from the i_u vivo and in vitro studies. To investigate the possible mediators for these E and P effects, EGF and IGF-1 have been examined herein in vitro, in serum-free 101 cultures. EGF and IGF-1 both stimulated epithelial cell proliferation, however, neither EGF nor IGF-I stimulated the growth of fibroblasts. Thus, it is possible that neither EGF nor IGF-I are the in vivo growth mediators of E or P in stromal cells. It is also possible that EGF or IGF-I may not be sufficient to mediate E and P effect, and that additional factors may be required. Furthermore, EGF and IGF-1 receptor status needs to be examined in cultured fibroblasts to see if the lack of the responses are due to the lack of the respective receptors. Even though EGF and IGF-1 have shown to have direct stimulatory effects in mammary epithelial cell cultures, further studies are needed to determine if they are responsible for the in vivo E-mediated cell proliferation. 2. Studies in mid-pregnant mammary gland: a. In_viv_o_ studies: Little is known about the mechanisms of E and P action in pregnant mammary gland m. E+P+Prl have been identified as the minimum requirement for lobuloalveolar development, a characteristic of pregnancy, in hypophysectomized+ovariectomized+adrenalectomized mice (N andi, 1958). b. _In_vitrp studies in serum-supplemented media: Previously, E was demonstrated to be able to stimulate proliferation of epithelial cells and stromal cells from early pregnant mice when the two types of cells are in contact (McGrath, 1983). In another study, mammary epithelial cells fiom mid-pregnant mice were cultured on collagen I coated surface, or in fibroblast conditioned media, or co-cultured with dead or live mammary fibroblasts. Only in presence of live fibroblasts did E stimulate cell proliferation, and the proliferative effect of E was on both epithelial cells and fibroblasts (Haslam, 1986). Thus, the presence of fibroblasts that are 102 metabolically active seems to be important for E regulated epithelial cell proliferation. c. In vitro studies in serum-free media: In serum-free culture conditions, Nandi's group had previously found that epithelial cells from mid-pregnant mice were only slightly stimulated by P, and that E had no effect (Imagawa et al., 1985). In the present studies, experiments were carried out to investigate the effects of E, P and E+P in cultured epithelial, fibroblast and mixed cell cultures from mid-pregnant mice. E, P and E+P did not show any significant proliferative effects in any types of cell cultures. In the mixed cultures where two type cells were present, E failed to stimulate cell proliferation. Thus, the proliferative effect of E which had been demonstrated previously in serum-supplemented media may involve the presence of serum factors. It should also be noted that the hormones 1, HC and Prl were also present in the serum-supplemented media in those previous studies. Whether or not these hormones were involved in E-mediated cell proliferation needs further investigation. In contrast to nulliparous cells, P did not stimulate proliferation of mid-pregnant epithelial cells in serum-free media. It is possible that under the present culture conditions, mid-pregnant cells are more differentiated, and are more oriented to differentiated fimctions such as secretion of milk, and have less proliferative potential overall. Another difference in E and P effects on epithelial cells from mid-pregnant mice and nulliparous mice was that E and/or P inhibited IGF-I effects on epithelial cells from mid-pregnant mice but not from nulliparous mice. Thus, E and P act differently on epithelial cells of mid- pregnant vs nulliparous mice. 103 II. Steroid hormone receptors: 1. Studies in nulliparous mammary gland: The effects of steroid hormones are mediated through steroid receptors. I_n uiv_o, adult mouse mammary glands contain ER and PR. ER and PR are present in both epithelial and stromal tissue (Haslam & Shyamala, 1981). PR in the epithelial compartment can be regulated by B. As demonstrated in Chapter Two, E (5 ng) implanted into the mammary gland increased PR about 1.5-fold in vivo and this increase of PR led to the synergistic cell proliferative effect of E+P. In this study in serum-free media, ER and PR were detected in epithelial and mixed cultures. E (23 nM) induced an 1.5-fold increase in PR. However, no cell proliferative effect mediated through E—induced PR was observed. Thus, it is possible that some component of the extracellular microenvironment which exists Lam may be necessary for the biologic effect of E-induced PR and may not be present in vitro. 