UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Hormone and growth factor effects on the growth of human mammary epithelial cells in serum-free primary.. Gabelman, Brett Michael 1992

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


ubc_1992_spring_gabelman_michael.pdf [ 4.74MB ]
JSON: 1.0087541.json
JSON-LD: 1.0087541+ld.json
RDF/XML (Pretty): 1.0087541.xml
RDF/JSON: 1.0087541+rdf.json
Turtle: 1.0087541+rdf-turtle.txt
N-Triples: 1.0087541+rdf-ntriples.txt
Original Record: 1.0087541 +original-record.json
Full Text

Full Text

LUMBIATHE Uto the required standardAPRIL, 1992HORMONE AND GROWTH FACTOR EFFECTS ON THE GROWTH OF HUMAN MAMMARYEPITHELIAL CELLS IN SERUM-FREE PRIMARY CULTUREbyBRETT MICHAEL GABELMANB. Sc. (Hons. Biol & Chem), University of British Columbia, 1988A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(DEPARTMENT OF ANATOMY)We accept this thesis as conforming©Brett Michael Gabelman, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of AN^m The University of British ColumbiaVancouver, CanadaDate  A e ,,\^1992_DE-6 (2/88)THESIS ABSTRACTThe ovarian steroid 17f3-estradiol (E2) is critically involved in the growth control of bothnormal and malignant mammary epithelial cells (MEC) in vivo. However, it has not yet beendetermined if E2 directly stimulates the growth of MEC, or if it stimulates growth via productionof locally acting autocrine and paracrine growth factors. Epidermal growth factor (EGF) andtransforming growth factor-a (TGF-a) are both peptide growth factors which interact with theEGF receptor (EGFR) to stimulate the growth of MEC in vivo and in vitro. E2-stimulatedgrowth in human breast cancer cell lines has been shown to be accompanied by increasedproduction of EGF, TGF-a and EGFR. In this thesis the effects of E2, EGF and TGF-a, aloneand in combination, on the growth of human MEC (HMEC) in primary culture were examined.HMEC from reduction mammoplasties, fibroadenomas and carcinomas were cultured oncollagen-coated dishes in serum-free medium (DME:F12 (1:1), 5 mg/ml BSA, 10 ng/ml choleratoxin, 0.5 µg/ml cortisol, 10 µg/ml insulin) in the presence and absence of E2, EGF and TGF-a.Tritiated-thymidine (3H-TdR) incorporation into DNA was used as a measure of cell growth.E2, at concentrations of 1-1000 nM, did not stimulate growth of any of the cultures examined inthe serum-free medium described above. However, E2 stimulated growth of 1 culture inmedium with a reduced insulin concentration (0.1 µg/ml). E2 inhibited the growth of HMEC insome cultures from all mammary tissue types examined. E2 effects on HMEC growth werestudied in cells grown on fibroblast feeder layers. E2 still failed to stimulate growth of the cells,but the growth-inhibitory effects of E2 differed in cells grown on collagen and fibroblasts. EGF,at concentrations of 1-100 ng/ml, consistently stimulated the growth of HMEC from allmammary tissue types examined. The EGF stimulation of growth was reduced by amonoclonal antibody (MAb 528) against the EGF receptor. TGF-a was equally or moreeffective in stimulating HMEC growth, although its dose response range was different than thatof EGF. E2 plus EGF synergized in the stimulation of HMEC growth in 33% of the samplesexamined. These studies suggest that E2 alone under the conditions used cannot directlystimulate the growth of HMEC in primary culture. However, E2 can exert effects on HMECgrowth via modulation of the cells' response to EGF.IITABLE OF CONTENTSTHESIS ABSTRACT^ iiLIST OF TABLES viiLIST OF ILLUSTRATIONS ^ viiiLIST OF ABBREVIATIONS ixACKNOWLEDGMENTS ^ xCHAPTER 1. INTRODUCTION 11. Growth and Differentiation in the Normal Mammary Gland ^ 1la. Anatomy of the Mammary Gland ^ 1lb. Summary of the Hormonal Control of Mammary Gland Growth andDifferentiation ^ 3lc. In vivo Evidence for the Role of E2 in Growth and Differentiation of theMammary Gland^ 7ld. In vitro Evidence for the Role of E2 in Growth and Differentiation of theMammary Gland ^ 101 e. Mechanism of E2 Action ^ 14lf. Role of Epidermal Growth Factor (EGF) and Transforming GrowthFactor-a (TGF-a) in Control of Mammary Gland Growth andDifferentiation 162. Involvement of E2 and Epidermal Growth Factor in Breast Cancer ^ 212a. Estrogens and Breast Cancer ^ 222b. EGF and TGF-a in Breast Cancer 26III3.Thesis Objectives ^ 28CHAPTER 2. MATERIALS AND METHODS ^ 301. Sample Procurement and Assessment 302. Tissue Preparation and Freezing ^ 303. Dissociation Procedure ^ 314. Cell Culture ^ 325. Growth Studies 34CHAPTER 3. RESULTS ^ 351. The Effects of 178-estradiol (E2) on Growth of Human Mammary Epithelial Cells(HMEC) in Primary Culture ^ 35la. The Effects of E2 on the Growth of Primary Cultures of Normal HMECObtained From Reduction Mammoplasties ^ 381b. The Effects of E2 on the Growth of Primary Cultures of HMEC ObtainedFrom Fibroadenom as ^ 441c. The Effects of E2 on the Growth of Primary Cultures of HMEC FromER+ Mammary Carcinomas ^ 441d. Summary of the Effects of E2 on the Growth of HMEC in Serum-FreePrimary Culture ^ 44le. The Effects of Insulin on E2 Modulation of HMEC Growth in PrimaryCultures ^ 522. The Effects of Epidermal Growth Factor (EGF) and Transforming Growth Factor-a(TGF-a) on the Growth of HMEC in Primary Culture ^ 56IV2a. The Effects of EGF on the Growth of Primary Cultures of HMEC fromReduction Mammoplasties ^ 562b. The Effects of EGF and TGF-a on the Growth of Primary Cultures ofHMEC from Fibroadenomas ^ 562c. The Effects of EGF and TGF-a on the Growth of Primary Cultures ofHMEC from Carcinomas ^ 602d. The Effects of a Monoclonal Antibody to EGFR on Growth Stimulationby EGF ^ 673 : The Combined Effects of E2 and EGF on the Growth of Primary Cultures of HMECFrom Reduction Mammoplasties, Fibroadenomas and Carcinomas ^ 724 : The Effects of E2 and EGF on the Growth of Primary Cultures of HMEC Cultured onFeeder Layers of Mitomycin-C Treated Fibroblasts ^ 80CHAPTER 4. DISCUSSION ^ 851. The Effects of E2 on the Growth of HMEC ^ 852. The Effects EGF and TGF-a on the Growth of HMEC ^ 943. The Effects of E2 plus EGF on the Growth of HMEC 974. Conclusions and Future Directions ^ 98APPENDIX 1 : Transport Medium 101APPENDIX 2 : Freezing Medium ^ 101APPENDIX 3 : Dissociation Medium 101APPENDIX 4 : Attachment Medium ^ 101APPENDIX 5 : Preparation of Rat Tail Collagen ^ 102APPENDIX 6 : Preparation of Pooled Normal Human Serum ^ 102VAPPENDIX 7 : Serum-free Medium ^ 102BIBLIOGRAPHY ^ 103VILIST OF TABLESTable #^ Page #Table I : The effects of 1713-estradiol (E2) on the growth of primary cultures of HMEC fromreduction mammoplasties. ^ 39Table II : The effects of 17[3-estradiol (E2) on the growth of primary cultures of HMEC fromfibroadenomas ^ 45Table III : The effect of 1713-estradiol (E2) on the growth of primary cultures of HMEC fromcarcinomas. ^ 46Table IV : Summary of the effects of 1713-estradiol (E2) on the growth of primary cultures ofHMEC from the 3 mammary tissue types^ 51Table V : The effects of insulin and 17f3-estradiol (E2) on the growth of primary cultures ofHMEC from fibroadenomas and carcinomas. ^ 53Table VI : The effects of EGF and TGF-a on the growth of primary cultures of HMEC fromfibroadenomas ^ 59Table VII : The effects of EGF on the growth of primary cultures of HMEC from carcinomas ^63Table VIII : Summary of the effects of EGF and TGF-a on the growth of primary cultures ofHMEC from reduction mammoplasties, fibroadenomas and carcinomas ^66Table IX : The effects of 17(3-estradiol (E2) plus EGF on the growth of primary cultures ofHMEC from a reduction mammoplasty, fibroadenomas and carcinomas ^73VIILIST OF ILLUSTRATIONSFigure #^ Page #Figure 1 : The effects of time after the last medium change on the incorporation of 3H-TdR into primary cultures of HMEC from 2 fibroadenomas (Fig. la. FA 49, Fig.lb. FA 50). ^ 36Figure 2 : The effects of 17f3-estradiol (E2) on the growth of primary cultures of humanmammary epithelial cells (HMEC) from 2 reduction mammoplasties (Redn 10 &Redn 11) ^ 40Figure 3 : The effect of time in culture on E2 regulation of growth in primary cultures ofHMEC from 2 reduction mammoplasties (3a. Redn 8„ 3b. Redn 9). ^42Figure 4 : The effects of E2 on the growth of primary cultures of HMEC from 2fibroadenomas (FA 47 & 67) ^47Figure 5 : The effect of E2 on the growth of primary cultures of HMEC from 2 ER+mammary carcinomas (HMC 101 and 102). ^ 49Figure 6 : The effects of E2 and insulin on the growth of primary cultures of HMEC from afibroadenoma (FA 67). ^ 54Figure 7 : The effects of EGF on the growth of primary cultures of HMEC from 3 reductionmammoplasties (Redn 12, 13 and 14). ^ 57Figure 8 : The effects of EGF and TGF -a on the growth of primary cultures of HMEC from2 fibroadenomas ( FA 74 & 80 ). ^ 61Figure 9 : The effects of EGF and TGF-a on the growth of primary cultures of HMEC from2 fibroadenomas ( FA 74 & 80 ). ^ 64Figure 10 : The effects of EGF and TGF-a on the growth of primary cultures of HMECfrom 2 mammary carcinomas (HMC 140 & HMC 141). ^ 68Figure 11 : The effect of MAb 528 on growth stimulation by EGF. 70Figure 12 : The effects of E2 plus EGF on the growth of primary cultures of HMEC from 2fibroadenomas (Fig. 12a., FA 51, Fig. 12b. FA 54) ^ 74Figure 13 : The effects of E2 plus EGF on the growth of primary cultures of HMEC from amammary carcinoma (HMC 111). ^ 76Figure 14 : The effects of E2 plus EGF on the growth of primary cultures of HMEC from afibroadenoma (FA 41). ^ 78Figure 15 : The effect of E2 and EGF on the growth of primary cultures of HMEC from afibroadenoma (FA 87 : 11 div). ^ 81Figure 16 : The effect of E2 on the growth of primary cultures of HMEC from 2carcinomas (HMC 129 and HMC 130). ^ 83VIIILIST OF ABBREVIATIONSE2^1 7p-estradiolEGF^Epidermal growth factorEGFR^Epidermal growth factor receptorER^Estrogen receptor3H-TdR Tritiated thym idineHMEC^Human mammary epithelial cellsMAb^Monoclonal antibodyMEC^Mammary epithelial cellsTGF-a^Transforming growth factor-alphaIXACKNOWLEDGMENTSNumerous people are owed great thanks and acknowledgment for the completion of thisthesis. It is with both sincere thanks and gratitude that I acknowledge my supervisor Dr.Joanne Emerman. Her great interest and patience in this work at all stages is greatlyappreciated. It has certainly been a privilege to work with her and I look forward to continuingwith related research in her lab. I would also like to thank the other members of my committee,Drs. Nelly Auersperg, Ross MacGillivray and Steve Pelech for both their help and guidancethroughout this project.The technical assistance of Darcy Wilkinson has been an immeasurable help in thecompletion of this thesis, as has her advice in the keeping of records. I would like to thank andacknowledge her for all her help and assistance. I would also like to thank Shannon Wilson forher technical assistance and sharing an office with me.I would also like to thank the Department of Anatomy. It has proven to be an excitingand diverse department in which to both carry out research and all the other aspects ofgraduate student training.Finally, I would like to especially thank my family, without whose support and tolerance Iwould never have been able to complete or even begin this work.This work was supported by a grant to Dr. J.T. Emerman from the National CancerInstitute, and by a Research Fellowship to B.M. Gabelman from the Evelyn Martin MemorialFoundation for Non-Animal Research. In respect to the Evelyn Martin Foundation, I would liketo acknowledge that no animals were used for the work contained in this thesis.XCHAPTER 1 INTRODUCTION 1. Growth and Differentiation in the Normal Mammary GlandThe mammary gland represents an exciting biological model system in which to study themechanisms involved in growth and differentiation. Compared to many other organ systems,where the majority of growth occurs during the fetal and early childhood periods of life, themammary gland undergoes maximal growth and differentiation in the adult. More specifically,with the onset of ovarian hormone production at puberty there is a dramatic increase inmammary gland growth. An even greater surge of growth occurs at pregnancy. Throughoutpregnancy the mammary gland continues to grow and differentiate until the onset of lactation.The specific changes in the hormonal milieu accompanying the various stages of mammarygland development control growth and differentiation in a distinct manner. This will bediscussed with reference to experimental evidence from animal model systems and cell culturemodels where applicable.1 a. Anatomy of the Mammary GlandThe mammary gland is defined as a complex tubuloalveolar gland, the function of whichis the production and secretion of milk that both protects and nourishes breast-feeding infants.In terms of a functional description, the mammary gland is composed mainly of a connectivetissue stroma (90 % resting, 13 % during pregnancy) and an epithelial parenchyma (10 %resting, 87 % during pregnancy) [Russo and Russo, 1987]. The epithelial and stromalcomponents of the mammary gland are separated by a basement membrane, the molecularcomponents of which are produced by both the epithelial and the stromal cells [Kimata et al.,1985]. The specific extracellular matrix composition of the basement membrane plays acritical role in the control of mammary epithelial cell differentiation [Emerman et al., 1977;1Emerman and Pitelka, 1977]. The majority of the nonpregnant adult female mammary gland iscomposed of fibrous and fatty connective tissue that plays an important structural andsupportive role in the gland.The epithelial component of the mammary gland is responsible for the production andsecretion of milk. The human mammary gland consists of 15-20 major ducts called lactiferousducts, each of which connects with the body surface through individual openings at the nipple.The regions of the mammary gland drained by each of the lactiferous ducts are termed lobes.Prior to exiting at the nipple, each lactiferous duct becomes dilated to form a lactiferous sinus.The sinuses function in the storage of secretory products during lactation. The lactiferousducts (interlobar) drain several smaller ducts (intralobar) that drain still smaller ducts(interlobular), which end in specialized epithelial buds or "end-buds". It is these terminal endbuds which are ultimately responsible for growth and differentiation into both the ducts and thelobuloalveolar units, depending on the hormonal milieu [Bresciani, 1965; 1968].Lobuloalveolar units are composed of numerous alveoli and the small ductules draining them.These lobuloalveolar regions of the mammary gland, also called lobules, comprise thesecretory portion of each mammary gland lobe. The secretory units are composed of twotypes of epithelial cells. Secretory cells, which are hormonally stimulated to synthesize andsecrete milk, line the lumens of the glands. At the basal surface of these cells are specializedcontractile cells, myoepithelial cells, which contract under hormonal stimulation to cause milkejection.21b. Summary of the Hormonal Control of Mammary Gland Growth and Differentiation Prior to puberty, the rate of mammary gland growth parallels growth in the other portionsof the body. This is referred to as isometric growth. During puberty, under the influence ofestrogens produced by the ovaries, both the connective tissue cells and the ductal end-budcells undergo dramatic increases in growth rates. The active form of estrogen produced by theovaries is 1713-estradiol (E2), and this is the estrogen that will be referred to throughout the restof this thesis. The growth at puberty is greater than the overall changes in body weight and isreferred to as allometric growth. The change in mammary gland size and shape with pubertyis due to increased proliferation of stromal cells, increased production of extracellular matrixand increased deposition of fat into mammary gland adipocytes. The increased growth of theductal epithelium contributes little to the change in size and shape of the mammary gland[Borellini and Oka, 1989].Growth in the epithelial component of the mammary gland during puberty is due to celldivision in the end-buds, leading to the elongation of the ductal system [Bresciani, 1965; 1968].Dichotomous branching also occurs at the level of the end-bud, and it is this branching that isresponsible for the tree-like pattern characteristic of the adult mammary ductal network.Growth and differentiation at the level of the end-buds also gives rise to structures known asalveolar buds. These buds will later develop into the functional secretory structures known asthe lobuloalveolar units or lobules [Russo and Russo, 1987].The lack of normal mammary gland development in ovariectomized animalsdemonstrates the critical requirement of ovarian hormones in this process. Replacement of E2in ovariectomized animals results in normal levels of ductal cell growth, whereas both E2 andprogesterone are required for normal development of the alveolar components of themammary gland [Tucker, 1974]. Variations in the rate of DNA synthesis by mammary gland3epithelial cells are observed during the menstrual cycle. By examining tritiated-thymidine ( 3H-TdR) incorporation using histoautoradiography, it has been shown that the DNA-labeling indexof mammary gland epithelium decreases during the follicular phase of the menstrual cyclewhen ovarian hormone levels are low [Masters et al., 1977; Meyer, 1977]. Also observed inthese studies was an increase in the labeling index of cells during the luteal phase, when E2and progesterone reach peak levels. Other researchers characterized the peak of mitoticdivision, as indicated by the labeling index, to occur at day 25 of the menstrual cycle [Fergusonand Anderson, 1981]. Following this peak at day 25, there is a peak in programmed celldeath, or apoptosis, at day 28. However, the level of monthly growth in the mammary gland isin excess of the level of apoptosis [Ferguson and Anderson, 1981]. This results in a netincrease in mammary gland development with each ovulatory cycle, until pregnancy ormenopause [Vorherr, 1977].With the onset of pregnancy there is an increase in the levels of both E2 andprogesterone. Although E2 in combination with either growth hormone or prolactin is sufficientfor stimulation of the growth of the ductal epithelium, E2 plus progesterone, in combination witheither growth hormone or prolactin leads to the massive increases of growth seen in thealveolar component of the mammary gland during pregnancy [Topper and Freeman, 1980].The degree of growth is such that by the end of pregnancy the epithelial tissue, both ductaland alveolar, has largely displaced the surrounding connective tissue. Thus, in contrast topuberty, the changes in mammary gland size during pregnancy are due largely to increases ingrowth of the epithelial cells.Accompanying the progression of pregnancy are increases in the level of prolactin, whichtogether with E2 and progesterone causes increased alveolar epithelial cell growth [Topperand Freeman, 1980]. Another function of prolactin is the induction of differentiation of alveolarend-buds into functional secretory lobuloalveolar units. However, the elevated levels of4progesterone directly opposes the secretory function of prolactin [Salazar and Tobon, 1974].Specifically, progesterone inhibits prolactin stimulation of both protein (casein) andcarbohydrate (lactose) secretion by the alveolar epithelial cells. As a consequence of theinhibition of secretion by progesterone, the secretory alveolar cells undergo significanthypertrophy and become extremely swollen with secretory products in the weeks prior toparturition [Baldwin and Daniels, 1974; Davis and Bauman, 1974].After parturition there is a sharp drop in the levels of both E2 and progesterone, whereasthe level of prolactin remains elevated. At this point, secretion of milk into the lumen occurs.The initial secretion produced by the lactating mammary gland is referred to as colostrum. Ithas an elevated level of maternal immunoglobulins, and functions in the transference ofpassive immunological protection from the mother to the newborn infant [Jenness, 1974].Colostrum secretion is limited to one week in humans and over a two to three week period oftransition the mammary gland secretion is altered to milk, which has lower levels ofimmunoglobulins but is rich in fats, sugars and proteins [Butler, 1974]. The differentiated,secretory state is maintained as long as the levels of prolactin remain elevated, which is aslong as breast feeding continues.The maintenance of prolactin levels during suckling involves a specific neuroendocrinearc [Grosvenor and Mena, 1974]. Suckling on the nipple activates the nervous system whichin turn acts on the hypothalamus to prevent the release of a prolactin-inhibiting factor whichtherefore permits secretion of prolactin in the anterior pituitary. Ejection of the milk through thenipple also involves interactions between the nervous system and the endocrine system,resulting in oxytocin release from the posterior pituitary stimulating contraction of themyoepithelial cells and forcing ejection of the milk.Following the termination of breast feeding, the epithelial portion of the mammary glandundergoes a massive regression due to apoptosis and reinfliltration with adipocytes, fibroblasts5and other connective tissue cells [Russo and Russo, 1987]. Numerous macrophages areresponsible for the digestion of the degenerate epithelial tissue. These same cycles of growthand differentiation are repeated with each subsequent pregnancy. It is important to note thatthe level of glandular development in the mammary gland after postlactational regressionremains greater than the levels observed in mammary glands of nulliparous women. Thepotential relevance of this observation is discussed later with regard to the epidemiology ofbreast cancer.The exact role of prolactin in the control of mammary gland growth and differentiation isunclear. Growth hormone, also produced in the anterior pituitary, has a 95 % cDNA and a 85% amino acid sequence identity to prolactin [Martial et al., 1979; Wallis, 1984]. The similarlocation of potential disulfide bridging cysteine residues further suggests that the proteins havevery similar secondary structures [Kohmoto et al., 1984]. Experimentally ovariectomized andhypophysectomized rodents require E2, progesterone and prolactin in order to undergo normallobuloalveolar development at pregnancy. However, in some strains of mice the requirementof prolactin can be replaced with growth hormone and normal growth and differentiation of themammary gland is observed during pregnancy in these mice. This ability of growth hormone tosubstitute for prolactin is a species-specific effect and varies even between different strains ofmice. These findings are difficult to extrapolate in terms of potential roles for these hormonesin normal development of the mammary gland [Imagawa et al., 1990].The circulating levels of prolactin in humans and rodents show cyclical variations duringthe ovulatory cycle. The peak levels of serum prolactin in humans are observed during themid-phase and luteal phase of the menstrual cycle [Vonderhaar, 1987b]. E2 has been shownto stimulate both hypertrophy and hyperplasia of pituitary lactotrophs that produce prolactin[Vonderhaar, 1987b]. As well, E2 stimulates the production of prolactin in ovariectomized rats,whereas progesterone inhibits this effect [Chen and Meites, 1970]. These two lines of6evidence suggest that the fluctuations in prolactin during the menstrual cycle are likely a directresult of the corresponding fluctuations in E2 and progesterone levels.A final major change in the growth and differentiation of the mammary gland occurs atmenopause. The cessation of ovulatory cycling and subsequent drop in circulating E2 andprogesterone levels leads to regression of the epithelial component of the mammary gland.The level of regression is such that the remaining epithelial portions of the gland are limited tothe large ducts and some of the secondary branches [Vorherr, 1974]. Accompanying thedecreases in epithelial content is the increased deposition of fat in adipocytes and an increasein the amount of fibrous connective tissue. This final stage in mammary gland development isof extreme importance from a clinical viewpoint as this is the stage in which the majority ofmammary gland neoplasms appear [Leis, 1978].1 e o f e 0 G e e iation o he mmar• 1 • ill 1 .1 • •Gland The readily apparent differences between mammary glands in male and female animalsis perhaps the first line of evidence suggesting a role for sex-linked hormones in the control ofmammary gland growth and development. Very early work in this field involved thetransplantation of ovarian tissue into male rodents and subsequent documentation of thefeminization of the male mammary gland [Engle, 1929; Gardner, 1935]. In the 30-40 yearsfollowing these pioneering studies, the majority of research investigating the regulation ofmammary gland growth and differentiation utilized four major strategies. These demonstratedthe role of E2 in the control of ductal mammary epithelial cell growth. These included:1. Administration of hormones to prepubescent animals to induce growth of theductal epithelium [Flux, 1954 a&b].72. Administration of hormones to animals that were endocrinectomizedprepubescently in order to determine which hormones could cause normal ductaldevelopment [Mumford, 1957; Richardson, 1955; Vonderhaar et al., 1978].3. Administration of hormones to mature endocrinectomized animals to determinewhich hormones were involved in maintenance of the ductal cells [Nandi, 1958;Traurig and Morgan, 1964].4. In animals in condition 3 above, regression of the mammary gland was allowed tooccur, and hormonal replacement was investigated [Ferguson 1956].Ovariectomy of nonpregnant adult mice results in a dramatic decrease of both ductal andalveolar components of the mammary gland. In these same mice, treatment with E2 alonecauses only ductal growth. However, if the mice are treated with both E2 and progesterone,there is growth of both the ductal and alveolar components of the mouse mammary gland[Bresciani, 1968; 1971]. Ductal growth is generally thought to be largely independent ofprogesterone [Sakakura and Nishizuka, 1967; Topper and Freeman, 1980]. However, veryhigh doses of progesterone can replace the standard combination of E2 and progesterone instimulating the growth of ductal and alveolar portions of the mammary gland in ovariectomizedmice [Seyle, 1940; Haslam, 1988a]. More recent studies have shown conclusively the ability ofE2 interaction with estrogen receptors (ER) to stimulate the production of progesteronereceptors directly via transcriptional regulation [Haslam and Levely, 1985]. These findings leadto the hypothesis that E2 may exert its growth effects via modulation of mammary epithelialcells responsiveness to other factors, such as progesterone. This hypothesis is consideredfurther in a later section.Haslam's group have also investigated the effect of implanting E2-releasing polymerpellets directly into the mammary glands of ovariectomized mice [Haslam, 1988b]. The result8of this treatment was the stimulation of end-bud cell growth. The details of this important studyare discussed further with respect to the mechanism of E2 action in Sec. 1e. Other groupshave also demonstrated the ability of E2 implants to stimulate ductal growth in animals whichhave been ovariectomized [Daniel et al., 1987]. The details of these studies are alsodiscussed in Sec. le. McManus and Welsch used subcutaneous implants of E2 to study theeffects of E2 on the growth of human mammary tissue transplanted into nude mice [McManusand Welsch, 1981 & 1984]. Their findings also demonstrated an E2 stimulation of growth inthe ductal epithelium in the human mammary tissue from noncancerous biopsies. In summary,it appears that E2 functions principally in the stimulation of ductal growth. This may occur bystimulation of cell growth directly, or by alteration of the cells responsiveness to other growthcontrolling factors such as progesterone or prolactin, as is discussed further in Sec. le.Another interesting observation has been the sharp rise in plasma E2 levels in mice justprior to parturition [McCormack and Greenwald, 1974; Shaikh, 1971]. This is followed by ashort burst of proliferation in the ductal epithelial cells observed immediately after parturition[Brookreson and Turner, 1959; Griffith and Turner, 1961]. This correlation also supports a rolefor E2 in the control of ductal cell growth. E2, as well as progesterone and either prolactin orgrowth hormone, is also required for proper growth of the lobuloalveolar portions of themammary gland during pregnancy [Nandi, 1958; Traurig and Morgan, 1964].In addition to its role in growth control, E2 may also have a role in differentiation of themammary gland. Using the C3H/He Crgl mouse strain, Nandi and coworkers demonstratedthat ovariectomy and/or hypophysectomy lead to the loss of alveolar structures in pregnantmice. In these same mice, the administration of E2 and progesterone together was required tomaintain the alveolar structures in their normal form [Nandi, 1958; 1959]. When these miceare ovariectomized, hypophysectomized and adrenalectomized, the requirements forlobuloalveolar maintenance are increased to include prolactin, as well as E2 and progesterone.9E2 stimulates the production of prolactin receptors in mouse mammary tissue [Sheth et al.,1978]. By increasing the number of prolactin receptors, E2 may act as a modulator ofmammary epithelial cells ability to respond to differentiation signals.The requirement of E2 for the maintenance of alveolar structures during pregnancy andlactation may not be common to all animals. In a different strain of mice than used by Nandiand coworkers, animals undergoing ovariectomy during lactation were still able to continuelactating [Griffith and Turner, 1962; Kuramitzu and Loeb, 1921], suggesting that E2 was notrequired to maintain the differentiated state. However, another study using a third strain ofmice has shown that lactating mammary gland tissue itself may produce sufficient E2 tomaintain Iactational activity [Sheth et al., 1978]. It is also possible that sufficient levels of E2 tomaintain lactation may be produced by endocrine organs other than the ovaries, such as theadrenal glands, or by peripheral tissues such as adipocytes. Therefore, it is difficult toconclude by ovariectomy alone that E2 is not required for lobuloalveolar maintenance in vivo.Conflicting findings such as these, all from the same type of animal model, differing only in thestrain of animal used, demonstrate the potential difficulties in extrapolating information fromanimal models directly to the human situation.1 d. In vitro Evidence for the Role of ELin Growth and Differentiation of the MammaryGland Determination of the hormonal requirements of mammary epithelial cell growth has beenfacilitated greatly by the in vitro techniques of organ culture and cell culture. The developmentof chemically defined media for mammary gland tissue aided our understanding of the controlof mammary gland development [Elias, 1957; Ichinose and Nandi, 1964]. These mediaallowed studies on the hormonal requirements of mammary gland development to "dearly"define the essential hormones involved. The early studies investigated alveolar development10and the hormonal signals involved in regulating both the growth and differentiation of thesestructures [Banerjee et al., 1973; Ichinose and Nandi, 1966; Wood et al., 1975]. These studiesutilized organ culture to study virgin mammary tissue removed from mice that had been"primed" with injections of both E2 and progesterone. This treatment alone is insufficient tostimulate alveolar differentiation in vivo. If the whole mammary glands are then removed andcultured as organ cultures, alveolar differentiation is induced by the addition of prolactin, insulinand a glucocorticoid (hydrocortisone). These results initially seem contradictory to the in vivostudies, in which both E2 and progesterone are required for lobuloalveolar development tooccur. However, it is quite likely that in organ cultures, residual hormones from the priming arecarried over with the gland to the culture medium. This carry over effect may also include anynumber of other hormones and growth factors which might also be involved in alveolardevelopment. The results of organ culture studies are further complicated by the finding thatorgan cultures of unprimed rats undergo DNA synthesis and lobuloalveolar differentiation inresponse to treatment with only insulin and prolactin. The addition of E2 further stimulated thisresponse but was not required [Dilley and Nandi, 1967].The discrepancies between the in vivo studies and the in vitro studies utilizing organculture are intriguing. Presently, there is no sufficient explanation as to why mouse and ratmammary tissue, both of which require E2 for ductal and alveolar mammary epithelial cellgrowth in vivo, differ so dramatically in vitro in organ culture with respect to the requirement ofin vivo pretreatment with E2 and progesterone. Interpretation of the organ culture studies isfurther complicated by the inability (at the time of the original research) to determine the targetcells and effector cells of the various hormonal treatments. The importance of this ishighlighted by the finding of Shymala and coworkers who demonstrated that the dramaticincrease in DNA synthesis in the mammary gland tissue of rodents injected with E2 occurs firstin the connective tissue (16 h) and subsequently in the epithelial tissue (24 h) [Shymala andFerenczy, 1984]. Although the advent of chemically defined media was hoped to eliminate the11complexities of hormonal interactions in vivo, many questions are left unanswered by the useof organ culture.The finding that enzymatic digestion of finely chopped mammary glands with the bacterialcollagenase enzyme produced by Clostridium histolyticum resulted in viable fragments of themammary epithelial tree was a major step in the development of primary cultures of MEC[Lasfargues, 1957]. Lasfargues showed that the epithelial fragments from collagenasedigestion were relatively free of fat and stromal cells, which allowed enriched populations ofepithelial cells to be examined in vitro. Several investigators have studied the effects of E 2and other mammogenic hormones on the growth of primary cultures of rodent MEC grown onplastic in serum-containing medium. E2 has either no or a very slight stimulatory effect underthese conditions [Ceriani and Blank, 1977; Hallowes et al., 1977; Richards and Nandi, 1978].The data from studies utilizing primary cultures of rodent MEC grown on biologicalsubstrates may provide more relevant information on growth effects of E2 and othermammogenic hormones, given that cells grown on these substrates are capable of undergoingextensive cytodifferentiation in response to hormones [Emerman and Pitelka, 1977; Emermanet al., 1977]. Studies examining the effects of E2 on the growth of both mouse and rat MEChave shown that E2 is unable to stimulate the growth of these cells when grown upon collagen-coated dishes [Edery et al., 1984; Imagawa et al., 1985]. In contrast to these studies, Ethierand coworkers have demonstrated an E2 stimulation of rat MEC growth on cells grown oncollagen-coated dishes in serum-free medium [Ethier, 1986; Ethier et al., 1987]. It is importantto note, however, that the serum-free medium used by Ethier's group contained epidermalgrowth factor (EGF), whereas the serum-free medium used in the studies showing no E2 effectdid not contain EGF. The importance of this is discussed in detail later.Interpretation of the large body of literature on E2 effects on the growth and/ordifferentiation of mammary epithelial cells in primary cell culture is made difficult by the12variation in experimental conditions used by different investigators. The results of the relativelyfew studies investigating the effects of E2 on the growth of human MEC (HMEC) have been atleast as mixed as those done with rodent cells. E2 in primary culture can stimulate theproliferation of normal HMEC grown on plastic in the presence of human serum, and it appearsthat in this same culture system progesterone stimulates differentiation [Mauvais-Jarvis et al.,1986]. Also in normal HMEC in primary culture, E2 has been shown to stimulate DNAsynthesis, possibly due to shortening of the cell cycle length [Calaf et al., 1986]. A commonfeature of these studies and others demonstrating E2 stimulation of growth is the presence ofserum in the medium [van Bogaert et al., 1982; Longman and Buehring, 1987]. Other effectsof E2 on cultures of mammary epithelial cells include increases in the number of microvilli[Chambon et al., 1984], increased casein production and lactose synthetase activity [Sankaranet al., 1984], and increased tyrosyl-kinase activity [Sheffield et al., 1987]. There are few resultsof the effects of E2 on the growth of HMEC in serum-free medium. Yang et al. [1980, 1982]have examined the effect of E2 on the growth of normal and fibroadenoma HMEC. E2 has nogrowth effects in their system in either the presence or absence of serum.One common criticism of organ culture and cell culture as tools for studying the control ofgrowth and differentiation of the mammary gland has been that investigators are simplystudying novel growth requirements for mammary tissues in abnormal conditions. Althoughthis may be partially true, information regarding the direct effects of hormones and interactionsamong them on growth and differentiation of mammary tissue is still greatly lacking. Thetechnique of cell culture, both of epithelial and stromal cells alone and in combination, providesa tool to more completely understand the actions and mechanisms of action of hormones suchas E2 in the regulation of mammary gland development.13In blood serum, E2 is present bound to steroid binding protein. In this form, E2 cannotreadily enter target cells. However, in its free unbound form, E2 readily migrates through lipidcell membranes due to its cholesterol based structure. Once inside the cell, it interacts with aspecific intracellular receptor, the estrogen receptor (ER). The nature of the ER and itsintracellular location is still a topic of some debate. The debate centers on whether the ER inits free form resides exclusively inside the nucleus or in both the cytoplasmic and nuclearcorn partm ents.If ER is found in the cytoplasm, then immediately upon binding to E2, the E2-ER complexis translocated to the nucleus. Clearly if ER is normally present in the nucleus thentranslocation is not a relevant step following hormone binding. In either case, the E2-ERcomplex has a high affinity for specific regions of DNA. These regions have been termedhormone responsive elements (HRE)[Darbre, 1990]. Upon binding to its specific HRE, the E2-ER complex functions as a transcriptional regulator. The end result is transcriptional activationof genes involved in the control of target cell growth and differentiation. Specific examples ofgenes directly effected include c-myc, c-fos, EGF, TGF-a, and many others [van der Burg etal., 1989, Lippman and Dickson, 1989]. The particular importance of some of these specifictranscriptional activations is discussed further in a later section.The finding that E2 is generally unable to stimulate the growth of mammary epithelial cellsin serum-free primary culture has led to considerable speculation of how E2 actually functionsin vivo and in organ culture to stimulate the growth of ductal epithelial cells. Three majorhypotheses have been proposed to account for the effects of E2:1. E2 directly stimulates the proliferation of target cells in which it is bound [Aitken andLippman, 1982].142. E2 indirectly stimulates growth by the production of systemically acting factorsproduced at sites distal to the mammary gland [Sirbasku et al., 1985].3. E2 acts both by affecting target cells responsiveness to local autocrine and paracrinegrowth factors, and by affecting the production of these locally acting growth factors[Lippman et al., 1988].Both the first and third hypotheses of stimulation of HMEC growth are supported by thefinding that E2 implants into the mammary glands of 5 week old ovariectomized micestimulates