2. Studies in mid-pregnant mammary gland: During pregnancy, there is a decrease in ER and PR. The cause of the decrease in ER is not clear. The decrease in PR might be due to the negative feedback of elevated P concentration (Haslam & Shyamala,1980). In ovariectomized and hysterectomized pregnant mice, E was able to regulate PR concentration in the epithelial compartment. In in_viug studies, most of the work on ER and PR of mid-pregnant mammary gland were examined in serum-supplemented media. ER and PR were present in mixed cultures, and ER was functional as judged by induction of PR (Haslam & Levely, 1985). However, in epithelial cell cultures, no E induced PR were demonstrated until epithelial cells were cultured on collagen I coated surface, with fibroblast conditioned media, or on a layer of dead or live 104 fibroblasts. When cultured in those conditions PR were E inducible, and the basal level of PR was significantly increased. Thus, the substratum laid down by fibroblasts and/or soluble factors present in conditioned media appear to be important for the maintenance and synthesis of PR. In the present studies, ER and PR were examined in epithelial and mixed cells cultured on collagen surface in serum-free media. In this culture condition, ER and PR were present in epithelial and mixed cultures of mid-pregnant mice, however, these cell cultures were less sensitive to E regulation of PR compared to nulliparous cell cultures. This result is different from the results obtained fi'om i_n_ M studies and from ur_v_it_rp studies in serum-supplemented conditions. Thus, it is possible that some factor is missing in serum-free conditions that is required for E regulation of PR in mammary epithelial cells of the pregnant stage. Another possibility is that under these culture conditions (i.e. collagen I culture substratum and absence of serum) that the cells are expressing a more differentiated phenotype and resemble more closely the lactational state. In this regard, it has been previously shown in_viv_o during lactation, that the mammary gland is refractory to estrogenic stimulation of cell proliferation and PR regulation (Shyamala & Haslam, 1980). Summg: In this study, by culturing epithelial and fibroblasts separately and culturing them together, the influence of epithelial cell and fibroblast interaction on specific hormone- and growth factor-mediated effects has been demonstrated. Specifically, E+R5020 stimulated proliferation of nulliparous fibroblasts only in presence of epithelial cells. Furthermore, by comparing cells from nulliparous and mid- pregnant mice, differences in hormonal regulation of cells at different developmental stages have also been identified. The PR level in nulliparous 105 epithelial cells could be regulated by E or H+Prl+I or IGF-I, whereas the regulatory effect of these hormones and growth factors on PR level in mid- pregnant epithelial cells was less pronounced. P directly stimulated proliferation of nulliparous epithelial cells only. In epithelial cells from mid-pregnant mice, no direct cell proliferative effect of P was demonstrated. In contrast to the effects of steroid hormones, the proliferative effects of the growth factors, EGF and IGF-1, on cells of nulliparous and mid-pregnant mice were similar. In view of these findings, it is possible that this culture system will help us to identify the . mechanisms underlying cell-cell interactions and their relation to hormone- and growth factor regulated events. Ultimately this may allow us to identify changes in cell-cell interactions which occur in breast cancer and influence tumor growth and invasion. Summary and Conclusions In this study the effects of E and P on mammary growth and the relative effects on epithelial and stromal cells have been investigated Mp. A serum-free monolayer culture system has been established for culturing mammary epithelial cells, fibroblasts and the mixed cells containing both epithelial cells and fibroblasts. The direct effects of E and P on epithelial cells and fibroblasts have been examined i_u_uitr_o in these serum-free monolayer cell cultures. The influence of epithelial cell and fibroblast interactions on the effects of E and P has been examined in epithelial cell or fibroblast cultures and in the mixed cell cultures where both epithelial cells and fibroblasts are present. In in vivo studies I found that: 1. The proliferative effects of E on nulliparous epithelial cells in vivo are likely mediated systemically, indicating that E may not be a direct mediator for epithelial cell proliferation. 