the growth of only the mammary gland into which the E2 implant is placed. In thiscase E2 is not exerting its growth regulatory effects via the production of systemic factors[Haslam, 1988b]. However, if these same experiments are performed on 10 week old mice,the effect of the implant is stimulation of epithelial cell growth in both the implant-receivingmammary gland and the contralateral control mammary gland. In this case E2 is exertinggrowth regulatory effects via the induction of a systemic factor. The lack of direct E2stimulation of HMEC growth in vitro is therefore likely due to the absence of required factorspresent in vivo. These investigators further demonstrated that the ability of E2 to increasefunctional levels of progesterone receptors is specific to the tissue from ten week old mice, anddoes not occur in tissue from 5 week old mice [Haslam, 1988b]. These findings suggest that inthis model system E2 can stimulate growth both directly and systemically, depending on theage of the animal. E2 only stimulates the production of progesterone receptors at an agewhen systemic effects are also occurring.In an elegant series of experiments, Daniel et al. showed that the implantation of E2-releasing pellets causes increased growth in a specific subset of the end-bud region MEC[Williams and Daniel, 1983] called the cap cells [Daniel et al., 1987]. However, the hormone-binding studies using radiolabelled hormones and autoradiography demonstrated that theactual hormone-binding cells are the stromal cells and another epithelial cell subset of the end15bud region, called the end bud cells, but not the cap cells. The work of Daniel's group alsosupports the model of E2 action whereby E2 exerts its growth regulatory effects via theproduction of paracrine acting growth factors.The ability of E2 to stimulate progesterone receptors in MEC supports the model for E2action in which E2 modulates the responsiveness of the MEC to other hormones and growthfactors. E2 has also been shown to regulate the level of EGF receptors (EGFR) in uterinetissue [Gardner et al., 1989]. The regulation of EGFR by E2 is discussed further in Sec. 1f.lf. Role of Epidermal Growth Factor (EGF) and Transforming Growth Factor-a (TGF-a) in Control of Mammary Gland Growth and Differentiation EGF is a small single chain polypetide consisting of 53 amino acids [Carpenter andCohen, 1979]. Both EGF and the closely related TGF-a exert their biological effects viainteraction with the same transmembrane receptor, the EGFR [Todaro et al., 1980, Massague,1983, Carpenter, 1987]. Interaction with the receptor activates the receptor kinase portion ofthe receptor and an intracellular increase in tyrosine phosphorylation is observed followingEGF binding. Other cellular responses to EGF include the EGFR activation of phospholipase-C and the subsequent increased production of inositol 1,4,5-triphosphate (IP3) [Schlessinger,1988]. The immediate substrates of the EGFR tyrosine kinase have not yet been conclusivelyidentified. However, both the progesterone receptor and HER2-neu receptor are rapidlytyrosine phosphorylated in response to EGF treatment [Ghosh-Dastidar et al., 1984; King etal., 1988]. The HER2-neu receptor is another transmembranous growth factor receptor with atyrosine kinase intracellular domain. HER2-neu is highly homologous to EGFR in its aminoacid and DNA sequence, which in the case of HER2-neu is encoded on the c-erbB-2 cellularoncogene [Coussens et al., 1985]. This gene and its encoded receptor will be discussedfurther with reference to its role in breast cancer in Sec. 2.16It is interesting to note that the EGFR gene belongs to a relatively large family of genesencoding growth factor receptors, many members of which have already been characterizedfor their potential to undergo oncogenic mutation or activation [Hunter and Cooper, 1985;Yarden and Schlessinger, 1988]. Although no directly oncogenic mutations in the EGFR genehave yet been discovered in breast cancer, amplification of the closely related c-erbB-2 geneand overexpression of its protein product are associated with advanced forms of breast cancer[Slamon et al., 1987]. The importance of EGFR as it applies to breast cancer is discussedlater.Both direct and indirect evidence links EGF to regulation of normal mammary epithelialcell proliferation. High levels of EGF are found in both milk and breast cyst fluid [Zwiebel et al.,1986; Connolly and Rose, 1988]. TGF-a activity has been isolated from normal mammarytissue [Valverius et al., 1987]. Unfortunately it is presently unknown whether normal mammarygland epithelial cells directly produce EGF or TGF-a, or if they sequester the high levels fromthe surrounding extracellular fluids. The demonstration of m RNA for both EGF and TGF-a innormal mammary tissue from both rodent and human sources suggests that the epithelial cellsare likely to produce at least a portion of the EGF and TGF-a isolated from mammary tissue[Brown et al., 1989; Liscia et al., 1990].The presence of specific high and low affinity receptors on normal MEC further supportsa potential role for EGF in the control of growth and/or differentiation [Taketani and Oka, 1983].EGFR levels vary according to the physiological state of the of the mammary gland. There aresignificant numbers of EGFR present on the MEC in virgin and lactating glands, which increaseto peak levels during pregnancy [Edery et al., 1985]. It is therefore possible that the hormonalfactors modulating growth and differentiation during pregnancy also modulate EGFR levels. Ifthese changes in EGFR subsequently increase the cells responsiveness to EGF, then it isconceivable that these effects may be involved in the stimulation of growth and differentiation17observed during pregnancy. This concept is not entirely speculative given that increasedproduction of both EGF and EGFR accompanies estrogenic stimulation of epithelial cell growthin the mouse uterus [Gardner et al., 1989; Huet-Hudson et al., 1990].Levels of circulating EGF produced by the submandibular gland in mice are regulated byprogesterone [Bullock et al., 1975], and progesterone stimulation of growth in breast cancercell lines is accompanied by increased expression of EGFR [Musgrove et al., 1991]. Asmentioned earlier, progesterone is critical to the development of lobuloalveolar structuresduring pregnancy [Haslam, 1987]. It is therefore possible that progesterone activity is alsopartially mediated by stimulation of EGF production. Given that E2 is directly able to regulateprogesterone receptor levels (Sec. le.), the regulation of progesterone levels in thesubmandibular gland may function as an important control pathway for the regulation of MECgrowth by distal organs, as proposed in the estromedin hypothesis for E2 action [Sirbasku,1978].Direct support for the role of EGF in mammary gland development in vivo comes fromstudies where the level of EGF is experimentally modulated. Pregestational removal of thesubmandibular gland, the principal source of EGF in mice, results in a dramatic decrease inboth mammary gland size and the volume of milk produced by subsequently impregnated mice[Okamoto and Oka, 1984; Sheffield and Welsch, 1987]. When replacement of EGF is providedby injection during pregnancy, both mammary gland size and milk volume return to normallevels [Oka et al., 1988]. However, it was noted that the level of lobuloalveolar developmentdoes not return to normal by supplementation with EGF alone. This leads to the hypothesisthat other growth regulatory factors are also being produced by the submandibular gland,which might be tentatively labeled as potential estromedins.Other investigators have used polymer implants similar to those in studies examining E2function to analyze the effect of EGF on mammary gland growth and differentiation in vivo.18Coleman et al. [1988] showed that implantation of EGF pellets in the mammary glands ofovariectomized mice leads to increased end bud formation. Coleman and co-workers alsodemonstrated that EGF binding occurred in end-bud cells, ductal myoepithelial cells and in thestromal cells surrounding the responsive epithelial cells. More recently the same groupshowed that implantation of EGF into intact animals leads to a time-dependent inhibition ofgrowth in the ductal epithelium [Coleman, 1990]. This effect is observed after an exposure tothe pellets of at least 3 days. A down regulation of EGFR levels in response to the implants isalso observed. The authors suggest that the growth inhibitory effect is likely due to this downregulation of EGFR by the implanted EGF.In vivo studies with implants have also suggested a possible synergistic effect betweenEGF and the ovarian steroids E2 and progesterone [Vonderhaar, 1987]. Vonderhaarexamined the effects of implants of both EGF and TGF-a on ductal branching and end-budformation and growth in ovariectomized mice. EGF stimulates both ductal branching and end-bud formation if both E2 and progesterone are added in conjunction with the EGF implant. Incontrast to EGF, TGF-a stimulates both branching and growth independently of additionalhormonal supplementation [Vonderhaar, 1987]. This study suggests that not only may E2and/or progesterone be required for the EGF effect, but that the closely related growth factorsEGF and TGF-a may have very different biological responses despite interaction with acommon receptor pathway.Tissue culture experiments have demonstrated the ability of EGF to stimulate the growthof breast cancer cell lines [Lippman and Dickson, 1989], and normal MEC from numerousanimal models [Yang et al., 1980 & 1986; Salomon et al., 1981; McGrath et al., 1985; Ethier etal., 1990]. In mouse organ culture, Tonelli and coworkers [1980] were able to stimulate a cycleof growth, differentiation and regression by addition of the insulin, prolactin, aldosterone andhydrocortisone to the medium in cultures of mammary glands from virgin mice that have been19primed with E2 and progesterone injections. However, when EGF was added to the medium incombination with insulin, prolactin, aldosterone and hydrocortisone, the investigators were ableto stimulate a second round of development, differentiation and regression, which could not bedone without the presence of EGF. The effects of EGF added to primary cell cultures ofnormal and benign MEC are varied in terms of the growth response. EGF addition to serum-free media has been shown to stimulate growth in cultures of normal and malignant MEC fromintact postpubertal nonpregnant mice [Imagawa et al., 1982], ovariectomized andovariectomized + adrenalectomized mice [Levay-Young et al., 1990] and intact maturenonpregnant rats [McGrath et al., 1985; Ethier, 1985]. Studies by Ethier's group demonstrateda requirement for EGF in serum-free medium for either progesterone, prolactin or E2 to exertgrowth modulating effects on cultures of normal rat mammary epithelial cells in monolayercultures. The growth effects of progesterone and prolactin were reported to vary in individualexperiments, although E2 was reported to cause a slight but consistent stimulation of growth ofrat mammary epithelial cells in medium containing EGF [Ethier, 1986; Ethier et al., 1987].The ability of EGF to stimulate the growth of rat mammary epithelial cells in culture is alsodependent on the nature of the substrate upon which the cells are grown. Although EGFstimulates the growth of cells in serum-free medium on either a plastic or Type I collagensubstrate, it has no effect on the growth of the same cells on Type IV collagen [Salomon et al.,1981; Kidwell and Shaffer, 1984]. The investigators hypothesize that EGF stimulation ofgrowth on plastic or Type I collagen is due to the stimulation of Type IV collagen production bythe cells in response to EGF treatment. The Type IV collagen (basement membrane collagen)likely provides a better substrate for epithelial cell growth than either plastic or Type I collagenand stimulates growth on its own. More recent studies have shown that TGF-a also stimulatesthe production of Type IV collagen [Liu et al., 1987]. Therefore, the above model explainingEGF stimulation of growth may also be applied to the growth-stimulatory effects of TGF-a onmammary epithelial cells in culture. The importance of substrate and the effects of hormones20and growth factors on both the production and degradation of substrate components isdiscussed in further detail in the Discussion section of this thesis.In contrast to the numerous experiments showing stimulation of MEC growth in responseto EGF, Ehmann and coworkers reported a growth inhibitory effect of EGF on mouse MECgrowth [Ehmann et al., 1984] in serum-containing media. The epithelial cells were grown on afeeder layer of irradiated rat mammary tumor cells. The possible production of unidentifiedgrowth factors by the feeder layer or the presence of other growth factors in the serum werenot accounted for and make the growth inhibition by EGF difficult to interpret.Only one other report has shown an absence of growth stimulation by EGF in primarycultures of HMEC. Yang et al. [1986] utilized the same serum-free medium as was used forthe experiments described in this thesis and showed that EGF stimulated cells fromfibroadenomas only if the cells were grown in three-dimensional cultures. In the two-dimensional culture system used in this thesis, Yang and coworkers found no growth effectwith EGF. Potential explanations for the discrepancy in results from Yang's group and thosepresented here are presented in the Discussion section.I •^- 1 -^ •I•e 1I^• Al 1Studies of hormones and growth factors in the genesis and progression of breast cancerhave examined the involvement of all the hormones and growth factors already discussed withregard to growth control of the normal mammary gland. The literature reviewed here is limitedto the roles of E2 and EGF in the process of breast cancer.2a. Estrogens and Breast CancerAnimal studies investigating the role of estrogens in breast cancer have shown thatprolonged exposure to estrogens can lead to the induction of mammary tumors [Dunning et al.,211947; Cutts and Noble, 1964]. Cutts and Noble showed that the incidence of estrogen-induced tumors varies dramatically between different species of rats. The mammary tumors inthese animals undergo partial remission in response to ovariectomy or adrenalectomy and totalremission in response to hypophysectomy. Therefore, estrogen's role in the control of growthin these tumors is neither complete nor direct, as removal of endogenous estrogens onlycaused partial regression of tumors, whereas hypophysectomy resulted in total regression ofthe tumors. These observation have led to debate as to how estrogens might function in thegenesis and growth control of breast cancer.Further animal studies have suggested E2 likely acts as a permissive or promoting agentrather than a carcinogen. When 3-methylcholanthrene (3-MC), a polycyclic aromatichydrocarbon that is highly carcinogenic, is administered to female Sprague-Dawley rats in asingle feeding results in an incidence of mammary tumors up to 100%, depending on the ageof the rats at the time of 3-MC exposure [Huggins, 1961]. In male rats, no mammary tumorsare caused by the same treatment, suggesting that female sex hormones play a key role aspermissive agents in mammary tumor induction. Ovariectomy prior to 3-MC administration alsoeliminated the induction of mammary tumors by 3-MC. If ovarian grafts are performed on thesame day as carcinogen exposure, there is a partial restoration in the incidence of mammarytumors [Dao, 1962]. However, if ovarian grafting is performed subsequent to carcinogenexposure, the incidence of mammary tumors is unaffected, demonstrating that ovarianhormones are required as permissive agents for the carcinogenic effect.Animal studies examining the role of ovarian hormones in growth control of establishedmammary tumors have also utilized carcinogen-induced tumors to a great extent. Thesestudies have shown that ovariectomy leads to temporary regression of mammary tumors[Huggins et al., 1961; Dao, 1962; Gullino et al., 1975]. Although these studies suggest a rolefor E2 in the initiation and growth control of breast cancer, they are far from conclusive.22Ovariectomy does reduce circulating E2 levels, however, it also reduces progesterone levelsand possibly levels of other unidentified factors as well. Furthermore, the regression of tumorsis only temporary in response to ovariectomy, further demonstrating the importance of otherfactors in the growth control of mammary tumors. It is possible that mammary tumors areinitially dependent on ovarian hormones for growth control and the progression to a hormone-independent state accounts for the temporary regression resulting from ovariectomy. Thiscritical hypothesis is discussed further in consideration of growth factors in breast cancer.Treatment of numerous ER+ breast cancer cell lines with E2, both in vitro and in nudemice, leads to an increase in growth in the tumor cells [Lippman et al., 1976; Soule andMcGrath, 1980; Darbre et al., 1983; Lippman and Dickson, 1989]. Accompanying theincreases in growth are increases in the production of numerous peptide growth factors andtheir receptors by the cancer cells [Dickson et al., 1986]. The list of such growth factorsincludes, but is not limited to, EGF and the closely-related TGF-a, insulin-like growth factor-1(IGF-1), platelet-derived growth factor (PDGF), and basic fibroblast growth factor (bFGF).Growth inhibition of ER+ cell lines by the antiestrogen tamoxifen is accompanied by adecrease in the level of production of these growth factors, as well as an increase in theproduction of growth inhibitory substances in some cell lines. In ER - breast cancer cell lines,which grow at a higher basal rate and are not stimulated by the addition of E2, these samegrowth factors are produced autonomously at elevated levels [Davidson and Lippman, 1989].These observations in cell lines also support a model of mammary tumor progression bythe deregulation of the production of autocrine and paracrine growth factors. According to thismodel, an initial event leads to abnormal production of growth factors in response to E2. Atthis stage the tumor is still hormone-dependent, requiring E2 for continued growth. The animalstudies and epidemiological data showing temporary remission of mammary tumors inresponse to ovariectomy provide strong supportive evidence for this portion of the model. A23second defect in the growth control of these hormone-dependent tumors then results inautonomous production of the same growth factors that were initially under E2 control. Thisstage would be the hormone-independent tumors characterized by higher basal growth ratesand independence of E2 for growth. The next section (Sec. 2b.) will consider further evidencesupporting an important role of EGF and its receptor in this mechanism of disease progression.Considerable epidemiological evidence exists linking female sex steroids to a role in theonset and progression of malignant breast disease. Women with early menarche and latemenopause appear to have a greater risk of developing breast cancer [Zumoff et al., 1975;Kelsey, 1970], whereas an early onset of menopause or ovariectomy correlates with a reducedrisk of breast cancer [Trichopoulos et al., 1972]. This epidemiological evidence is the basis ofthe estrogen window hypothesis which states that the longer a woman's exposure to E2, thegreater the risk of breast cancer. The total duration of E2 exposure is defined as the time frompuberty to menopause. It has been suggested that it is the exposure to unopposed E2, that islow progesterone levels, that might increase the risk of breast cancer [Korenman, 1980]. Lowlevels of progesterone are characteristic of the follicular phase of the menstrual cycle, or theymay occur due to deficiencies in normal production of progesterone in the luteal phase.However, this theory is largely speculative and investigators have been unable to find anyassociation between anovulatory cycles (low progesterone) and women with a high incidenceof breast cancer [MacMahon et al., 1980]. Proponents