2. The proliferative effects of P on nulliparous epithelial cells are likely a direct effect. 3. E+P synergistically stimulated nulliparous epithelial cell proliferation, and this synergistic effect is likely due to an E induced increase in PR in epithelial cells. 4. The proliferative effects of E or P on nulliparous stromal cells in vivo is likely mediated systemically, indicating that E or P may not be the direct mediators for stromal cell proliferation. Correspondingly in vitro: 1. E did not stimulate epithelial cells to proliferate. Since in vitro studies examine direct effect of hormones, this result supports the in vivo observation that B may not be the direct mediator for epithelial cell proliferation. 106 107 2. In agreement with in vivo findings, P stimulated the proliferation of nulliparous epithelial cells. Thus, a direct effect of P on mammary epithelial cell proliferation is further supported. 3. E+P did not show synergistic cell proliferative effect in nulliparous epithelial cells. One possible explanation of this lack of a synergistic effect is the reduced induction of PR by E in cells in serum-free media. In addition, some component of the extracellular microenvironment which exists in vivo may be necessary for the biologic effect of E-induced PR, and may not be present in_vi_t;g. 4. E or P did not stimulate fibroblasts to proliferate. This result also supports the m findings and suggests that E or P may not be the direct mediators for stoma] cell proliferation in_vflg. EGF or IGF-I stimulated epithelial cell proliferation, however, EGF and IGF-1 failed to stimulate fibroblast Cultures. These results indicate that EGF or IGF-I may not be the in vivo mediators of E or progestin on stoma] cell proliferation. Alternatively, there might be other factors required for the E or P responses in cell culture. From the ul_vi_tp studies I also observed that epithelial-stoma] cell interactions influence hormonal or growth factor responsiveness. For example, E+P stimulated fibroblasts in nulliparous mixed cell cultures, but not in fibroblast enriched cell cultures, therefore the presence of epithelial cells is necessary for this hormone regulated response. I also found that the developmental state of the gland (nulliparous vs. pregnant) can modulate hormonal responsiveness. The proliferation of pregnancy- delived epithelial cells and fibroblasts was only stimulated by growth factors, and failed to be stimulated by steroids. This did not appear to be due to the lack of ER or PR. The lack of responsiveness to E and P suggests that increased differentiation results in a loss of responsiveness to E+P. 108 In conclusion, this study has advanced our understanding of the growth regulation of mammary epithelial cells and fibroblasts by E and P, and increased our knowledge about the growth contol of mammary cells at different developmental stages. By separately culturing epithelial cells and fibroblasts which coexist in mammary ducts m, the differential effects of E, P and growth factors EGF and IGF-1 on epithelial cells and fibroblasts have been identified. The influence of epithelial cell-fibroblast interactions on E, P and growth factor-mediated effects has been demonstated by comparing the effect of E, P and growth factors on mixed cell cultures containing both epithelial cells and fibroblasts to epithelial or fibroblast enriched cell cultures. The changes in mammary grth conto] at different developmental stages have been revealed by comparing the effect of E, P and EGF and IGF-1 on mammary cells from nulliparous and mid-pregnant mice. This study opened an avenue for future studies toward discovering the effects of hormones and growth factors on other types of mammary cells, such as adipocytes, and the influence of interactions of different cell types on hormone and growth factor mediated events. In addition, the findings herein on growth regulation of normal mammary cells are of value by providing reference for identifying abnormal changes in regulation of E, P and growth factors and cell-cell interactions in hormone and growth factor mediated effects in breast cancer. Finally, since breast tumors increase with age and have their highest incidence in menopause, an understanding of the changes in growth conto] by hormones and growth factors on mammary cells at different