of this theory of unopposed estrogensuggest that the reduction in breast cancer incidence in women with a first pregnancy at ayoung age is due to a protective function of the high doses of progesterone produced duringpregnancy [Davidson and Lippman, 1989].The incidence of breast cancer is significantly increased with obesity and may be due tothe fact that fat cells are the principal site for the conversion of androstendione to estrone, theprecursor to the family of biologically active estrogens [Davidson and Lippman, 1989].24However, it has not been shown that circulating levels of E2 are different in obese women thanaverage body weight controls. Also, the levels of E2 in breast tissue are significantly higherthan in serum [James et al., 1971; Witliff, 1974]. The higher levels of adipocytes (whichproduce E2 from inactive precursors) present in the breasts of obese women could lead toincreased local levels of E2 in the breast.Due to the possible role of E2 in the initiation of breast cancer, numerous studies havebeen conducted examining the risk of breast cancer in women using hormonal contraceptivescontaining E2 and in women utilizing E2 replacement therapy post-menopausally [Drill, 1981;Thomas, 1982]. The findings in this large body of studies have been varied with respect toincreased risk of breast cancer and further studies are required to determine the roles ofduration of exposure and dosage of the exogenous E2. To date there is no convincingevidence that E2 used medically has any role in the initiation of breast cancer, however, thepotential for such a risk warrants further study in this area. In summary then, the experimentaland epidemiological evidence suggests a role for estrogens in the breast cancer process.The strongest evidence for the importance of E2 as the specific female sex steroidinvolved in growth control of mammary tumors comes from epidemiological studies on the ERstatus of breast cancer patients. Approximately 66% of human breast cancers are classifiedas ER+ [Clarke et al., 1990]. In this ER+ group of patients, remission is observed in 70% ofcases treated with either E2 removal via ovariectomy, high dose E2 therapy or treatment withthe antiestrogen tamoxifen. Due to the severity of side effects with E2 removal or high dosetreatment, the current treatment is the use of the relatively low side-effect inducingantiestrogen tamoxifen. These same endocrine treatments yield remission in only 5-10% ofER- breast cancers. This difference in response represents a progression of the tumors fromhormonally-responsive ER+ cancers to hormonally-independent ER - tumors [King, 1989].Progression of the tumors to a hormonally-independent state could occur by a number of25defects in the cell. Unregulated and elevated expression of autocrine and paracrine growthfactors which are normally under endocrine control could be a mechanism by which breastcancer cells progress to hormonal independence. Another possibility is that breast cancercells have an exaggerated response to the autocrine and paracrine growth factors that areproduced in response to endocrine hormones. Exaggerated responses to hormonally-regulated growth factors could occur with amplified receptor levels.2b. EGF and TGF-a in Breast CancerThe roles of EGF and TGF-a in the growth control of normal mammary tissue has alreadybeen discussed and includes a description of both the growth factors and their receptor,EGFR. Experimental animal evidence links EGF to the development of malignant breastdisease in mice with a high incidence of breast cancer. If the mice undergo sialoadenectomy(removal of the submandibular gland) prior to 30 weeks of age, the incidence of breast cancerin these mice drops from 63% to 13% at 52 weeks of age [Kurachi et al., 1985]. Theimportance of the 30 week age limit is that at this time the production of EGF by thesubmandibular glands in mice is greatly increased and the incidence of mammary tumors risesdramatically following this event. Removal of the submandibular glands after the appearanceof mammary tumors in these mice results in a regression of the tumors. If the mice receiveEGF injections after sialoadenectomy, the incidence of new and growth of established tumorsis returned to elevated levels. These findings demonstrate both promotional and growth-regulatory roles for EGF in mammary carcinogenesis.Considerable evidence from breast cancer cell lines links both EGF and TGF-a toregulatory roles in breast cancer cell growth. EGF and TGF-a are both produced by breastcancer cell lines [Salomon et al., 1984; Dickson et al., 1985 and 1986] and numerous studiesshown that both EGF and TGF-a stimulate the growth of the same cell lines [Lippman and26Dickson, 1989]. Furthermore, E2 stimulation of ER+ breast cancer cell line growth isaccompanied by increased production of both EGF and TGF-a [Murphy et al., 1988, Lippmanet al., 1988] and increases in EGFR levels [Berthois et al., 1989; Bates et al., 1990]. Additionof antibodies to EGFR results in a partial block of E2 stimulation of growth [Bates et al., 1988].This finding suggests that although E2 stimulation of growth is partially due to production offactors interacting with EGFR, other factors are also relevant. A possible role for insulin-likegrowth factor-1 in this function is presented in the Discussion.Epidemiological evidence from breast cancer patients also supports a role for EGF, TGF-a, and EGFR in breast cancer. Primary tumor samples have been shown to contain significantlevels of TGF-a m RNA and protein [Nickell et al., 1983; Salomon et al., 1984]. Tumors show alarge degree of variability in the level of EGFR expressed, with higher EGFR levels in breastcancer biopsies corresponding to significantly poorer prognosis [Spyratos et al., 1990;Nicholson et al., 1991]. This finding is intriguing as studies of normal mammary tissue haveshown that the highest levels of EGFR observed in breast cancer biopsies are also observedin normal tissue [Pekonen et al., 1988; Barker et al., 1989]. The best explanation for thecorrelation between high EGFR levels in breast cancer and poor prognosis is that it is not theabsolute level of EGFR that is important, but the interrelationship between EGFR levels andother prognostic factor such as ER status. The highest levels of EGFR are observed in ER'breast cancers, the hormonally-independent form of the disease, which has the poorestprognosis. This supports the importance of the interrelationship between these two pathways[Toi et al., 1989]. It is possible that the elevation of EGFR levels observed in ER- mammarytumors represents a key step in their progression to hormone independence. Theseepidemiological findings are yet another line of evidence supporting the hypothesis thatderegulation in growth factor production or responsiveness is likely involved in the progressionof breast cancer from a hormonally responsive disease to a hormonally independent disease.273.Thesis ObjectivesAlthough the literature contains much information on the hormonal and growth factorregulation of mammary gland growth, there is still much that is not understood regarding theeffects of E2, EGF, and TGF-a on the growth of HMEC. The principal objective of this thesiswas to investigate the effects of each of these factors, alone and in combination, on HMEC inserum-free primary culture. Although E2 has been shown to stimulate the growth of MECgrowth in vivo, it is presently unclear whether E2 can directly stimulate the proliferation ofHMEC in serum-free primary culture. This thesis investigates the growth effects of E2 atconcentrations ranging from 1-1000 nM on the growth of HMEC from reductionmammoplasties, fibroadenomas, and ER+ carcinomas. The objectives of this portion of thethesis were to determine if E2 could directly effect the growth of HMEC from either of the threemammary tissue types in serum-free primary culture and to describe any differences in E2responses that exist among the different tissue types. Another objective was to determine iffactors present in the serum-free medium are involved in modulation of any observed E2effects on HMEC growth.Both EGF and TGF-a have important growth regulatory roles in both normal andmalignant mammary tissue. This thesis investigates the effects of both EGF and TGF-a onHMEC from reduction mammoplasties, fibroadenomas, and ER+ carcinomas. The objective ofthis portion of the thesis was to determine if these growth factors were able to directly stimulatethe growth of HMEC in serum-free primary culture and if there were any significant differencesin the responses of HMEC from the three different mammary tissue types. Another objective ofthis portion of the thesis was to compare the effects of EGF and TGF-a in order to determine ifthere were any differences in HMEC growth responses to the two factors.Evidence indicates that E2 stimulated production of paracrine growth factors and E2modulation of HMEC responsiveness to growth factors is likely involved in the E2 stimulation of28HMEC growth. This thesis investigates the effects of E2 plus EGF on the growth of HMEC inserum-free primary culture. The objective of this portion of the thesis was to determine if theaddition of exogenous EGF would be capable of increasing E2 stimulation of growth in HMECin serum-free primary culture.29CHAPTER 2 MATERIALS AND METHODS1. Sample Procurement and AssessmentNormal tissue was obtained from reduction mammoplasties; fibroadenoma andcarcinoma samples were obtained from biopsies and mastectomies. All surgical procedureswere performed at local hospitals by collaborating surgeons. All carcinoma samples chosenfor this study were ER+ in situ or infiltrating ductal carcinomas. ER levels and pathologyreports were provided by the pathology departments of the hospitals.2, Tissue Preparation and FreezingAn insulated container equipped with sterile cups each containing transport medium(Appendix 1) on ice was delivered to the operating room on the morning of the surgeries.Samples were aseptically placed in the transport medium by operating room nurses andbrought back to the tissue culture room as soon as possible after surgery. Under sterileconditions, excess fat was trimmed from the tissues with scalpels. The remaining tissue wasminced into approximately 1 mm 3 pieces using 2 scalpels in a cross-cutting manner. Withlarger reduction mammoplasty samples it was necessary to change scalpel blades frequentlydue to dulling of the blades, which lead to difficulty in sufficiently mincing the tissue. Usingforceps, the minced tissues from small reduction mammoplasties, biopsies and mastectomieswere transferred to 1.7 ml cryotubes until they were 1/2 full. The tube was filled with freezingmedium to a volume of 1 ml (Appendix 2). The freezing tubes containing minced tissue andfreezing medium were inverted gently to insure mixing of freezing medium and the tissuepieces. The tissue was then slowly frozen and stored in liquid nitrogen until dissociated for cell30culture. Large reduction mammoplasties were dissociated as described below prior tofreezing.3. Dissociation ProcedureFrozen tissue was removed from storage in liquid nitrogen and rapidly thawed by firstwarming with rotation of the vial between the hands, followed by immersion in a 37°C waterbath. The vial was wiped with 70 % ethanol prior to opening. A Kim-wipe wetted with 70 %ethanol was held over the lid to avoid aerosol release upon opening of the tube. The tissuewas aseptically transferred from the freezing vial to a 15 ml conical centrifuge tube and 5 ml ofDME/F12 (1:1) (Terry Fox Media Preparation Services) + 10 mM Hepes (Sigma) pre-warmedto 37°C was added to the tube. The mixture was spun in a clinical centrifuge for 3 min at 1000rpm (100 x g). The supernatant was discarded and 5 ml dissociation medium (Appendix 3)was added to the tube. The dissociation procedure has been described previously [Emermanet al., 1990]. The mixture of tissue and dissociation medium was transferred into 250 mldissociation flasks. The centrifuge tube was rinsed with a further 5 ml of dissociation mediumthat was also transferred to the dissociation flask. An additional 40 ml of dissociation mediumwas added to the flask bringing the total volume of dissociation medium to 50 ml. In thesmaller samples, such as minced tissue from biopsies, a 125 ml dissociation flask with 25 ml ofdissociation medium was used. The dissociation flask was covered in sterile tin foil andparafilm and placed in a gyrating shaker inside a 37°C incubator and shaken for approximately18 h. Starting at 15 h the dissociation mixture was examined every 1 h and dissociation wasconsidered complete when no large pieces of tissue remained. Typically, the solution wascloudy with stringy appearing aggregates of dissociated cells in suspension. The dissociationsolution was divided equally into four 15 ml or 7 ml conical centrifuge tubes, depending on theamount of dissociation medium and dissociated tissue. The cell suspension was centrifuged ina clinical centrifuge for 4 min at 800 rpm (80 x g). This centrifugation speed was chosen to31pellet preferentially the epithelial cells present in the suspension. The supernatant was usuallydiscarded and the cell pellets were combined and resuspended in 10 ml of DME. Forexperiments utilizing normal fibroblast feeder layers, the supernatant from dissociations ofreduction mammoplasty tissues was collected and centrifuged for 4 min at 1000 rpm (100 x g).The supernatant of this centrifugation was discarded and the pellet was resuspended in 10 mlDME, then treated the same as the epithelial cell pellet, described as follows. The cells wereresuspended and centrifuged for 4 min at 1000 rpm (100 x g). The supernatant was removedand discarded, and the pellet was again resuspended in 10 ml of DME/F12/Hepes. Thesolution was again centrifuged for 4 min at 1000 rpm (100 x g) and the pellet resuspended in 5ml of DME/F12/Hepes. The purpose of the repeated washings was to remove any remainingenzymes from the dissociation medium. To determine cell numbers, 0.1 ml of the cellsuspension was removed and placed in a clean 2 ml glass tube. A small drop of trypan blue(pH 7.2) was added to the solution in order to distinguish between viable and dead cells. Theplasma membranes of dead cells are not able to prevent trypan blue from entering thecytoplasm, and therefore, the dead cells stain blue. Viable cells were counted on ahemocytometer. For large reduction mammoplasty samples, the total cell yield was calculatedand the dissociated cells were mixed with freezing medium at a concentration of 1 x le cells /ml , then slowly frozen and stored in liquid nitrogen as described earlier. If the dissociatedcells were not to be frozen, they were cultured as described below.4. Cell CultureFollowing the determination of viable cell number, the cells were centrifuged for 3 min at1000 rpm (100 x g). The supernatant was discarded and the cell pellet resuspended inattachment medium (Appendix 4). The cells were seeded on dehydrated collagen-coated 24well tissue culture dishes at 3 x 105 cells / cm 2. In the case of fibroblasts grown to be feederlayers, the cells were directly seeded onto plastic. Collagen coated wells were made by32adding of one drop of rat tail collagen (Appendix 5) to each well [Emerman and Wilkinson,1990]. The dish was swirled gently so that the collagen covered the entire lower surface of thetissue culture well and any excess collagen was removed. The freshly coated dishes wereallowed to dry inside the laminar flow hood under UV. light to insure sterility. All cultures wereincubated at 37°C in 95 % air: 5 % CO2. For the first 24 h of culture, the medium consisted ofthe attachment medium containing 5 % pooled human serum from normal donors (Appendix6). This medium allowed the cells to attach to the substrate [Emerman and Wilkinson, 1990].After 24 h in culture, the medium for cultures of epithelial cells was changed to a phenol-red-free, serum-free medium (Appendix 7) with no extra additives; cultures of fibroblasts were leftin the serum-containing medium until they were confluent. It has been observed thatfibroblasts in serum-containing media grow rapidly [Emerman and Wilkinson, 1990]. Allepithelial cultures remained in the serum-free medium for an additional 24 h to insure maximalremoval of serum from the medium. After 24 h, the medium was changed and serum-freemedium containing varying amounts of E2 (1 to 1000 nM) and/or EGF (1 - 100 ng/ml) or TGF-a(1 - 100 ng/ml) were added to the cultures. All growth factors and media supplements (exceptfetal calf serum, Gibco) were obtained from Sigma and prepared according to manufacturersinstructions. It has been observed that the phenol-red used as a pH indicator in tissue culturemedium possesses estrogenic activity that might mask the growth effects of exogenouslyadded E2 [Benthois et al., 1986]. In order to avoid this, all experiments were carried out inphenol-red free medium. The insulin concentration in some experiments was varied from 0.1to 10.0 µg/ml to determine if high concentrations of insulin were masking an estrogenic effecton growth, as has been reported [van der Burg et al., 1988, Ruedl et al., 1990]. In somestudies, a monoclonal antibody to the EGF receptor, MAb 528 (Oncogene Science) , wasadded to the medium at 1.5 µg/m1 to block EGF binding. This MAb binds to the EGF receptor,blocking EGF binding as well as blocking EGF stimulation of the receptor tyrosine-kinaseactivity [Arteaga et al., 1988]335. Growth StudiesMedia were changed every 3 d and cultures were observed daily using a phase contrastmicroscope. When the fastest growing cultures were between 70 - 80 % confluent, a finalmedia change was done. The growth assay used has been described previously [Furlanettoand DiCarlo, 1984]. Fourteen h after the last media change, tritiated thymidine (3H-TdR) at 1pei/m1 was added to each well. After 6 h the media were removed and the cultures were fixedin 10 % trichloroacetic acid (TCA) at 4°C for 15 min. The 10 % TCA was removed and eachwell was washed 3 x with 5 % TCA at 4°C for 5 min each. The acid-insoluble material wasdissolved in 2N NaOH at room temperature for 24 h. Aliquots (250 ILI) were removed fromeach well and transferred into plastic scintillation vials. To each vial 2.75 ml of organicscintillation cocktail was added. Glacial acetic acid (25 RI) was also added to each scintillationvial to neutralize the NaOH solution. The addition of acid is necessary to allow dissolving ofthe aqueous solvent into the scintillation cocktail. The amount of 3H-TdR incorporation intoDNA was measured on a Beckman p-counter. Incorporation was measured in disintegrationsper minute (dpm). These values were subsequently converted to percents of controls to allowfor comparison of results among experiments. In all cases, the controls were cells grown in theabsence of E2, EGF, and TGF-a. The mean values were compared for significant differenceswith a two-tailed students T-test for differences between means.34CHAPTER 3 RESULTS7 • • 1 it^•^11 -_ 1 Li - 1 11 - • -(HMEC) in Primary CultureGiven the evidence for an important role of E2 in the growth control of MEC in vivo, andin and in ER+ human breast cancer cell lines in vitro, the effects of different concentrations ofE2 on the growth of primary cultures of cells from reduction mammoplasties, fibroadenomas,and carcinomas were examined. In all experiments, cells were seeded at 3 x 10 5 cells/cm2 inserum containing attachment medium (Appendix 4). After 24 h, the medium was changed toserum-free medium (Appendix 7) plus different concentrations of E2. When the fastestgrowing cultures were 70 - 80 % confluent, all cultures were labeled with 1 RCi/mI 3H-TdR for 4h prior to termination. In all experiments, growth was assessed by the level of incorporation of3H-TdR relative to the control condition which was the serum-free medium with no added E2.Two separate experiments utilizing cultures of HMEC from fibroadenomas were carried out todetermine the optimal time after the last medium change to assay growth using the 3H-TdRincorporation assay. The results of these experiments are shown in Figures la and lb. Inboth experiments, it was observed that reasonable incorporation of 3H-TdR occurred between12 and 16 hours after a medium change. For this reason, cell growth assays were done 14 hafter the last medium change. There was no E2 stimulation of growth in cells from eithersample at any of the time points examined. The growth of cells from FA 49 was notsignificantly (p < 0.05) effected by either concentration of E2 at any of the timepointsexamined. Growth of cells from FA 50 was not significantly effected at the 8 h timepoint;however, the inhibition of growth in FA 50 by 10 nM E2 at 12 and 16 h, and by 1000 nM E2 at16 h were all statistically significant (p < 0.05).35Figure 1 : The effects of time after the last medium change on the incorporation of 2.W.TdR into primary cultures of HMEC from 2 fibroadenomas (Fia. 1a. FA 49. Fig. lb. FA 50). Cells were cultured in serum-free medium containing different concentrations of E2. Thevalues shown for 3H-TdR incorporation are the means ± SEM of triplicate wells in the samecondition. The level of 3H-TdR incorporation is dependent on the time at which the 3H-TdR isadded after the last medium change. High level of incorporation were observed at 12 and 16h. At none of the concentrations examined was there any effect of E2 on the incorporation of3H-TdR into FA 49 (Fig. la.). E2 at both 10 and 1000 nM dramatically reduced incorporationof 3H-TdR into the cells at all three timepoints, however, the inhibition of incorporation wasonly statistically significant (p < 0.05) at the 12 and 16 h timepoint in FA 50 (Fig. lb.).3615000o. 12500fl 10000.07500c c-0 50002500—1E-- 0 nM Estrogen10 nM Estrogen—4-1000 nM Estrogen0 2 4 6 8 10 12 14 16 18 200 nM Estrogen10 nM Estrogen—0-1000 nM Estrogen150001250010000750050002500Figure I a.Figure lb. 0 2 4 6 8 10 12 14 16 18 20Time after last medium change (h)371a.