developmental stages might be helpful in the further development of endocrine stategies for breast cancer teatnent and prevention. References Adams EF, Newton CJ, Tait GH, Braunsberg H, Reed M], James VH (1988) Paracrine influence of human breast stroma] fibroblasts on breast epithelial cells: Secretion of a polypeptide which stimulates reductive 17 beta-oestradiol dehydrogenase activity. Int. J. Cancer 42(1): 1 l9 Aitken SC, Lippman ME (1983) Hormonal regulation of de novo pyrimidine synthesis and utilization in human breast cancer cells in tissue culture. Cancer Res. 43:4681 Aitken SC, Lippman ME (1985) Effects of estrogens and antiestrogens on growth- regulatory enzymes in human breast cancer cells in tissue culture. Cancer Res. 45:16]] Ances IG (1973) Serum concentration of epidermal growth factor in human pregnancy. Amer. J. Obstet. Gynecol. 115:357 Asch BB, Burstein NA, Vidrich A, Sun TT (1981) Identification of mouse mammary epithelial cells by immunofluorescence with rabbit and guinea pig antikeratin antisera. Proc. Natl. Acad. Sci. USA 7825643 Bandyopadhyay G, Imagawa W, Wallace D, Nandi S (1987) Linoleate metabolites enhance the in vitro proliferative response of mouse mammary epithelial cells to epidermal grth factor. J. Biol. Chem. 26222750 Banerjee MR (1976) Responses of mammary cells to hormones. Int. Rev. Cyto]. 46:1 Banerjee MIL Wood BG, Lin FK, Crump LR (1976) Organ culture of whole mammary gland of the mouse. Tissue Culture Assoc. Manual 2:457 Bartley JC, Emerrnan JJ, Bissell MJ (1981) Metabolic cooperativity between epithelial cells and adipocytes of mice. Am. J. Physiol. 2411204 Bates SE, Davidson NE, Valverius EM, Freter CE, Dickson RB, Tam JP, Kudlow IE, Lippman ME, Salomon DS (1988) Expression of transforming growth factor or. and its messenger ribonucleic acid in human breast cancer: Its regulation by estrogen and its possible significance. Mol. Endocrinol. 2:543 Beck JC, Hosick HL (1988) Growth of mouse mammary epithelium in response to serum- fi'ee media conditioned by mammary adipose tissue. Cell Biol. Int. Rep. 12:85 Bemfield MR, Banerjee SD, Koda IE, Rapraeger AC (1984) Remodeling of the basement membrane as a mechanism of morphogenetic tissue interaction. In: The Role of Extracellular Matrix in Development. Eds Trelstad RL, New York: A.R. Liss, p 545 109 110 Berthois Y, Katzenellenbogen J A, Katzenellenbogen BS (1986) Phenol red in tissue culture media is weak estrogen : implications concerning the study of estrogen-responsive cells in culture. Proc. Natl. Acad. Sci. USA 83:2496 Bowden RT, Hissom JR, Moore MR (1989) Growth stimulation of T47D human breast cancer cells by the anti-progestin RU486. Endocrinology 12422642 Bresciani F (1968) Topography of DNA synthesis in the mammary gland of the C3H mouse and its control by ovarian hormones. Cell Tissue Kinetics 1:51 Brown JB (1981) Hormone profiles in young women at risk of breast cancer. In: Hormones and Breast Cancer. Eds Pick MC, Sitteri PK, Welsch CW; Banbury Report 8, Cold Spring Harbor Laboratory, Cold Spring Harbor, p33 Butler WB, Kirkland WL, Jorgensen TL (1979) Induction of plasminogen activator by estrogen in a human breast cancer cell line (MCF-7). Biochem. Biophys. Res. Commun. 90: 1328 Carmichael J, DeGraff WG, Gazdar AF, Minna JD, Mitchell JB (1987) Evaluation of a tetrazolium-based semiautomatic colorimetric assay: assessment of chemosensitivity testing. Cancer Res. 472936 Ceriotti GA (1952) A microchemical determination of deoxyribonucleic acid. J. Biol. Chem. 198:297 Chablos D, Vignon F, Keydar I, Rochefort H (1982) Estrogens stimulate cell proliferation and induce secretory proteins in a human breast cancer cell line ( T47D). J. Clin. Endocr. Metab. 552276 Chablos D, Rochefort H (1984) Dual effects of progestin R5020 on proteins released by the T47D human breast cancer cells. J. Biol. Chem. 259: 1231 Cho H, Aroniea SM, Katzenellenbogen BS (1994) Regulation of progesterone receptor gene expression in MCF-7 breast cancer cells: a comparison of the effects of cyclic adenosine 3', 5'-monophosphate, estradiol, insulin-like growth factor-1, and serum factors. Endocrinology 1342658 Coleman S, Silberstein GB, Danie] CW (1988) Ductal morphogenesis in the mouse mammary gland: Evidence supporting a role for epidermal growth factor. Dev. Biol. 1272302 Daniel CW, Berger JJ, Strickland P, Garcia R (1984) Similar growth pattern of mouse mammary cells cultivated in collagen matrix in vivo and in vitro. Dev. Biol. 104257 1]] Daniel CW, Silberstein GB, Strickland P, (1987) Direct action of estradiol on mouse mammary ducts analyzed by sustained release implants and steroid autoradiography. Cancer Res. 4726052 Darbre P, Yates J, Curtis S, King RJB (1983) Effect of estradiol on human breast cancer cells in culture. Cancer Res. 43 2349 Dennizo F, Lang R (1986) Rapid colorimetric assay for cell growth and survival modifications to the tetrazolium dye procedure giving improved sensitivity and reliability. J. Immunol. Methods 892271 Decks S, Richards J, Nandi S (1988) Maintenance of normal rat mammary epithelial cells by insulin and insulin-like growth factor. Exp. Cell Res. 1742448 DeLarco EJ, Todaro GJ (1978) Growth factors from murine sarcoma virus transformed cells. Proc. Nat]. Acad. Sci. USA 752400] Dickson RB, Huff KK, Spencer EM, Lippman ME (1986) Induction of epidermal growth factor-related polypeptides by 17B-estradiol in MCF-7 human breast cancer cells. Endocrinology 1182138 Dickson RB, McManaway ME, Lippman ME (1986) Estrogen-induced factors of breast cancer cells partially replace estrogen to promote tumor growth . Science 3221540 Dubik D, Demboiniski TC, Shiu RP (1987) Stimulation of c-myc oncogene expression associated with estrogen-induced proliferation of human breast cancer cells. Cancer Res. 4726517 Durning P, Schor SL, Sellwood RA (1984) Fibroblasts from patients with breast cancer show abnormal migratory behaviour in vitro. Lancet 2(8408)2890 Edery M, Imagawa W, Larson L, Nandi S (1985) Regulation of estrogen and progesterone receptor levels in mouse mammary epithelial cells grown in serum-free collagen gel cultures. Endocrinology 1162105 Edery M, McGrath M, Larson L, Nandi S (1984) Correlation between in vitro growth regulation of estrogen and progesterone receptors in rat mammary epithelial cells. Endocrinology 1 15: 1691 Edery M, Pang k, Larson L, Colosi T, Nandi S (1985) Epidermal grth factor receptor levels in mouse mammary glands in various physiological states. Endocrinology 1172405 Edwards DP, Murphy SR, McGuire WL (1980) Effect of estrogen and antiestrogen on DNA polymerase in human breast cancer. Cancer Res. 4021722 112 Emerman JT, Enami J, Pitelka DR, Nandi S (1977) Hormonal effects on intracellular and secreted casein in cultures of mouse mammary epithelial cells on collagen membranes. Proc. Natl. Acad. Sci. USA 7424466 Enami J, Enami S, Koga M (1983) Growth of normal and neoplastic mouse mammary epithelial cells in primary culture: stimulation by conditioned medium from mouse mammary fibroblasts. Gann 742845 Feil PD, Glasser SR, Tofi DO, O’Mally BW (1972) Progesterone binding in the mouse and rat uterus. Endocrinology 912738 Fisher HW, Puck TT, Sato G (1958) Molecular growth requirements of single mammalian cells: the action off fetuin in promoting cell attachment to glass. Proc. Natl. Acad. Sci. USA 4424 Haggie JA, Sellwood RA, Howell A, Birch JM, Schor SL (1987) Fibroblasts from relatives of patients with hereditary breast cancer show fetal-behaviour in vitro. Lancet 1(8548): 1455 Hahm HA, [P M. (1990) Primary culture of normal rat mammary epithelial cells within a basement membrane matrix. I. regulation of proliferation by hormones and growth factors. In Vitro Cell. Dev. Biol. 262791 Haslam SZ, Shyamala G (1979) Effect of oestradio] on progesterone receptors in normal mammary glands and its relationship to lactation. Biochem. J. 1822127 Haslam SZ, Shyamala G (1980) Progesterone receptors in normal mammary gland: receptor modulations in relation to differentiation. J. Cell Biol. 862730 Haslam SZ, Shyamala G (1981) Relative distribution of estrogen and progesterone receptors among the epithelial, adipose, and connective tissue components of the normal mammary gland. Endocrinology 1082825 Haslam SZ, Levely ML (1985) Estrogen responsiveness of normal mouse mammary cells in primary cell culture: association of mammary fibroblasts with estrogenic regulation of progesterone receptors. Endocrinology 1 1621835 Haslam SZ (1986) Mammary fibroblast influence on normal mouse mammary epithelial cell responses to estrogen in vitro. Cancer Res. 462310 Haslam SZ (1987) Role of sex steroid hormones in normal mammary gland function. In: The Mammary Gland: Development, Regulation and Function. Eds Neville MC, Daniel CW; Plenum Press, New York, p449 113 Haslam 82 (1988) Progesterone effects on DNA synthesis in normal mouse mammary glands. Endocrinology 1222464 Haslam SZ (1988) Acquisition of estrogen-dependent progesterone receptors by norrna] mouse mammary gland. Ontogeny of mammary progesterone receptors. J. steroid Biochem. 