^e ff e G h f Pri a Cul No ma ed roso I • .• 1 • IIReduction MammoplastiesThe effects of different concentrations of E2 on the growth of cultured cells from reductionmammoplasties are summarized in Table I. Also included in Table I are the results ofexperiments examining the effects of E2 on the growth of MCF-7 breast cancer cells. Theseexperiments served as a positive control for estrogen activity as MCF-7 are well characterizedas being growth stimulated by E2 at the concentrations examined [Lippman and Dickson,19891 In none of the 8 experiments examining the effect of E2 at concentrations ranging from1 to 1000 nM was any significant stimulation of growth of the primary cultures of HMECobserved. However, in 3 of 8 experiments there was a significant inhibition (p < 0.05) ofgrowth. Inhibition of growth was generally observed at the high (pharmacological)concentrations of E2, but was also seen at the lower (physiological) concentrations of E2 in afew individual cases. Examples of the inhibitory effects of E2 on the growth of HMEC fromreduction mammoplasties are presented in Figure 2.Two experiments examined the effects of E2 on the growth of normal HMEC at differenttimes in vitro and the results are presented in Figures 3a and 3b. In both experiments, thenumber of days in vitro had no significant effect on E2 regulation of cell growth. However, theSEM increased with time in culture. Cell growth was therefore ideally measured as early aspossible in an attempt to minimize the variation within individual conditions.38T^ Ifrom reduction mammoplasties. Specimen no. (div aConcentration of E2 (nM)0 1 10 100 1000Redn 7 (5 div) 100±5b 95±3 100± 5 109± 15 115±24Redn 8 (5 div) 100±3 90± 11 68± 13* 75± 8* 60± 10*Redn 9 (6 div) 100 ± 7 98 ± 12 84 ± 17 91 ± 8 82 ± 8Redn 10 (6 div) 100 ± 20 112 ± 33 97 ± 11 101 ± 11 92 ± 7Redn 11 (7 div) 100 ± 9 90 ± 11 92 ± 12 63 ± 19* 69 ± 18Redn 12 (8 div) 100 ± 22 N.D•c 80 ± 26 N.D. 80 ± 11Redn 13 (8 div) 100±7 72±9* 60±5* N.D. 62 ± 6*Redn 14 (9 div) 100 ± 27 98 ± 6 77 ± 4 N.D. 83 ± 3MCF-7 100 ± 9 174 ± 16 126 ± 6Summary 100 ±12 91 ±12 85 ± 12 85 ±12 80 ± 11n=8d n=7 n=8 n=5 n=8a : div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = 0 nMestrogen). Values are expressed as the mean ± SEM of each condition done in triplicate.c : N.D. = no data for these conditions due to limitations in cell numbers from various samplesd : n = number of individual samples examined in each condition* significant inhibition of growth (P < 0.05)39I on the Growth of primary cultures of human mammary epithelial cells (HMEC) from 2 reduction mammoplasties (Redn 10& Redn 11). Growth was assessed by the incorporation of 3H-TdR (1 1 Ci/m1) into the cells over a 4 hperiod. The absolute levels of incorporation have been converted to % of control values, withthe level of incorporation into the control condition (0 nM E2 in the serum-free medium,Appendix 7) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. E2 caused no significant effects on the growth of Redn10, however a significant ( p < 0.05) inhibition of growth of Redn 11 is shown at 100 nM E 2.The inhibition at 1000 nM E2 in Redn 11 is not statistically significant.- - a •40Figure 2.110 ^■^Redn 11 c 7 divr^Redn 13 c 8 div100 —90 —8070 perC060.c^500C.16."^40^—30 —2010 —0 A0^1^10^100^1000Concentration of E2 (nM)41HMEC from 2 reduction mammop asties (3a. Redn 8.. 3b. Redn 9). Growth was assessed by the incorporation of 3H-TdR (1 [LCi/m1) into the cells over a 4 hperiod. The absolute levels of incorporation have been converted to % of control values, withthe level of incorporation into the control condition (0 nM E2 in the serum-free medium,Appendix 7) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. Growth of Redn 8 was significantly (p < 0.05) inhibitedby 10, 100 and 1000 nM E2 at 5 div (Fig. 3a.). Due to the considerably higher SEM values at14 div, the inhibitory effects of E2 are not significant at any of the concentrations examined.Growth of Redn 9 was significantly (p < 0.05) inhibited by 1000 nM E2 at 6 div, however, therewere no significant effects of E2 at any concentration on the growth of Redn 9 on either 8 or 11div (Fig. 3b.).4210 100^100010^10010 1000■ Redn 8 c 5 divRedn 8 c 14 div--r-■ Redn 9 c 6 divRedn 9 c 8 divRedn 9 c 11 div0 1Figure 3a. 130120110O 1004.2 90• 8070tg 60=0 406 3020100Figure 3b. 1301201100100▪ 900• 80• 70ate 605-5 50.40a 3020100Concentration of E2 (nM)43• 1^1^• ill •^11 %. • 1 • 1FibroadenomksThe effects of different concentrations of E2 on the growth of HMEC from fibroadenomas, atype of benign breast disease, are shown in Table II. In none of 19 samples was a significantstimulation of growth by E2 observed at any of the concentrations used. In a total of 5 of 10samples used to examine the effects of E2 at 1000 nM, a significant (p < 0.05) inhibition ofgrowth was observed. The results of two representative experiments demonstrating theseeffects are presented in Figure 4.MammaryCarcinomasThe effects of E2 on the growth of HMEC from ER+ mammary carcinomas wereexamined in 12 samples and the results are shown in Table III. In none of the 12 carcinomasamples examined was a significant stimulation of growth by E2 observed at any of theconcentrations examined. Significant inhibition of growth by E2 at 1000 nM was observed in 4of 8 samples. The results of two individual experiments demonstrating these effects arepresented in Figure 5.11 11^1^ 1^• it^• a^ 11- AA summary of all the experiments examining the effects of E2 on the growth of HMECfrom the 3 types of tissue is presented in Table IV. E2 failed to stimulate the growth of any ofthe cultures at all concentrations examined. There is no significant differences between thedifferent mammary tissue types. However, there is a trend towards inhibition of growth at thehigher concentrations of E2.44on the growth of primary cultures of HMECfrom fibroadenomasSpecimen no. (div)Concentration of E2 (nM)0 1 10 100 1000FA 43 (8 div)a 100 ± 17b N.D.c 146 ± 26 N.D. N.D.FA 45 (8 div) 100 ± 8 N.D. 104 ± 15 N.D. N.D.FA 47 (8 div) 100 ± 21 92 ± 24 81 ± 23 39 ± 15* 31 ± 6*FA 48 (9 div) 100 ± 15 93 ± 31 156 ± 44 130 17 113 ± 46FA 49 (7 div) 100 ± 10 N.D. 94 ± 7 N.D. 96 ± 14FA 50 (10 div) 100 ± 12 N.D. 66 ± 2* N.D. 63 ± 12*FA 51 (11 div) 100 ± 9 N.D. 129 ± 17 N.D. N.D.FA 52 (8 div) 100 ± 37 101 ± 22 102 ± 20 92 ± 7 76 ± 30FA 54 (9 div) 100 ± 11 N.D. 81 ± 30 N.D. 80 ± 25FA 58 (9 div) 100 ± 22 N.D. 136 ± 27 N.D. N.D.FA 59 (9 div) 100 ± 16 N.D. 125 ± 16 N.D. N.D.FA 60 (8 div) 100 ± 14 N.D. 101 ± 10 N.D. 100 ± 4FA 62 (11 div) 100 ± 40 N.D. 160 ± 35 N.D. 44 ± 29FA 63 (11 div) 100 ± 8 N.D. 82 ± 6* N.D. N.D.FA 65 (11 div) 100 ± 4 N.D. 89 ± 2* N.D. N.D.FA 67 (11 div) 100 ± 15 N.D. 60 ± 4* N.D. 61 ± 5*FA 81 (7 div) 100 ± 28 N.D. 110 ± 25 N.D. N.D.FA 82 (7 div) 100 ± 17 N.D. 74 ± 6 N.D. N.D.FA 86 (9 div) 100 ± 20 N.D. 111 ± 18 N.D. 92 ± 22Summary 100 ± 17 95 ± 25 105 ±18 90±13 74 ± 21n = 18d n = 3 n = 18 n = 3 n = 9a: div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = 0 nMestrogen). Values are expressed as the mean ± SEM of each condition done in triplicate.c : N.D. = no data for these conditions due to limitations in cell numbers from various samplesd : n = number of individual samples examined in each condition* significant inhibition of growth (P < 0.05)45-_ •^- 1- • • 1^•^•^1 • •^11 -^_ • i llfrom carcinomas.Specimen no. (div)aConcentration of E2 (nM)0 10 100 1000HMC 96 (13 div) 100 ± 16b 93 ± 17 N.D. 82 ± 13HMC 97 (13 div) 100 ± 5 N.D•c 92 ± 10 N.D.HMC 101 (11 div) 100± 15 114±23 74±35 48± 7*HMC 102 (11 div) 100±39 92± 1 59±19 53±2*HMC 107 (11 div) 100±58 37±8 N.D. 33 ± 4*HMC 108 (13 div) 100±23 57±21 N.D. 84 ± 38HMC 109 (12 div) 100±40 130±29 N.D. N.D.HMC 110 (12 div) 100 ± 10 139 ± 21 N.D. N.D.HMC 111 (13 div) 100±38 112± 19 N.D. 105 ± 17HMC 127 (15 div) 100± 15 120±24 N.D. N.D.HMC 129 (13 div) 100 ± 26 89 ± 24 N.D. 81 ± 16HMC 130 (13 div) 100 ± 7 56 ± 10* N.D. 27 ± 4*Summary 100 = 24 94=18 73=21 64 = 14n=12d n=11 n=3 n=8a : div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = 0 nMestrogen). Values are expressed as the mean ± SEM of each condition done in triplicatec : N.D. = no data for these conditions due to limitations in cell numbers from various samplesd : n = number of individual samples examined in each condition* significant inhibition of growth (P < 0.05)46• 1^1^• • IL 1 •^• 11^ • 1fibroadenomas ( 47 & 67). Growth was assessed by the incorporation of 3H-TdR (1 RCi/m1) into the cells over a 4 hperiod. The absolute levels of incorporation have been converted to % of control values, withthe level of incorporation into the control condition (0 nM E2 in the serum-free medium,Appendix 7) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. E2 at 100 and 1000 nM concentrations significantly (p <0.05) inhibited the growth of FA 47, whereas E2 at 10 and 1000 nM significantly (p < 0.05)inhibited the growth of FA 67.47■ FA 47 c 8 div22 FA 67 c 11 div0^1^10^100^1000Concentration of E2 (nM)Figure 4.13012011010090O.%06.80C0a"c;tt70moio6050LI40302010048 hlo - - I - on e I •mammary carcinomas (HMC 101 and 102). Growth was assessed by the incorporation of 3H-TdR (1 [A.Ci/m1) into the cells over a 4 hperiod. The absolute levels of incorporation have been converted to % of control values, withthe level of incorporation into the control condition (0 nM E2 in the serum-free medium,Appendix 7) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. E2 at 1000 nM caused a significant (p < 0.05) inhibitionof the growth of both HMC 101 and HMC 102.49Figure 5. ^140 ^^130 ^120 ^110 ^100 —0,_^90...c00 80 ^"a ^-de 703 600..5040 —3020 —100 ^■ HMC 101 c 11 divz HMC 102 c 11 div0^10^100^1000Concentration of E2 (nM)5011 11 -.. i^•^1^L.^• a• o 1 . 4 r s• 1 1^• • II 1 • •of HMEC from the 3 mammary tissue typesConcentration of E2 (nM)Tissue Type^0^1^10^100^1000Reduction^100 -± 12a^91 ± 12^85 ± 12^85 ± 12^80± 11n = 11 12^n = 10^n = 11^n = 8^n = 11MammoplastiesFibroadenomas^100 ± 17^95 ± 25^105 ± 18^90 ± 13^74 ± 21n = 19^n = 3^n = 19^n = 3^n = 10Carcinomas 100 ± 24 94 ± 18^73 ± 21^64 ± 14N.D. on = 12^n = 11^n = 3^n = 8Summary^100 ± 18^92 ± 15^97 ± 16^84 ± 14^74 ± 15n = 42^n = 13^n = 41^n = 14^n = 29a : Values shown represent a percentage of growth relative to the control (control = 0 nMestrogen). Values are the average of individual experiments, and are expressed as the mean ±SEM of each tissue typeb : n = number of individual samples examined in each conditionc : N.D. = no data for these conditions due to limitations in cell numbers from various samples51To determine if the high concentration (10 µg/ml) of insulin in our serum-free medium wasmasking a growth stimulation by E2, 8 experiments comparing the growth effects of E2 inserum-free medium with 10 µg/ml of insulin and medium with 0.1 µg/ml of insulin wereconducted using HMEC from 7 fibroadenomas and a carcinoma. Two additional experimentsinvestigated the effects of insulin on the growth of HMEC from carcinoma samples, however,insufficient cell numbers did not allow examination of the effects of E2 in these cultures. Theresults of these experiments are presented in Table V. In all of the cases examined, thereduction in the concentration of insulin significantly (p < 0.05) reduced the growth of culturesof HMEC from both fibroadenomas and carcinomas. Reduction in growth ranged from 33% to93%, the average reduction being 59%. The effects of E2 in low concentrations of insulin weregenerally the same as the effects in high concentrations of insulin. There was no significantstimulation of growth in 7 of the 8 cases examined. Cells from 3 fibroadenomas, FA 63, FA 65,and FA 67, were growth inhibited by E2 in medium containing 10 µg/ml insulin. However, incells from FA 67 a stimulation of growth was observed in response to the same concentrationof E2 in 0.1 µg/ml insulin (Figure 6), that inhibited growth in 10 µg/ml insulin, as alreadydiscussed.52of HMEC from fibroadenomas and carcinomas.  Concentration of E2 (nM)Specimen no.^Concentration of(div)^insulin (µg/m1) 0^10^1000FA 58 (9 div)aFA 59 (9 div)FA 60 (8 div)FA 62 (11 div)FA 63 (11 div)FA 65 (11 div)FA 67 (11 div)HMC 127 (15 div)HMC 140 (11 div)HMC 141 (13 div)10^100 ± 22b^136 ± 27^N.D•c0.1 67 ± 8 44 ± 10 N.D.10^100 ± 16^125 ± 16^91 ± 210.1 51 ±6 63 ± 2 N.D.10^100 ± 14^101 ± 10^100 ± 40.1 56±1414 46 ± 8 52 ± 1310^100 ± 40^160 ± 35^44 ± 290.1 17 ± 1 18 ± 4 19 ± 410^100 ± 8^82 ± 6*^N.D.0.1 7 ± 2 4 ± 3 N.D.10^100 ± 4^89 ± 3*^N.D.0.1 68 ± 15 50 ± 3 N.D.10^100 ± 13^60 ± 4*^61 ± 5*0.1 35± 1 48 ± 5**^54 ± 4**10^100 ± 15^120 ± 24 N.D.0.1 41 ± 8 30 ± 11^N.D.10^100 ± 13^N.D. N.D.0.1 38 ± 17 N.D.^N.D.10^100 ± 14^N.D. N.D.0.1 31 ± 2* N.D.^N.D.Summary 10 µg/ml0.111,g/m1 100 ± 16^107 ± 1641±7 39 ± 6(n=10)d^(n=8)54 ± 15 (n=4)42 ± 7 (n=3)a : div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = 0 nMestrogen). Values are expressed as the mean ± SEM of each condition done in triplicatec : N.D. = no data for these conditions due to limitations in cell numbers from various samplesd : n = number of individual samples examined in each condition* significant inhibition of growth (P < 0.05)** significant stimulation of growth compared to 0.1 ng/ml insulin with 0 nM estrogen (p < 0.05)53a fibroadenoma A 67). Growth was assessed by the incorporation of 3H-TdR (1 iCi/mI) into the cells over a 4 hperiod. The absolute levels of incorporation have been converted to % of control values, withthe level of incorporation into the control condition (0 nM E2 in the serum-free medium,Appendix 7) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. In medium containing 10 µg/ml insulin, E 2 at 10 and1000 nM caused a significant inhibition (p < 0.05) of growth in FA 67. In medium containing0.1 itg/m1 insulin, E2 at 10 and 1000 nM caused a significant stimulation of growth compared to0 nM E2 in 0.1 µg/ml medium.54Figure 6,120110■ 10 ug/mI InsulinP2 0.1 ug/ml Insulin100 —90....O 60Z00 70I rft 60f.3o 504030 —20 —1000 10 1000Concentration of E2552. The Effects of Epidermal Growth Factor (EGF) and Transforming Growth Factor-a(TGF-a) on the Growth of HMEC in Primary CultureAs discussed in the introduction, EGF and TGF-a have been shown to be involved in thedirect control of both normal rodent MEC growth in vivo (EGF) and human breast cancer cellline growth in vitro (EGF and TGF-a). For this reason the effects of these two growth factorson the growth of HMEC in serum-free primary culture were examined.2a. The Effects of EGF on the Growth of Primary Cultures of HMEC from Reduction MammoplastiesThe effects of different concentrations of EGF on the growth of HMEC from 3 reductionmammoplasties were examined and the results are shown in Figure 7. In all cases, EGFsignificantly (p < 0.05) stimulated growth in a dose-dependent manner. Stimulation rangedfrom 188% to 1698% of the growth in control cultures without EGF (average stimulation ofgrowth by EGF was 894 % of controls). Although stimulation of growth was seen at allconcentrations examined, ranging from 1 to 100 ng/ml of EGF, peak stimulation of growth wasobserved between 5 - 10 ng/ml. Based on these findings, a value of 10 ng/ml was chosen tostudy the effects of EGF on HMEC in subsequent experiments.2b. The Effects of EGF and TGF-a on the Growth of Primary Cultures of HMEC From FibroadenomasHMEC from 12 fibroadenomas were examined for the effects of EGF on growth. In allcases, EGF was found to stimulate growth significantly (p < 0.05). The results of theseexperiments are shown in Table VI. The degree of growth stimulation ranged from 133% to3455% (average stimulation of growth was 923%) of the levels observed in control culturesgrown without EGF. The growth effects of TGF-a were compared to those of EGF in culturesof HMEC from 6 fibroadenomas (Table VI).56Figure 7 : The effects of EGF on the growth of primary cultures of HMEC from 3reduction mammoplasties (Redn 12.13 and 14). Growth was assessed by the incorporation of 3H-TdR (11.1.Ci/m1) into the cells over a 4 hperiod. The absolute levels of incorporation have been converted to % of control values, withthe level of incorporation into the control condition (0 ng/ml EGF in the serum-free medium,Appendix 7) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. EGF significantly (p < 0.05) stimulated growth in a dose-dependent manner (1-100 ng/ml). Peak stimulation of growth was observed between 5 - 10ng/ml.570 5 200 1 10 10020001500■ Redn 14 c 9 div5000Concentration of EGF (ng/mI)--.400 —300  ■ Redn 12^c0f Zx0° 200a -ae10000900800 ^ ■ Redn 13^c700 ^77; 600 ^-2. 0 500xo-;-,-400 ^300200 ^1000z a .8 div II 15^10^408 div58Table VI : The effects of EGF and TGF-a on the Growth of primary cultures of HMEC from fibroadenomasCulture conditionsSpecimen no.(div) Control + EGF (10 ng/ml) + TGF-a (10 ng/ml)FA 43 (8 div)a 100 ± 17b 167 ± 43 N.D•cFA 45 (8 div) 100 ± 8 197 ± 15* N.D.FA 51 (11 div) 100 ± 9 1047 ± 283* N.D.FA 52 (8 div) 100 ± 37 188 ± 22* N.D.FA 54 (9 div) 100 ± 11 753 ± 161* 2006± 57**FA 61 (8 div) 100 ± 17 284 ± 67* 394 -± 12*FA 74 (11 div) 100 ± 6 1616 ± 182* 3984 ± 181**FA 80 (11 div) 100 ± 23 589 ± 99* 1113 -± 119**FA 81 (7 div) 100 ± 28 3455 ± 583* N.D.FA 82 (7 div) 100 ± 17 2203± 284* N.D.FA 84 (13 div) 100± 19 583 ± 58* 392 ± 32*FA 85 (13 div) 100 ± 16 615 ± 109* 543* ± 4SUMMARY 100 ± 17 (n=12)d 923 ± 159 (n=12) 1088 ± 63(n=6)a : div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = 0 addedgrowth factor). Values are expressed as the mean ± SEM of each condition done in triplicatec : N.D. = no data for these conditions due to limitations in cell numbers from various samplesd : n = number of individual samples examined in each condition* significant stimulation of growth compared to control with 0 ng/ml growth factor (p < 0.05)**growth significantly greater than corresponding + EGF condition (p < 0.05)59TGF-a was found to be significantly (p < 0.05) growth stimulating in all culturesexamined. Although individual samples varied greatly in response to TGF-a , TGF-a wasequal to, or greater than EGF in stimulating growth of HMEC from fibroadenomas. Stimulationof growth ranged from 392% to 3984% (average was 1088%) of control cultures, whichcontained no added growth factor. A difference in the dose responses between EGF andTGF-a was also observed in 2 experiments comparing their effects on HMEC growth (Figure8a and 8b). TGF-a appeared to be active over a lower concentration range, as indicated bythe drop in stimulation of HMEC growth at 100 ng/ml compared to 10 ng/ml. In contrast toTGF-a, EGF effects were similar or greater at 100 ng/ml than at 10 ng/ml of EGF.c. The Effects of EGF and TGF-a on the Growth of Primary Cultures of HMEC From CarcinomasPrimary cultures of HMEC from 9 carcinomas were examined for the effects of EGF oncell growth and the results are shown in Table VII. The effects of TGF-a on growth were alsoinvestigated in HMEC from 2 of the carcinomas where sufficient cells were obtained. Theseresults are shown in Figure 9. In all cases examined, EGF and TGF-a significantly (p < 0.05)stimulated growth. The degree of stimulation by EGF ranged from 150% to 2912% (averagewas 1033%) of the controls with no EGF. TGF-a stimulated growth 330% and 639% overcontrols. In the 2 cases where the effects of EGF and TGF-a were determined, bothstimulated growth to the same magnitude. Table VIII presents a summary of the averagegrowth effects of EGF and TGF-a on HMEC from the three different tissue types.60Figure 8 : The effects of EGF and TGF-a on the arowth of primary cultures of HMEC from ? fibroadenomas ( FA 74 & 80 ). Growth was assessed 14 h after the last medium change, by the incorporation of 3H-TdR (1 RCi/mI) into the cells over a 4 h time period. The absolute levels of incorporation have beenconverted to % of control values, with the level of incorporation in the control condition (0 ng/mlgrowth factor in the serum-free medium, Appendix 7) being assigned a value of 100 %. Valuesshown represent the mean ± SEM of triplicate wells of the same condition. EGF and TGF-a at10 and 100 ng/ml both significantly (p < 0.05) stimulated the growth of HMEC from FA 74 (Fig.8a) and FA 80 (Fig. 8b). In both examples TGF-a at 10 ng/ml stimulated growth to a greaterdegree than EGF at the same concentration. At 100 ng/ml the stimulation by TGF-a wassignificantly (p < 0.05) reduced relative to level of stimulation by the same growth factor at 10ng/ml. EGF at 100 ng/ml caused a greater increase in growth than EGF at 10 ng/ml, however,the effect was only significant in cells from FA 74 (Fig. 8a.).61—R3— FA 74 + EGF-a- FA 74 + TGF-aFA 80 + EGF0 FA 80 + TGF-a'^I'^I'^I'^I'^I'^I'^I'^I'^I'^I'^I1200110010009000 800. C 700oV 6005004003002001000Figure 8a. 40003500i"-; 3000-00 25000(917 2000150010005000Figure 8b. 0 10 20 30 40 50 60 70 80 90 100 1100 10 20 30 40 50 60 70 80 90 100 110Concentration (ng/mI)62Table VII : The effects of EGF on the growth of primary cultures of HMEC from carcinomasCulture conditionsSpecimen no.(div) Control + EGF (10 ng/ml)HMC 101 (11 div)a 100 ± 15b 2912 ± 643*HMC 102 (11 div) 100 ± 39 668 ± 91*HMC 107 (11 div) 100 ± 58 1582 ± 533*HMC 108 (13 div) 100 ± 23 150 ± 17*HMC 109 (12 div) 100 ± 40 1457 ± 403*HMC 110 (12 div) 100 ± 10 612 ± 20*HMC 111 (11 div) 100 ± 38 1014 ± 130*HMC 140 (11 div) 100 ± 14 403 ± 16*HMC 141 (13 div) 100 ± 13 502 ± 168*Summary 100 = 28 (n=9)C 1033 = 224 (n=9)a : div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = 0 addedgrowth factor). Values are expressed as the mean ± SEM of each condition done in triplicatec : n = number of individual samples examined in each condition* significant stimulation of growth (p < 0.05)63Figure 9 : The effects of EGF and TGF-a on the arowth of primary cultures of HMEC from 2 fibroadenomas ( FA 74 & 80 ). Growth was assessed 14 h after the first medium change, by the incorporation of 3H-TdR(1 pei/m1) into the cells over a 4 h time period. The absolute levels of incorporation have beenconverted to % of control values, with the level of incorporation in the control condition (0 ng/mlgrowth factor in the serum-free medium, Appendix 7) being assigned a value of 100 %. Valuesshown represent the mean ± SEM of triplicate wells of the same condition. EGF and TGF-a at10 ng/ml both significantly (p < 0.05) stimulated the growth of HMEC from HMC 140 and 141.There was no difference in the degree of growth stimulation by EGF and TGF-a.64■ HMC 140 c 11 divr HMC 141 c 13 divcontrol EGF7006005004003002001000TGF-aFigure 9. Culture conditions65Table VIII : Summary of the effects of EGF and TGF-a on the growth of primary culturesof HMEC from reduction mammoplasties. fibroadenomas and carcinomasAverage percentage of growth stimulationTissue Type Control + EGF [10 ng/m1] + TGF-a [10ng/ml]ReductionMammoplastiesFibroadenomasCarcinomas100 ± 12a(n=11)b100 ± 17 (n=19)100 ± 24 (n=24)894 ± 97 (n=3)*923 ± 159 (n=12)*1033 ± 224 (n=12)*N.D•c1088 ± 63 (n=6)*484 ± 168 (n=2)*a : Values are the averages of individual experiments, and are expressed as the mean ± SEMof each tissue typeb : n = number of individual samples examined in each conditionc : N.D. = no data for these conditions due to limitations in cell numbers from various samples* significant stimulation of growth compared to control with 0 ng/ml growth factor (p < 0.05)662d. The Effects of a Monoclonal Antibody to EGFR on Growth Stimulation by EGFTwo experiments utilized HMEC from fibroadenomas to investigate the effects of EGFand TGF-a immediately following the first medium change. The rational for these experimentswas to determine if these growth factors were able to stimulate the growth of HMEC after only1 day of exposure to them. These findings were relevant in planning experiments investigatingthe effects of a monoclonal antibody (MAb) against the EGFR on the growth of cells incubatedwith EGF. It would be ideal to investigate the effects of the MAb after a minimal amount oftime in vitro in order to minimize the amount of antibody required for each experiment. Theresults of these experiments are shown in Figure 10. Sixteen hours after the addition of EGFor TGF-a, growth was only significantly (p < 0.05) stimulated by EGF in HMEC from FA 78.The effects of MAb 528, a monoclonal antibody which competitively blocks EGF bindingto EGFR, were examined in cultures of HMEC from 2 fibroadenomas and the results shown inFigures lla and 11b. Since these growth factors do not significantly stimulate growth afteronly 1 medium change (Figure 10), the medium in these experiments was changed every otherday for 6 days. Fresh growth factor and antibody were added with each change. The MAbreduced the degree of growth stimulation in both cases examined, however, the reduction instimulation was statistically significant (p < 0.05) for HMEC from only one of the fibroadenomas(FA 81, Figure 11a.).67Figure 10 : The effects of EGF and TGF-a on the growth of primary cultures of HMECfrom 2 mammary carcinomas (HMC 140 & HMC 141). Growth was assessed 14 h after the last medium change, by the incorporation of 3H-TdR(1 [A,Ci/m1) into the cells over a 4 h time period. The absolute levels of incorporation have beenconverted to % of control values, with the level of incorporation in the control condition (0 ng/mlgrowth factor in the serum-free medium, Appendix 7) being assigned a value of 100 %. Valuesshown represent the mean ± SEM of triplicate wells of the same condition. EGF caused asignificant (p < 0.05) stimulation of growth in HMEC from FA 78. None of the other treatmentscaused any significant effects on the growth of HMEC from either FA 77 or FA 78.68Control + TGF-aFigure 10. 300■ FA 77 c 3 divF2 FA 78 c 3 div2500 • • •-.0....c0U0iit.0. 1 .X01 .a20015010050 —0+ EGFCulture Conditions69Figure 11 : The effect of MAb 528 on growth stimulation by EGF. The effects of MAb 528, which competitively blocks EGF binding to EGFR, and EGF aton the growth of primary cultures of HMEC isolated from 2 fibroadenoma samples (FA 81,Fig.11a., FA 82, Fig.12a.) was examined. Growth was assessed by the incorporation of 3H-TdR (1 [LCi/m1) into the cells over a 4 h time period. The absolute levels of incorporation havebeen converted to % of control values, with the level of incorporation in the control condition (0ng/ml EGF) being assigned a value of 100 %. Values shown represent the mean ± SEM oftriplicate wells of the same condition. EGF caused a significant (p < 0.05) stimulation of growthin both FA 81 and FA 82. MAb 528 reduced the degree of stimulation by EGF in both cases,however, this effect was statistically significant only in FA 81 (Fig.11a.).70Figure 11 a.  45003750300022501500 —7500■ FA 81 - MAb 528ra FA 81 + MAb 528.....v-r-z- zy,-,zControl + EGFFigure 11 b. 30002500,0 2000mz5 00° 1500iiT,:ft 10005000• FA 82 - MAb 528z FA 82 + MAb 528ControlI+ EGFCulture conditions71From Reduction Mammop asties. Fibroadenomas and CarcinomasThe effects of E2 and EGF added in combination were examined in primary cultures ofHMEC from 1 reduction mammoplasty, 6 fibroadenomas and 2 carcinomas. The results ofthese experiments are shown in Table IX. There was no additive or synergistic effect of E2and EGF in combination in HMEC from normal tissue; that is, the growth effect of E2 and EGFtogether was the same as the stimulation observed in response to EGF alone. In 2 of the 6cultures of HMEC from fibroadenomas, a significant (p < 0.05) synergistic effect on thestimulation of growth by the addition of E2 and EGF in combination was observed. The resultsof 1 experiment are illustrated in Figure 12. A synergistic effect of E2 plus EGF was alsoobserved in HMEC from 1 of the 2 carcinomas and the results of this experiment are illustratedin Figure 13.Prior to the initiation of these experiments measuring growth responses of HMEC to E2and EGF, a number of experiments were attempted to determine the effects of these sametreatments on gene expression. Though these experiments were unsuccessful, in oneexperiment sufficient cells were grown to yield cell weight measurements for each of theexperimental conditions. These results are presented in Figure 14. Each condition consistedof only one culture dish, therefore, it is not possible to determine the statistical error in thesemeasurements. However, these results are not presented as statistically significant, but ratheras supportive evidence of a synergistic effect of E2 plus EGF on the growth of HMEC from afibroadenom a.72 G •I le •r• 11 • •^re• •A^I" "I"HMEC from a reduction mammoplasty. fibroadenomas and carcinomasCulture ConditionsSpecimen no.(div)a Control + E2 (10 nM) + EGF (10ng/ml)+ E2 (10 nM)+ EGF (10 ng/ml)Redn 12 (8 div) 100 = 22b 80 = 26 264 ± 30* 292 ± 12*FA 43 (8 div) 100 = 17 146 ± 26 167 = 43* 167 = 30*FA 45 (8 div) 100 = 12 120 = 10 197 = 15* 201 = 23*FA 51 (11 div) 100 = 9 129 = 17 1047 = 283* 1643 = 245**FA 54 (9 div) 100 = 11 81 = 30 753 ± 161* 1866 = 147**FA 81 (7 div) 100 ± 28 110 = 25 3455 = 583* 3249 = 138"FA 82 (7 div) 100 = 17 74 = 6 2203 = 284* 1917 = 331*HMC 108 (13 div) 100 ± 23 57 = 21 150 = 17* 200 = 39*HMC 111 (13 div) 100 ± 40 112 = 19 1457 = 403* 3195 ± 795**a : div = days in vitrob : Values shown represent a percentage of growth relative to the control (control = serum-free medium with no added E2 or EGF, Appendix 7). Values are expressed as the mean =SEM of each condition done in triplicate* significant (p < 0.05) stimulation of growth compared to the control**significant (p < 0.05) stimulation of growth compared to the plus EGF condition730 r C f o11 1 •^.•^11- 1111.fibroadenomas (Fig. 12a.. FA 51. Fig. 12b. FA 54). Growth was assessed by the incorporation of 3H-TdR (1 RCi/m1) into the cells over a 4 htime period. The absolute levels of incorporation have been converted to % of control values,with the level of incorporation in the control condition (0 nM E2 and 0 ng/ml EGF ) beingassigned a value of 100 %. Values shown represent the mean ± SEM of triplicate wells of thesame condition. E2 alone caused no significant effect on growth in either FA 51 or FA 54. EGFcaused a significant (p < 0.05) stimulation of growth in both FA 51 and FA 54. E2 and EGF incombination caused a significantly greater (p < 0.05) increase in growth over EGF alone inboth samples.742000■ FA 51 c 0 EGFm FA 51 c 10 ng/mI EGF15001000 —500 —00 10Figure 12a. Figure 12b.  2000 ■ FA 54 c 0 EGFr FA 54 c 10 ng/mI EGF1500100050000^ 10Concentration of E2 (nM)75Figure 13 : The effects of E2 plus EGF on the growth of primary cultures of HMEC from amammary carcinoma (HMC 111). Growth was assessed by the incorporation of 3H-TdR (1 tiCi/m1) into the cells over a 4 htime period. The absolute levels of incorporation have been converted to % of control values,with the level of incorporation in the control condition (0 nM E2 and 0 ng/ml EGF ) beingassigned a value of 100 %. Values shown represent the mean ± SEM of triplicate wells of thesame condition. E2 alone caused no significant effect on the growth of HMC 111. EGF causeda significant (p < 0.05) stimulation of growth. E2 and EGF in combination caused a significantlygreater (p < 0.05) increase in growth than EGF alone.76Figure 13. 4000■ 0 EGFn 10 ng/mI EGF3000200010000 N.D.0 10^1000Concentration of E2 (nM)77f^I s G on t^ C ffibroadenoma l A 41). Cells were harvested from the dishes by mild trypsinization and the cell suspension waspelleted by centrifugation. Cell pellets were then weighed. Estrogen at 10 nM caused a 10%reduction in cell weight. EGF at 10 ng/ml caused a 46% increase in cell weight. Estrogen andEGF in combination caused a 100% increase in cell weight.78Figure 14. 20■ 0 ng/ml EGF2 10 ng/ml EGF0 10Concentration of E2 (nM)79GFeeder Layers of Mitomvcin-C Treated FibroblastsSince E2 does not appear to have a direct effect on the growth of HMEC , the effects ofE2 may be mediated by production of paracrine growth factors by fibroblasts or may requiredirect contact of HMEC with mammary fibroblasts. For these reasons, three preliminarystudies examined the effects of E2 and/or EGF on the growth of HMEC grown on mitomycin-Ctreated fibroblasts feeder layers. A dose response curve for mitomycin-C demonstrated that1.0 - 10.0 µg/ml mitomycin-C rendered the fibroblasts incapable of growth. However, thehigher concentrations of mitomycin-C (5.0 and 10.0 µg/ml) were observed to cause celldetachment and death. For this reason fibroblasts were treated with 1 µg/ml of mitomycin-C.After mitomycin-C treatment, the fibroblasts remained in serum-free medium for a further 24 hto remove traces of serum or mitomycin-C. The results of these studies are presented inFigures 15 and 16. E2 at 1000 nM significantly (p < 0.05) inhibited the growth of HMEC fromFA 87 grown on mitomycin-C treated fibroblasts (Figure 15). EGF was observed to stimulatethe growth of FA 87 grown on fibroblasts. The growth of HMEC from HMC 130 on collagen-coated dishes was significantly inhibited by 1000 nM E2 (Figure 16). In contrast to this E2 hadno effects on the growth of HMEC from the same sample grown on fibroblasts. HMC 129grown on either collagen or fibroblasts did not respond to E2 (Figure 16).800 1^1^• 'kit'^•^11 s^• 11fibroadenoma l A 87 :11 dive. Cells were grown on a monolayer of mitomycin-C treated fibroblasts. Growth wasassessed by the incorporation of 3H-TdR (1 RCi/m1) into the cultures over a 4 h period. Theabsolute levels of incorporation have been converted to % of control values, with the level ofincorporation in the control condition ( 0 nM E2 & 0 ng/ml EGF) being assigned a value of 100%. Values shown represent the mean ± SEM of triplicate wells of the same condition. E2 at1000 nM caused a significant inhibition (p < 0.05) of growth in FA 87. EGF at 10 ng/ml causeda significant stimulation (p < 0.05) of growth.810 10 1000■ 0 ng/ml EGF2 10 ng/ml EGFFigure 15Concentration of E2 (nM)82carcinomas (HMC 29 and HMC 130). Cells were grown on either dehydrated collagen or a monolayer of mitomycin-C treatedfibroblasts. Growth was assessed by the incorporation of 3H-TdR (RCi/m1) into the culturesover a 4 h period. The absolute levels of incorporation have been converted to % of controlvalues, with the level of incorporation in the control condition (0 nM E2) being assigned a valueof 100 %. Values shown represent the mean ± SEM of wells of the same condition. E 2 at1000 nM caused a significant inhibition (p < 0.05) of growth in HMC 130 grown on collagen,however in none of the other conditions was any significant effect of estrogen observed.83Figure 16120100800a-Z00"O" 60st.c50 403._020■ HMC 129 on collagentz HMC 129 on fibroblastsKi HMC 130 on collagenal HMC 130 on fibroblasts0 10 1000Concentration of E2 (nM)84CHAPTER 4 DISCUSSIONel^1"^• it I •^iiOur studies as well as those of others have shown that E2 at physiologicalconcentrations can stimulate the growth of HMEC in primary cultures in the presence of serum[Emerman et al., 1990; Mauvais-Jarvis et al., 1986; Calaf et al., 1986b; Longman andBeuhring, 1987], however, we were unable to demonstrate a stimulatory effect on the growthof cells from normal, benign and malignant mammary gland tissues in the serum-free mediumused in the experiments described in this thesis (Appendix 7). This is consistent withpreviously published reports showing no effect of E2 on the growth of human or rodent MEC inserum-free primary culture [Yang et al., 1982; Imagawa et al., 1985; Hahm and Ip, 1990].In the carcinoma samples, there was no growth response to E2 in spite of the fact that allsamples were ER+. Unfortunately, ER levels were not determined for the normal andfibroadenoma samples. However, both of these tissue types have been shown to containsignificant numbers of ER+ cells in a majority of specimens [Petersen et al., 1987; Giri et al.,1989]. Furthermore, it has been shown that ER+ cells from normal samples andfibroadenomas cultured under similar conditions to ours remain ER+ throughout the cultureperiod [Balakrishnan et al., 1987; Malet et al., 1991]. Therefore, it is unlikely that the absenceof ER+ cells in our samples is a factor related to the absence of estrogen-stimulated growth.We are planning to measure the levels of ER in the original tissues and to study the effects ofour cell-culture conditions on ER levels using the immunohistochemical ER detection kitproduced by Abbot laboratories [Malet et al., 1991]. Assuming that the absence of anestrogenic stimulation of growth in serum-free medium is not due to a lack of ER, the data85suggest that either factors present in the medium block or mask an estrogenic stimulation ofgrowth or, alternatively, that factors absent from the medium are required for an E2 effect.We examined the possibility that the pharmacological concentration of insulin in ourserum-free-medium (10 tA. g/m1) was masking an estrogenic effect on cell growth, as has beendemonstrated for the MCF-7 breast cancer cell line [van der Burg et al., 1988; Ruedl et al.,1990]. A 100-fold reduction of the initial insulin concentration caused a significant reduction incell growth in all of the cases examined (Table V). Differences in the growth response of thecells to E2 were observed in the low-insulin medium. E2 failed to inhibit growth in the lowinsulin medium and there was a E2 stimulation of growth observed in a fibroadenoma (Fig. 6.).These experiments suggest that insulin is indeed blocking E2 growth-stimulating effects. It ispossible that high insulin concentrations block a growth regulatory pathway normally utilized inE2 stimulation of growth.The role of insulin in the development of alveolar structures, or simply for maintenance ofmammary epithelial cells in vitro, is unclear. Early studies showed that injections of E2 andprogesterone into diabetic (no significant insulin present) male rabbits induced the formation ofextensive lobuloalveolar structures in mammary gland tissue [Norgren et al., 1968]. Manystudies have shown that mammary epithelial cells in serum-free medium are stimulated toreplicate by the addition of insulin [Stockdale and Topper, 1966; Wang and Amor, 1971], butchanges in the cells' response to insulin during the time in vitro have led to debate over thephysiological relevance of insulin to mammary gland tissue [Friedberg et al., 1970].Friedberg's group observed that mammary epithelial cells on day 1 in vitro were insensitive toinsulin, but after day 2 in vitro the cells acquired insulin sensitivity. Although this findingsuggests that insulin sensitivity in vitro is an acquired effect, it is equally possible that thetrauma to the tissue during transplant procedures or culturing the tissue temporarily rendersthe mammary epithelial tissue incapable of