31(1):9 Haslam $2 (1988) local versus systemically mediated effects of estrogen on normal mammary epithelial cell deoxyribonucleic acid synthesis. Endocrinology 122:860 Haslam SZ (1989) The ontogeny of mouse mammary gland responsiveness to ovarian steroid hormones. Endocrinology 12522766 Haslam SZ (1991) Stromal-epithelial interactions in normal and neoplastic mammary gland. In: Regulatory Mechanisms in Breast Cancer. Eds Lippman, M.; Dickson R., Kluwer Academic Publishers, Boston, p40] Haslam SZ, Counterman LJ (1991) Mammary stroma modulates hormonal responsiveness of mammary epithelium in vivo in the mouse. Endocrinology 12922017 Haslam SZ, Counterman LJ, Nummy KA (1992) EGF receptor regulation in normal mouse mammary gland. J. Cell. Physiol. 1522553 Haslam SZ, Nummy KA (1992) The ontogeny and cellular distribution of estrogen receptors in norrna] mouse mammary gland. J. Steroid Biochem. Molec. Biol. 422589 Haslam SZ, Counterman LJ, Nummy KA (1993) Effects of epidermal grth factor, estrogen, and progestin on DNA synthesis in mammary cells in vivo are determined by the developmental state of the gland. J. Cell. Physiol. 155272 Hayman EG, Barnes D, Pierschacher M, Engvall E, Ruoslshti E (1982) Serum spreading factor is a major attachment protein in fetal bovine serum and the active principle in attachment-promoting fetuin. (Abstract 5038.) J. Cell Biol 95: 121a Hissom JR, Bowden RT, Moore MR (1989) Effects of progestins, estrogens, and antihorrnones on growth and lactate dehydrogenase in the human breast cancer cell line T47D. Endocrinology 1252418 Huff KK, Lippman ME (1984) Hormonal control of plasminogen activator secretion in ZR-75—1 human breast cancer cells in culture. Endocrinology 11421665 Ichinose RR, Nandi S (1964) Lobuloalveolar differentiation in mouse mammary tissues in vitro. Science 1452496 114 Ichinose RR Nandi S (1966) Influence of hormones on lobuloalveolar differentiation of mouse mammary glands in vitro. J. Endocrinology 35:33] Imagawa W, Tomooka Y, Nandi S (1982) Serum-free growth of normal and tumor mouse mammary epithelial cells in primary culture. Proc. Natl. Acad. Sci. USA 7924074 Imagawa W, Tomooka Y, Hamamoto S, Nandi S (1985) Stimulation of mammary epithelial cell growth in vitro: interaction of epidermal grth factor and mammogenic hormones. Endocrinology 1 16: 1514 Imagawa W, Spencer M, Larson L, Nandi S (1986) Somatomedin-C substitutes for insulin for the growth of mammary epithelial cells fiom normal virgin mice in serum-free collagen gel cell culture. Endocrinology 11922695 Imaguma Y, Kusakabe M, Mackie EJ, Pearson CA, Chiquet-Ehrismann R, Sakakura T (1988) Epithelial induction of stromal tenacin in the mouse mammary gland from embryogenesis to carcinogenesis. Dev. Biol. 1282245 Kaufman U, Zapf J, Torretti B, Froesch ER (1977) Demonstration of a specific serum carrier protein of nonsuppresible insulin-like activity in viva. J. Clin. Endocr. Metab. 442160 Koyama H, Sinha D, Dao TL (1972) Effects of hormones and 7, 12-dimethyl-benz- [a]anthracene on rat mammary tissue grown in organ culture. J. Natl. Cancer Inst. 4821671 Kurachi H, Oka T (1985) Changes in epidermal growth factor concentration of submandibular gland, plasma and urine of normal and sialoadenectomized female mice during various reproductive stages. J. Endocrinology 1062172 Levine JF, Stockdale FE (1984) 3T3-L1 adipocytes promote the growth of mammary epithelium. Exp. Cell Res. 15121 12 Levine JF, Stockdale FE (1985) Cell-cell interactions promote mammary epithelial cell differentiation. J. Cell Biol. 100(5): 1415 Lippman ME, Bolan, Huff K (1976) The effects of estrogens and antiestrogens on hormone-responsive human breast cancer in long term culture. Cancer Res. 3624595 Lippman ME, Dickson RB, Bates S, Knabbe C, Huff K, Swain S, McManaway M, Bronzert D, Kasid A, Gelmann EP (1986) Autocrine and paracrine grth regulation of human breast cancer. Breast Cancer Res. Treat. 1259 Liotta LA, Rao CN, Weiner UM (1986) Biochemical interactions of tumor cells with the basement membrane. A. Rev. Biochem. 5521037 115 McGrath CM (1983) Augmentation of the response of normal mammary epithelial cells to estradiol by mammary stroma. Cancer Res. 43‘: 1355 Masters JRW, Drife JO, Scarisbreck JJ (1977) Cyclic variation of DNA synthesis in human breast epithelium. J. Natl. Cancer Inst. 5821263 Mathieu M, Vignon F, Capony F, Rochefort H (1991) Estradiol down-regulates the mannose-6-phosphatefrnsulin-like growth factor-II receptor gene and induces cathepsin D in breast cancer cells: a receptor saturation mechanism to increase the secretion of lysosomal proenzymes. Mol. Endocrinol. 