responding to insulin.86Although insulin does not appear to be required for normal mammary gland developmentin vivo, the large number of studies showing a variety of effects of insulin on mammaryepithelial cells in vitro suggest a role for insulin in mammary gland development. Insulin addedto serum-free explant cultures of mammary glands causes extension of cell viability andstimulation of DNA synthesis [Oka et al., 1974]. The in vitro studies of insulin effects on MECalso support a role for insulin in lactational activity. The presence of insulin is required incombination with prolactin and a glucocorticoid to induce terminal differentiation (milk productproduction) in mammary explant cultures of hormonally-primed virgin mice [Oka et al., 1974].Furthermore, addition of insulin to the medium of MEC cultures stimulates an increase in theamount of rough endoplasmic reticulum, an increase in the number of Golgi complexes, milkprotein production and lactose synthetase activity [Mills and Topper, 1970; Emerman et al.,1977; Emerman and Pitelka, 1977; Katiyar et al., 1978; Vonderhaar, 1977].It has been demonstrated that insulin at 10 µg/ml interacts with the insulin-like growthfactor 1 (IGF-1) receptor [Rechler et al., 1986]. Therefore, studies demonstrating therequirement of high concentrations of insulin for significant growth of serum-free cultures ofMEC may have actually been demonstrating a requirement for IGF-1 [Ethier et al., 1987,Deeks et al., 1988]. Addition of IGF-1 to serum-free cultures of E2-dependent breast cancercell lines has been found to be growth stimulatory [Karey and Sirbasku, 1988], and it will beimportant to determine if lower concentrations of IGF-1 are able to replace the highconcentrations of insulin used in our serum-free cultures. The importance of the IGF-1receptor in E2-responsive growth is implicated by the finding that E2 stimulation of growth inhuman breast cancer cell lines is accompanied by increased production of IGF-1 [Huff et al.,1986]. It is possible that in E2 stimulation of growth in HMEC may involve production of growthfactors (eg. IGF-1) that interact with the IGF-1 receptor, but this interaction is blocked in thepresence of high concentrations of insulin due to insulin occupation of the IGF-1 receptor [vander Burg et al., 1990].87Using the same serum-free medium used in the experiments of this thesis, but withouthydrocortisone, Yang and coworkers fail to show any growth effect by E2 [Yang et al., 1982].Although their study examined only two normal specimens and one cancerous sample, theexperiments suggest that hydrocortisone is not blocking or masking a growth effect of E2.However, these experiments utilized the high concentration of insulin, which could still block ormask an E2 effect on growth, even in the absence of hydrocortisone. Clearly a more detailedstudy is required to conclude that hydrocortisone does not modulate E2 responses in vitro.The effects of cholera toxin in the serum-free medium on E2 regulation of growth are alsonot known. The addition of cholera toxin to the medium is based on the finding that itselectively stimulates the growth of the MEC, whereas mammary stromal cells do not grow inits presence [Taylor-Papadimitriou et al., 1980]. The growth stimulatory effects of cholera toxinon MEC are thought to be due to increased intracellular levels of cAMP, as analogues of cAMPare also able to stimulate the growth of MEC. No studies have been done to show howcholera toxin may modulate the growth effects of E2 or other mammotrophic hormones in vitro.Our lab is currently investigating the effects of significantly reducing or eliminating cholera toxinfrom the serum-free medium in order to eliminate the problem of uncharacterized effects ofcholera-toxin. It will be interesting to compare the growth effects of E2 and other hormones inmedium without cholera toxin present. It is possible that via an increase in intracellular cAMPlevels, cholera toxin could mask growth modulatory effects of any hormones or growth factorsalso operating through a cAMP second messenger system.In summary, the medium components required to support growth in serum-free primaryculture are quite likely responsible, at least in part, for the finding that E2 is unable to stimulatethe proliferation of HMEC from any of the samples examined (with the notable exception ofone fibroadenom a in the low insulin medium). However, the finding that addition of EGF to themedium is consistently able to stimulate the proliferation of HMEC from normal, benign and88malignant tissues demonstrates that there is no factor in the medium which renders HMECincapable of growing at increased rates. The growth stimulation by EGF also shows that theabsence of stimulation by E2 is not due to the fact that the cells have already reached amaximal growth rate.As already discussed, E2 stimulation of HMEC growth in serum containing media haspreviously been demonstrated by us [Emerman et al., 1990] and others [Mauvais-Jarvis et al.,1986; Calaf et al., 1986b; Longman and Beuhring, 1987]. The E2 stimulation of growth that wereported was observed in medium containing dextran charcoal-treated serum. This treatmentremoves steroid hormones present in the serum, but does not remove peptide growth factors.However, if the serum is further chemically treated to break disulfide bridges, and thereforeinactivate peptide growth factors, serum is no longer able to support E2 stimulation ofproliferation to the same magnitude [Ruedl et al., 1990 ]. This finding suggests that there isone or more peptide factors present in serum which are required for E2 stimulation of growth invitro; it is likely that the significant levels of EGF in serum are involved in this effect (discussedin more detail later in the Discussion).The absence of other factors in the serum-free culture system used in these experimentsmay also be related to the absence of growth stimulation by E2. One factor that is verydifferent in vitro than in vivo and is directly involved in modulating both growth anddifferentiation responses to hormones is the extracellular substrate to which the HMECadheres [Blum et al., 1989]. The ability of a biological substrate to modulate MECdifferentiation in vitro was clearly demonstrated when Emerman and coworkers cultured normalmouse MEC on floating collagen Type I gels [Emerman et al., 1977; Emerman and Pitelka,1977]. Previously, cytodifferentiation of these cells in vitro had not been possible. However,when grown on floating collagen gels the MEC could be induced to undergo cytodifferentiation.Hormonal regulation of casein synthesis and secretion was also demonstrated. In a89modification of the technique developed by Emerman et al., Yang and co-workers culturedMEC inside of 3-dimensional collagen Type I gels. In their system, MEC was shown to bothproliferate and differentiate into duct-like structures [Yang et al. 1980 and 1982]. Although acollagen Type I substrate provides a sufficient substrate for cell growth and cytodifferentiation,it may not be sufficiently able to mimic the in vivo situation to allow E2 stimulation of growth.Co-culture studies involving both fibroblasts and/or adipocytes grown together withepithelial cells provide an in vitro condition more like the in vivo setting. When mouse MEC arecultured in 5 % serum on irradiated adipocytes (3T3-L1), the growth level is greater than that ofcells grown on non-adipocyte 3T3 cells [Levine and Stockdale, 1984]. The effect of theadipocytes is partially due to direct substrate effects, as demonstrated by the observation thatcell-free ECM preparations of the adipocytes are able to stimulate the growth of mouse MEC.A role for soluble factors produced by the adipocytes is also shown by the ability of conditionedmedia from adipocytes to stimulate growth of the mouse MEC. In another paper by the sameresearchers, adipocytes as a substrate are shown to induce duct-like morphogenesis and theproduction of a basement membrane [Wiens et al., 1987]. In experiments similar to those ofEmerman et al. [1977] and Emerman and Pitelka [1977], these researchers showed thataddition of lactogenic hormones to the medium of mouse MEC cultured on 3T3-L1 adipocytesare able to stimulate secretory differentiation, as indicated by morphological criteria [Wiens etal., 1987]. The ability of hormones to stimulate secretory differentiation requires that theadipocytes be alive, in comparison to the stimulation of growth which is observed in cellsgrown on lethally irradiated adipocytes.Other studies have examined the ability of mammary fibroblasts to modulate both growthand differentiation responses of MEC to E2 in co-culture experiments. Haslam's group showedthat E2 stimulation of progesterone receptor synthesis does not require the presence of livefibroblasts [Haslam and Levely, 1985, Haslam, 1986]. Gluteraldehyde treated fibroblasts,90fibroblast conditioned medium or collagen Type I coating of the plates are all sufficient to allowE2 stimulation of progesterone receptor levels. In contrast to the progesterone receptorstimulation, E2 stimulation of DNA synthesis requires the presence high numbers of livefibroblasts, or if the number of fibroblasts is reduced, direct contact of the fibroblast andepithelial cells is required. The reason for the apparent discrepancies between the findings ofHaslam's group, who showed that live fibroblasts are required for E2 stimulation of growth, andWien's group, who showed that lethally irradiated adipocytes, but not live 3T3 fibroblasts, aresufficient is not clear. The most intuitive hypothesis is that the findings in Haslam's laboratoryare more relevant as the stromal cells used in those studies were of mammary origin, asopposed to the 3T3 derivatives used in the studies by Levine and Stockdale [1984] and Wienset al. [1987].The different in vitro requirements for the two different E2 responses studied by Haslamand co-workers are likely indicative of the existence of at least two separate mechanisms of E2action. More specifically, it is likely that E2 stimulation of progesterone receptors is via amechanism involving alteration of the extracellular matrix composition. Collagen type I, asubstitute for the fibroblasts in E2 stimulation of progesterone receptors, is a substratesufficient to allow production of extracellular matrix components by MEC. Cells grown onplastic do not produce a basement membrane and progesterone receptors are not stimulatedin MEC grown on plastic. The growth stimulatory effects of E2 require the presence of livefibroblasts. This observation supports the hypothesis that E2 stimulation of growth is viaproduction of diffusible paracrine-acting growth factors by E2-responsive cells. The presenceof multiple mechanisms of E2 action on the parameter of cell growth is also demonstrated bythe findings in this thesis that E2 is able to significantly inhibit the growth of HMEC from alltissue types. This finding demonstrates that growth inhibition by high dose E2 is not affectedby the conditions which prevented E2 stimulation of growth.91Although preliminary, the results of the experiments in this thesis examining the growth ofHMEC grown on mitomycin-C treated normal mammary fibroblasts were exciting. As shown inFigure 16, the growth of HMEC from a carcinoma sample was significantly inhibited by E 2when grown on collagen Type 1 but not when grown on mitomycin-C treated fibroblasts.However, E2 did inhibit the growth of HMEC from a fibroadenoma grown on mitomycin-Ctreated fibroblasts (Figure 15), showing that the inhibitory effects of high dose E2 are notblocked by the presence of mitomycin-C treated mammary fibroblasts. Further experimentsare required to determine if the growth effects of E2 observed in cells grown on collagen Type Iare different in cells grown on mitomycin-C treated fibroblasts. It was expected that growthwould be inhibited by high dose E2 in cells grown on mammary fibroblasts, as in vivo growth ofmammary tumor cells in breast cancer patients is frequently inhibited by high dose E2 [Clarkeet al., 1990]. The inhibition of tumor growth by high dose E2 was the rationale for high doseE2 as a treatment modality.Haslam's group also demonstrated that in co-culture experiments of stromal andepithelial cells, E2 also stimulates DNA synthesis in the fibroblasts [Haslam, 1986]. This effectis also dependent on the presence of live epithelial cells. The bidirectionality of E2 stimulationhas also been observed in vivo where E2 stimulation of DNA synthesis is observed to occurfirst in the stromal cells followed several hours later by stimulation in the epithelial cells[Shyamala and Ferenczy, 1984]. Bidirectionality of these responses is supportive of a E2growth response model incorporating the production of a locally diffusible growth factor(s).McGrath [1983] utilized histoautoradiography to demonstrate that in mixed cultures of MECand mammary fibroblasts from mice, E2 stimulated DNA synthesis in MEC only where there isdirect contact or very close juxtaposition between the fibroblasts and epithelial cells. However,he noted that the growth-stimulatory effect of E2 is not common to all colonies of cells in thecultures and, at this time, it is difficult to interpret the data presented by McGrath.92The results of the studies on E2 modulation of HMEC growth presented in this thesishave led to the proposal of a number of important studies to be carried out in the work leadingto completion of my Ph.D. research. As already mentioned briefly, I am interested in furtherpursuing the experiments on the effects of insulin concentration on HMEC growth. If IGF-1 atlow concentrations can replace the high levels of insulin required for maintenance of HMEC inserum-free primary culture, then the growth effects of E2 will be examined in the new medium.The effect of reducing the hydrocortisone levels on E2 stimulation of growth in low insulinmedium will also be examined. Another goal of my future research is also to modify the serum-free medium composition presently used in order to minimize or eliminate the presence ofcholera toxin. I will investigate the ability of cAMP or related analogues to replace choleratoxin, if in fact they are required at all, prior to continuation of further growth studies.The results of the co-culture studies also suggest a number of follow-up studies. It hasalready been discussed that E2 stimulation of HMEC growth in vivo is preceded by stimulationof growth in the surrounding stromal cells [Shyamala and Ferenczy, 1984] and that E2-stimulated growth of HMEC in vitro is accompanied by DNA synthesis in the fibroblast feederlayer [Haslam, 1986]. This observation supports a critical role of the stromal cells. I amplanning to compare the E2 regulation of growth of HMEC seeded directly on to feeder layersof mitomycin-C treated mammary fibroblasts or irradiated fibroblasts, HMEC seeded on tonontreated mammary fibroblasts and HMEC co-cultured with actively dividing mammaryfibroblasts but physically separated from them by microporous collagen-coated filters. This willdetermine if actively dividing fibroblasts are needed for E2 stimulation of HMEC growth and ifdirect contact with the fibroblasts is required or if locally diffusible paracrine growth factors willresult in E2 growth regulation. I am also planning to compare the growth of HMEC grown onfeeder layers of mammary fibroblasts from all three types of tissue utilized in this thesis, as it ispossible that defects in the stromal response to growth regulatory signals are involved in thederegulation of growth in the epithelial cells comprising the mammary tumor.932. The Effects EGF and TGF-a on the Growth of HMECEGF has been shown to stimulate MEC growth, both in vivo [Gardner et al., 1989,Coleman et al., 1988] and in vitro [Stoker et al., 1976, Stampfer et al., 1980]. In mice itappears that EGF plays a role in both the initiation and maintenance of the breast cancerprocess [Kurachi et al., 1985]. Our results indicate that EGF is indeed a potent and directmitogen for normal, benign, and malignant HMEC in serum-free primary culture. Although themajority of studies examining the effect of EGF and TGF-a on MEC growth demonstrate agrowth-stimulatory effect, two studies have shown the opposite results [Ehman et al., 1984;Yang et al., 1986]. Ehman's group observed an inhibitory effect of EGF on the growth of MECgrown on irradiated fibroblasts in serum-containing media. This result is difficult to interpret,particularly in light of the results presented here, showing a growth-stimulatory effect of EGFon HMEC grown on both collagen Type I and mitomycin-C treated fibroblasts. It has beensuggested that the irradiated fibroblasts in Ehman's study were perhaps producing growth-inhibitory factors in response to EGF [Imagawa et al., 1990]. The explanation is purelyspeculative; however, the serum-free medium used in this thesis is the same as that describedby Yang et al. [1982]. Using this medium, Yang et aL [1986] failed to see any growth-stimulatory effect of EGF, unless the HMEC are cultured in 3-dimensional collagen gels orhydrocortisone is deleted from the medium. The reason for this discrepancy is not clear, giventhe high degree of similarity in protocol between the experiments in this thesis and those ofYang's group.The variation in response to EGF among individual samples (167-3455%) is ofconsiderable interest due to the large degree of variability of receptor levels observed inmammary tumor biopsy samples [Nicholson et al., 1988]. Currently we are using thetetrazolium dye-reduction (MTT) assay to measure growth rather than 3H-TdR incorporation,which was used prior to completion of this thesis. The MTT assay uses far fewer cells per94experimental condition, so it is now possible to measure both EGF-binding to EGFR and EGFgrowth-stimulating effects on HMEC from the small tissue samples that we receive from theoperating room. The ability to compare both of these parameters will allow us to investigate ifdifferences in EGF-binding (receptor levels) can explain the large variability in the degree ofgrowth stimulation by EGF.The variation in magnitude of growth stimulation by EGF may also be due to differencesamong the samples in the levels of other receptors that also interact directly or indirectly withEGF in stimulating cell growth. Synergism between EGF and IGF-1 has been shownpreviously in cultures of bovine MEC [Shamay et al., 1988]. The interaction of highconcentrations of insulin used in our serum-free medium with the IGF-1 receptor has alreadybeen discussed with respect to its potential role in the inhibition of growth stimulation byphysiological concentrations of E2. With regard to the variation in the magnitude of responsesto EGF, it is necessary to consider further the possibility that high concentrations of EGF +insulin may mimic the synergistic effect of EGF + IGF-1 observed in primary cultures of bovineMEC [Shamay et al., 1988]. If this is indeed occurring, then variations in IGF-1 receptor levelsamong individual samples could cause variable degrees of synergism with EGF, which couldbe a factor contributing to the large variability in the magnitude of growth stimulation by EGF.TGF-a also significantly stimulates the growth of HMEC in primary culture (Table VIII).However, samples vary as to whether TGF-a is equal to or more potent than EGF instimulating HMEC growth. At present there is no explanation for the differences in the growth-promoting activity between EGF and TGF-a, considering they are generally thought to act via acommon receptor pathway [Korc et al., 1991]. One possibility is that interaction betweenexogenously added growth factors and endogenously produced factors may occur, withdifferent interactions between EGF and TGF-a and endogenous growth factors Both EGF andTGF-a, as well as numerous other factors such as platelet-derived growth factor, IGF-1, IGF-2,95and transforming growth factor-p, are all likely candidates for such endogenously producedgrowth factors [Davidson and Lippman, 1989]. Synergism between EGF and IGF-1 hasalready been demonstrated [Shamay et al., 1988]; however, it was not investigated if TGF-aalso synergized with IGF-1. If EGF and TGF-a differ in the degree to which they synergizewith IGF-1 (or high concentrations of insulin), then this could account for the differences inEGF and TGF-a growth responses as well. It is worth noting that TGF-a is also a more potentmediator than EGF in the stimulation of both bone resorption and neovascularization [lbbotsonet al., 1986; Schreiber et al., 1986].Another possibility is that EGF and TGF-a interact in different ways with EGFR [King,1988]. Relatively little is known about the mechanism of growth stimulation by EGF or TGF-a.The degree of growth stimulation in HMEC as presented in this thesis is considerably greaterthan that observed in the epidermoid carcinoma A431 cell line usually used to study themechanistic effects of EGF. A431 cells are growth inhibited by the concentrations of EGFused in this thesis, and are only stimulated at extremely low concentrations of EGF (pg/mI).The inhibitory effect of physiological levels of EGF on A431 is probably due to the extremelyamplified levels of EGFR in these cells. Given the finding that HMEC is so significantly growthstimulated by EGF at physiological levels, and the finding that we can successfully subcultureHMEC in the presence but not in the absence of EGF in our serum-free cultures (unpublishedresults, Emerman et al.), our cell culture system should provide an excellent model system inwhich to investigate the mechanistic effects of EGF and TGF-a in stimulating cell growth.The two experiments demonstrating a difference in the dose responses of EGF and TGF-a (Figures 8a and 8b) also indicate another variation in the growth stimulatory effects of thesetwo growth factors. The use of the MTT assay will allow us to characterize differences inbinding characteristics for both EGF and TGF-a with respect to their growth stimulatory effects.96The results of these experiments in this thesis are the first report of any such differencesbetween EGF and TGF-a in primary cultures of HMEC.In planning the experiments comparing hormone and growth factor effects in cells fromthe three different mammary gland tissue types, it was hypothesized that differences amongthe tissue types would be observed. The results of this thesis show that in general, cells frombreast cancers are not different from cells obtained from normal tissue or fibroadenomas intheir growth response to E2 or EGF (Tables IV & VIII). However, it is possible that suchdifferences may occur between normal and cancerous cells from individual samples. Bycomparing growth responses of normal cells obtained distal to the tumor site and cells from thetumor proper from individual mastectomy samples, it will be possible to address this questiondirectly. Alternatively, gross differences in responses to individual factors may play a small rolein malignancy as compared to multiple defects in growth responses to a number of factors.Future experiments investigating the effects of multiple growth factors and hormones aloneand in combination will address this issue.