5:815 Moguilewsky M, Philibert D (1984) RU3 8486: potent antiglucocorticoid activity correlated with strong binding to the cytosolic glucocorticoid receptor followed by an impaired activation. J. Steroid Biochem. 202271 Mukku VR, Stancel G (1985) Regulation of epidermal growth factor receptor by estrogen. J. Biol. Biochem. 26029820 Murphy LC, Murphy LJ, Dubik D, Bell GI, Shiu RPC (1988) Epidermal frowth factor gene expression in human breast cancer cells: regulation of expression by progestins. Cancer Res. 48:4555 Murphy LC, Murphy LJ, Shiu RPC (1988) progestin regulation of EGF-receptor mRNA accumulation in T-47D human breast cancer cells. Biochem. Biophys. Res. Commun. 1 50: 192 Nandi S (1958) Endocrine control of mammary-gland development and fimction in the C3H/HeCrgl mouse. J. Natl. Cancer Inst. 2121039 Nandi S, Bern GA (1960) Relation between mammary-gland responses to lactogenic hormone combinations and tumor susceptibility in various strains of mice. J. Natl. Cancer Inst. 242907 Nandi S, Imagawa W, Tomooka Y, Shiurba R, Yang J (1981) Grth requirements for mammary epithelial cells in primary culture in the presence and absence of serum: effects of in viva mammogenic hormones. In: Cold Spring Harbor Conferences on Cell Proliferation, Book B. Eds Pardee AB, Sato GH, Sirbasku DA, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, vol. 9-79 Oka T, Tsutsumi O, Kurachi H, Okamoto S (1988) The role of epidermal growth factor in normal and neoplastic growth of mouse mammary epithelial cells. In: Breast Cancer: Cellular and Molecular Biology. Eds Lippman ME, Dickson R. Kluwer Academic Publishers, Boston, p343 116 Okamoto S, Oka T (1984) Evidence for physiological firnction of epidermal growth factor: Pregestational sialoadenectomy of mice decreases milk production and increases offspring mortality during lactation period. Proc Natl Acad Sci., USA, 8126059 Orrnerod EJ, Warburton MJ, Gusterson B, Hughes CM, Rudland PS (1985) Abnormal deposition of basement membrane and connective tissue components in dimethylbenzan- thracene-induced rat mammary tumors: An immunocytochemical and ultrastructural study. Histochem. J. 17 (10):1155 Osborne CK, Arteaga CL (1990) Autocrine and paracrine growth regulation of breast cancer: clinical implications. Breast Cancer Res. Treat. 15:3-11, 1990 Papa V, Hartmann KKP, Rosenthal SM, Maddux BA, Siiteri PK, Goldfine II) (1991) Progestins induce down-regulation of insulin-like grth factor-I (IGF-I) receptors in human breast cancer cells: potential autocrine role of IGF-II. Mol. Endocrinol. 52709 Pasco D, Quan A, Smith S, Nandi S (1982) Efi‘ect of hormones and EGF on proliferation of rat mammary epithelium enriched for alveoli. An in vitro study. Exp. Cell Res. 1412313 Pearlman WH, Peng LH, Pearlman MR] (1981) In vitro uptake and metabolism of [3H]corticosterone by mammary glands from pregnant, lactating and post-lactational rats and by parametrial adipose tissue from lactating rats. J. Endocr. 91281 Pearson CA, Pearson D, Shibakara S, Hofsteenge J, Chiquet-Ehrismann R (1988) Tenascin: cDNA cloning and induction by TGF-or EMBO J. 1022977 Philbert D, Raynaud JP (1974) progesterone binding in the immature rabbit and guinea pig uterus. Endocrinology 942627 Philbert D (1984) RU3 8486: an original multifaceted antihorrnone in viva. In: Adrenal Steroid Antagnism. Eds Walter de Gruyter & Co, Berlin, p77 Ries LAG, Hankey BF, Miller BA, Hartman AM, Edwards BK (1991) Cancer statistics review 1973-1988. National Cancer Institute. NIH Publ. No. 91-2789, Bethesda, MD. Rozengurt E, Sinnett-Smith J, Taylor-Papadimitriou J (1985) Production of PDGF-like growth factor by breast cancer cell lines. Int. J. Cancer 36:246 Rudland PS, Durbin H, Clingan D, Jimenez de Asua L (1977) Iron salts and transferrin are specifically required for cell division of cultured 3T6 cells. Biochem. Biophys. Res. Comm. 75:556 117 Salomon DS, Bano M, Smith KB, Kidwell WR (1982) Isolation and characterization of a grth factor (embryonin) from bovine fetuin which resembles A2_macroglobulin. J. Biol. Chem. 257214093 Salomon DS, Smith KB, Losonczy I, Bano M, Idwell WR, Alessandri G, Gullino PM (1984) (12 macroglobulin, a contaminant of commercially prepared Pedersen fetuin: isolation, characterization and biological activity. In: Cell Cult. Methods Mol. Cell. Biol. Eds Barnes DW, Sirbasku DA, Sato GI-I, Alan R, Liss Inc., New York, 3:125 Saceda M, Lippman ME, Chambon P, Lindsey RL, Ponglikitmongkol M, Puente M, Martin MB (1988) Regulation of the estrogen receptor in MCF-7 cells by estradiol. Mol. Endocrinology 221 157 Sakakura T, Nishizuka Y, Dawe CJ (1976) Mesenchyme-dependent morphogenesis and epithelium-specific cytodifi‘erentiation in mouse mammary gland. Science 19421439 Sakakura T (1987) Mammary embryogenesis. In: The Mammary gland: Development, Regulation, and Function. Eds Neville MC, Daniel CW. Plenum Press, New York, p 37 Sakakura T, Sakagumi Y, Nishizuka Y (1982) Dual origin of mesenchymal tissue participating in mouse mammary gland morphogenesis. Dev. Biol. 91 :202 Sappino AP, Skalli 0, Jackson B, Schurch W, Gabbiani G (1988) Smooth-muscle differentiation in stromal cells of malignant and non-malignant breast tissues. Int J Cancer 41(5)2707 Scatchard G (1947) The attractions of proteins for small molecules and ions. Ann. NY. Acad. Sci. 51:660 Scheffield LG, Welsch CW (1987) Influence of submandibular salivary glands on hormone responsiveness of mouse mammary glands. Proc. Soc. Exptl. Med. 1862368 Schor SL, Haggie JA, Duming P, Howell A, Smith L, Sellwood RA, Crowther D (1986) Occurrence of a fetal fibroblast phenotype in familial breast cancer. Int. J. Cancer 37(6):831 Shyamala G, Haslam S2 (1980) Estrogen and progesterone receptors in normal mammary gland during different firnctional states. In: Perspectives in Steroid Receptor Research. Eds Bresciani F, Raven Press, New York, p193 Siberstein GB, Daniel CW (1982) Elvax 4OP implants: sustained, local release of bioactive molecules influencing mammary ductal development. Dev. Biol. 93 2272 118 Soule HI), McGrath CM (1980) Estrogen responsive proliferation of clonal human breast carcinoma cells in athymic mice. Cancer Lett. 10: 177 Stewart AJ, Johnson MD, May FEB, Westley BR (1990) Role of insulin-like growth factors and the type I insulin-like growth factor receptor in the estrogen-stimulated proliferation of human breast cancer cells. J. Biol. Chem. 265221172 Stockdale FE, Wiens DJ, Levine JF (1986) Cell-cell interactions in the mediation of hormone dependent differentiation of mammary epithelium. In: Progress in Developmental Biology, Part B. Eds Slavkin HC, Alan R. Liss, New York, p 417 Sumida C, Pasqualini JR (1989) Antiestrogens antagonize the stimulatory effect of epidermal grth factor on the induction of progesterone receptor in fetal uterine cells in culture. Endocrinology 124:591 Thompson EW, Reich R, Martin GR, Albini A (1988) Factors regulating basement membrane invasion by tumor cells. In: Breast Cancer: Cellular and Molecular Biology. Eds Lippman ME, Dichson RB; Kluwer, Boston, p239 Tonelli QB, Sorof S (1980) Epidermal growth factor requirement for development of cultured mammary glands. Nature 2852250 Topper YJ, Oka T (1974) Some aspects of mammary gland development in the mature mouse. In: Lactation. Eds Larson BL, Smith VR, Academic Press, New York, 1:327 Vrgnon F, Bouton MM, Rochefort H (1987) Antiestrogens inhibit the mitogenic effect of growth factors on breast cancet cells in the total absence of estrogens. Biochem. Biophys. Res. Commun. 14621502 Vonderhaar BK, Greco AE (1979) Lobulo-alveolar development of mouse mammary glands is regulated by thyroid hormones. Endocrinology 1042409 Vonderhaar BK (1984) Hormones and growth factors in mammary gland development. In: Control of Cell Growth and Proliferation. Eds Veneziale CM, Van Nostrand-Reinhold, Princeton, NJ, p11 Wang S, Counterman LJ, Haslam ZS (1990) Progesterone action in normal mouse mammary gland. Endocrinology 12722183 Wei LL, Krett NL, Francis MD, Gordon DF, Wood WM, O'Malley BW, Horwitz KB (1988) Multiple human progesterone receptor messenger ribonucleic acids and their autoregulation by progestin agonists and antagonists in breast cancer cells. Mol. Endocrinol. 2262 119 Wicha MS, Liotta LA, Kidwell WR (1979) Effects of fiee fatty acids on the growth of normal and neoplastic rat mammary epithelial cells. Cancer Res. 392426 Wiens D, Park CS, Stockdale FE (1987) Milk protein expression and ductal morphogenesis in the mammary gland in vitro: hormone-dependent and -independent phases of adipocyte-mammary epithelial cell interaction. Dev. Biol. 1202245 Yee D, Cullen KJ, Paik S, Perdue JF, Hampton B, Schwartz A, Lippman ME, Rosen N (1988) Insulin-like growth factor II mRN A expression in human breast cancer. Cancer Res. 4826691 HICHIGRN STATE UNIV. LIBRARIES llHIlull"llllllllllllllll"lllllllllllllllmlllllllll 31293010201725