•^1^• it 1 • iiiA synergistic effect on growth of E2 and EGF on cells from several samples wasobserved (Table IX). This finding gives support to the idea that E2 may act as a modulator ofcellular responses to growth factors, rather than acting directly as a mitogen. The synergismwith E2 and EGF also supports the hypothesis presented earlier that factors absent in theserum-free medium but present in serum-containing medium are required for E2 growthstimulation in vitro. This hypothesis is further supported by the work of Ethier's group whohave shown that E2 can stimulate the growth of rat MEC in a serum-free medium containing 10ng/ml EGF [Ethier et al., 1987]. However, Ethier's group did not investigate the growth effectsof E2 in serum-free medium without EGF. Although E2 alone is unable to elicit a growth-97stimulatory effect, it is capable of enhancing the growth-stimulating effect of EGF in somesamples. E2 has been shown to regulate production of IGF-1 in MCF-7 cells [Huff et al.,1988], and EGF and IGF-1 [Shamay et al., 1988] can interact in a synergistic fashion.Therefore the synergistic effect of E2 and EGF could result from E2 stimulation of growthfactors and these growth factors could then directly synergize with the added EGF. However,it is unlikely that production of IGF-1 accounts for the synergistic effect observed in this thesis.The high concentrations of insulin in the serum-free medium are presumably already activatingthe IGF-1 receptor, and further effects through the receptor are therefore unlikely. Usingantibodies specific to the different growth factors stimulated by E2, it may be possible todetermine the role of various growth factors in the synergism of E2 and EGF. This strategy iscurrently being used in an attempt to determine the role of EGF and TGF - a in E2-stimulatedgrowth of breast cancer cell lines [Clarke et al., 1990]An E2-induced increase in EGFR levels could also explain the synergism between E2and EGF observed in our studies. Such increases in EGFR have been demonstrated in ER+cell lines [Dickson et al., 1986; Bates et al., 1990]. In uterine tissues it has been shown that E2is capable of modulating functional EGF and EGFR levels in vivo [Gardner et al., 1989; Huet-Hudson et al., 1990]. As already discussed, it is my intention to measure EGF-binding levels inHMEC in serum-free primary culture of HMEC. Investigating EGF binding levels in thepresence and absence of E2 will determine if the synergism between E2 and EGF is due tomodulation of EGF-binding characteristics by E2.4. Conclusions and Future DirectionsThe results of this thesis show that E2 may be incapable of directly stimulating HMECgrowth in a minimally supplemented serum-free culture medium. However, the results of theexperiments examining insulin concentration have shown that factors in the serum-free98medium can inhibit E2 stimulation of growth, and further experiments are required to establishwhat effects each of the individual supplements have on the growth effects of E2. In additionto the effects of medium supplements on modulation of E2 growth regulation, substrate effectshave also been demonstrated in this thesis. More complete investigation of the role of stromalcells in E2 regulation is required before any conclusions regarding the preliminary results in thisthesis can be discussed further.The results of this thesis also show that both EGF and TGF-a are potent growth-stimulatory factors for HMEC from normal, benign and malignant mammary tissue in serum-free primary culture. The results have shown a large degree of variability in the magnitude ofgrowth stimulation in response to EGF among individual samples. However, the differenttissue types did not differ significantly in their average responses to EGF. I have proposedexperiments to examine the potential causes of the variability among individual samples, aswell as additional experiments addressing the question of differences between normal andmalignant cells from the same patient. TGF-a is equal to or greater than EGF in its ability tostimulate HMEC growth. Preliminary studies in this thesis have implicated dose-responsedifferences as a potential explanation for the differences in the magnitude of growth responsesto EGF and TGF-a. Using the MU assay, I have proposed experiments to characterizefurther the differences in dose responsiveness to EGF and TGF-a.The results of this thesis showing that E2 can synergize with EGF in the growthstimulation of HMEC in serum-free primary culture support a role for E2 in growth regulation viathe modulation of HMEC responsiveness to growth factors. Alternatively, EGF may be acompetency factor required for E2 stimulation of growth, both in vitro and in vivo. I haveproposed to study the effects of E2 on EGF binding in order to determine if E2 alters HMECresponsiveness to EGF via modulation of EGFR levels.99In summary, the results of this thesis have directed my research interests to two specificand independent directions. I am interested in studying and characterizing further the cultureconditions which are required for E2 stimulation of HMEC growth in the absence of serum.Results of such studies may be able to identify the as of yet unidentified serum-factorresponsible for conveying E2 sensitivity to HMEC both in vitro and in vivo. The other researchinterest I have developed through completion of this thesis is regarding the nature of growthfactor stimulation of HMEC growth. I would like to study the mechanisms involved in thegrowth regulatory effects observed in response to both the E2 and growth factors, alone and incombination. The results presented in this thesis have been submitted and accepted forpublication in Experimental Cell Research in an article entitled "Hormone and growth factoreffectsl on the growth of human mammary epithelial cells in serum-free primary culture."100APPENDIX 1 : Transport Medium DME:F12 - ( 1:1 )Hepes buffer - 10 mMCalf serum - 5 %Insulin - 5 µg/mlDME - Dulbecco's Modified Eagles MediumAPPENDIX 2 : Freezina Medium DME - 50 %Calf serum - 44 %Dim ethylsulfoxide - 6 %APPENDIX 3 : Dissociation Medium DME:F12^- ( 1:1 )Hepes buffer^- 10 mMBSA - 2 %Insulin^- 5 µg/mlCollagenase^- 300 U/mIHyaluronidase^- 100 U/mlBSA - bovine serum albuminAPPENDIX 4 : Attachment Medium DME:F12^- ( 1:1 )Hepes buffer^- 10 mMPooled normal HuS - 5 %Insulin^- 5µg/mlHuS - human serum101APPENDIX 5 : Preparation of Rat Tail Collagen The collagen solution was prepared from rat tails by first placing the rat tails in 95 % ethanol for15 min. The tendons were dissected out and teased apart using scalpel blades and forceps,weighed and place in a 60 mm Petri dish containing sterile deionized water and exposed toultraviolet light in the laminar flow hood for 24 h. The fibers were then suspended in a diluteacetic acid solution ( 0.01 N) and stirred at 4°C for 48 h. They were left for another 24 h in thedilute acid solution without stirring at 4°C. The solution was transferred into 50 mlultracentrifuge tubes and spun in a Sorvall ultracentrifuge for 30 min at 10,000 x g. Thesupernatant consisted of the collagen solution and was bottled and stored at 4°C.APPENDIX 6 : Preparation of Pooled Normal Human Serum Serum samples were collected in the mornings from patients who had fasted over the previous8-12 h. Blood, received in non-heparinized tubes, was incubated for 30 min at 37° C, thencentrifuged at 100 x g and the serum collected. Pooled serum was stored at -20° C.APPENDIX 7 : Serum-free Medium DME:F12^- ( 1:1 )Hepes buffer^- 10 mMBSA - 5 mg/mlCholera toxin^- 10 ng/mlHydrocortisone^- 0.5 µg/mlInsulin^- 10.0 µg/ml102BIBLIOGRAPHY1. Aitken, S.C., and M.E. Lippman 1982 Hormonal regulation of net DNA synthesis in MCF-7 human breast cancer cells in tissue culture. Cancer Res., 42: 1727-1735.2. Arteaga, C.L., E. Coronado, and C.K. Osborne 1988 Blockade of the epidermal growthfactor receptor inhibits transforming growth factor-alpha induced but not estrogen-induced growth of hormone-dependent human breast cancer. Mol. Endocrinol., 2: 1064-1069.3. Balakrishnan, A., J. Yang, C.W. Beattie, T.K. das Gupta, and S. Nandi 1987 Estrogenreceptor in dissociated and cultured human breast fibroadenoma epithelial cells. CancerLett., 34: 233-242.4. Baldwin, R.L., and C.W. Daniels 1974 Enzymatic and metabolic changes in thedevelopment of lactation. In: Lactation: A Comprehensive Treatise. B.L. Larson and V.R.Smith, eds. Academic Press, New York, Vol 1, pp. 349-411.5. Banerjee, M.R., B.G. Wood, and D.L. Kinder 1973 Whole mammary gland organ culture:selection of appropriate gland. In vitro, 9: 129-133.6. Barker, S., C. Panahy, J.R. Puddefoot, A.W. Good, and G.P. Vinson 1989 Epidermalgrowth factor receptors and estrogen receptors in the non-malignant part of thecancerous breast. Br. J. Cancer, 60: 673-677.7. Bates, S.E., M.E. McManaway, M.E. Lippman, and R.B. Dickson 1986 Characterizationof estrogen responsive transforming activity in human breast cancer cell lines. CancerRes., 46: 1707-1713.8. Bates, S.E., N.E. Davidson, E.M. Valverius, C.E. Freter, R.B. Dickson, J.P. Tam, J.E.Kudlow, M.E. Lippman, and D.S. Salomon 1988 Expression of transforming growthfactor-a and its messenger ribonucleic acid in human breast cancer: its regulation byestrogen and its possible functional significance. Mol. Endocrinol., 2: 543-555.9. Bates, S.E., E.M. Valverius, B.W. Ennis, D.A. Bronzert, J.P. Sheridan, M.R. Stampfer, J.Mendelsohn, M.E. Lippman, and R.B. Dickson 1990 Expression of transforming growthfactor-alpha/epidermal growth factor receptor pathway in normal human breast epithelialcells. Endocrinology, 26: 596-607.10. Benthois, Y., JA. Katzenellenbogen, and B.S. Katzenellenbogen 1986 Phenol red intissue culture media is a weak estrogen: implications concerning the study of estrogenresponsive cells in culture. Proc. Natl. Acad. Sci. USA, 83: 2496-2500.11. Berthois, Y., X. Dong, F. Martin, and P.M. Martin 1989 Regulation of epidermal growthfactor-receptor by oestrogen and anti-oestrogen in the human breast cancer cell-lineMCF-7. Biochem. Biophys. Res. Commun., 159: 126-131.10312. Blum, J.L., M.E. Ziegler, and M.S. Wicha 1989 Regulation of mammary differentiation bythe extracellular matrix. Enviro. Health Perspect., 80: 71-84.13. Borellini, F., and T. Oka 1989 Growth control and differentiation in mammary epithelialcells. Enviro, Health Perspect., 80: 85-99.14. Bresciani, F. 1965 Effect of ovarian hormones on duration of DNA synthesis in cells ofthe C3H mouse mammary gland. Exp. Cell Res., 38: 13-32.15. Bresciani, F. 1968 Topography of DNA synthesis in the mammary gland of the C3Hmouse and its control by ovarian hormones: an autoradiographic study. Cell TissueKinet., 1: 51-63.16. Bresciani, F. 1971 Ovarian steroid control of cell proliferation in the mammary gland andcancer. In: Basic Actions of Sex Steroids on Target Organs. P. Hubinont, F. Leroy, andP. Galand, eds., Karger, Basel, pp.130-159.17. Brookreson, A.D., and C.W. Turner 1959 Normal growth of mammary gland in pregnantand lactating mice. Proc. Soc. Exp. Biol. Med., 102: 744-745.18. Brown, C.R., C.T. Teng, B.T. Pentecost, and R.B. Augustine 1989 Epidermal growthfactor precursor in mouse lactating mammary gland alveolar cells. Mol. Endocrinol., 3:1077-1083.19. Bullock, L., P.L. Barthe, I. Mowszowicz, D.N. Orth, and C.W. Bardin 1975 The effect ofprogestins on submaxillary gland epidermal growth factor: Demonstration of androgenic,synandrogenic and antiandrogenic actions. Endocrinology, 97: 189-195.20. Butler, J.E. 1974 Immunoglobulins of the mammary secretions. In: Lactation: AComprehensive Treatise, Vol. III. B.L. Larson and V.R. Smith, eds. Academic Press,New York, pp. 217-280.21. Calaf, G., I.H. Russo, L.D. Roi, and J. Russo 1986 Effects of peptides and steroidhormones on cell kinetic parameters of normal human breast tissue in organ culture. Invitro Cell. Dev. Biol., 22: 135-140.22. Calaf, G., F. Garrido, C. Moyano, and R. Rodriguez 1986 Influence of hormones on DNAsynthesis of breast tumors in culture. Breast Cancer Res. Treat., 8: 223-232.23. Carpenter, G., and S. Cohen 1979 Epidermal growth factor. Annu. Rev. Biochem., 48:198-216.24. Carpenter, G. 1987 Receptors for epidermal growth factor and other polypeptidemitogens. A. Rev. Biochem., 56: 881-914.25. Ceriani, R.L., and E.W. Blank 1977 Response to prolactin and ovarian steroids of normalmammary epithelial cell cultures. Mol. Cell. Endocrinol., 8: 95-103.10426. Chambon, M., G. Cavalie-Barthez, F. Veith, F. Vignon, R. Hallowes, and H. Rochefort1984 Effect of estradiol on nonmalignant human mammary epithelial cells in primaryculture. Cancer Res., 44: 5733-5743.27. Chen, C.L., and J. Meites 1970 Effects of estrogen and progesterone on serum andpituitary prolactin levels in ovariectomized rats. Endocrinology. 86: 503-505.28. Clarke, R., M.E. Lippman, and R.B. Dickson 1990 Mechanisms of hormone andcytotoxic drug interactions in the development and treatment of breast cancer. In:Molecular Endocrinology and Steroid Hormone Action, pp. 2243-278.29. Coleman, S., G.B. Siberstein, and C.W. Daniel 1988 Ductal morphogenesis in themouse mammary gland: evidence supporting a role of epidermal growth factor. Dev.Biol., 127: 304-315.30. Coleman, S., and C.W. Daniel 1990 Inhibition of mouse mammary ductalmorphogenesis and down regulation of the EGF receptor by epidermal growth factor.Dev. Biol., 137: 425-433.31. Connoly, J.M., and D.P. Rose 1988 Epidermal growth factor like proteins in breast fluidand human milk. Life Sci., 42: 1751-1756.32. Coussens, L., T.L. Yang-fen, Y.C. Liao, 1985 Tyrosine kinase receptor with extensivehomology to the EGF receptor shares chromosomal location with the neu oncogene.Science, 230: 1132-1139.33. Cutts, J.H., and R.L. Noble 1964 Estrone-induced mammary tumors in the rat. I.Induction and behavior of the tumors. Cancer Res., 24: 1116-1123.34. Daniel, C.W., G.B. Silberstein, and P. Strickland 1987 Direct action of 173-estradiol onmouse mammary ducts analyzed by sustained release implants and steroidautoradiography. Cancer Res., 47: 6052-6057.35. Dao, T.L. 1962 The role of ovarian hormones in initiating the induction of mammarycancer in rats by polynuclear hydrocarbons. Cancer Res., 22: 973-981.36. Darbre, P., J. Yates, S. Curtis, and R.J.B. King 1983 Effect of estradiol on human breastcancer cells in culture. Cancer Res., 43: 349-354.37. Darbre, P. 1990 Steroids and steroid receptors in growth control of cultured breastcancer cells. Int. J. Cancer, Suppl. 5: 67-75.38. David, G., and M. Bernfleld 1981 Type I collagen reduces the degradation of basallamina proteoglycan by mammary epithelial cells. J. Cell Biol., 91: 281-28639. Davidson, N.E., and M.E. Lippman 1989 The role of estrogens in growth regulation ofbreast cancer. Oncogenesis 1: 89-111.10540. Davis, C.L., and D.E. Bauman 1974 General metabolism associated with the synthesisof milk. In: Lactation: A Comprehensive Treatise. B.L. Larson and V.R. Smith, eds.Academic Press, New York, Vol 1, pp.3-30.41. Deeks, S., J. Richards, and S. Nandi 1988 Maintenance of normal rat mammaryepithelial cells by insulin and insulin-like growth factor 1. Exp. Cell Res., 174: 448-460.42. Dickson, R.B., K.K. Huff, E.M. Spencer, and M.E. Lippman 1985 Induction of epidermalgrowth factor-related peptides by 17-beta-estradiol in MCF-7 human breast cancer cells.Endocrinology, 118: 138-142.43. Dickson, R.B., H.K. Huff, M. McManaway, and M.E. Lippman 1986 Characterization ofestrogen responsive transforming activity in human breast cancer cell lines. CancerRes., 46: 1707-1713.44. Dilley, W.G., and S. Nandi 1968 Rat mammary gland differentiation in vitro in theabsence of steroids. Science, 161: 59-60.45. Drill, V.A. 1981 An overview of studies on estrogens, oral contraceptives and breastcancer. Prog. Drug Res., 25: 159-187.46. Dunnning, W.F., M.R. Curtis, and A. Segaloff 1947 Strain differences in response todiethylstilbesterol and the induction of mammary gland and bladder cancer in rats.Cancer Res., 7: 511-523.47. Edery, M., K. Pang, L. Larson, T. Colosi, and S. Nandi 1985 Epidermal growth factorreceptor levels in mouse mammary glands in various physiological states.Endocrinology, 117: 405-411.48. Ehmann, U.K., W.D.J. Peterson, and D.S. Misfeldt 1984 To grow mouse mammaryepithelial cells in culture. J. Cell Biol., 98: 1026-1032.49. Elias, J.J. 1957 Cultivation of adult mouse mammary gland in hormone-enrichedsynthetic medium. Science, 162: 842-844.50. Emerman, J.T., J. Enami, D.R. Pitelka, and S. Nandi 1977 Hormonal effects onintracellular and secreted casein in cultures of mouse mammary epithelial cells onfloating collagen membranes. Proc. Natl. Acad. Sci. USA, 74: 4466-4470.51. Emerman, J.T., and D.R. Pitelka 1977 Maintenance and induction of morphologicaldifferentiation in dissociated mammary epithelium on floating collagen gels. In vitro, 13:316-328.52. Emerman, J.T., and D.A. Wilkinson 1990 Routine culturing of normal, dysplastic, andmalignant human mammary epithelial cells from small tissue samples. In Vitro Cell.Devel. Biol. 26: 1186-1194.10653. Emerman, J.T., A.W. Tolcher, and P.M. Rebbeck 1990 In vitro sensitivity testing ofhuman breast cancer cells to hormones and chemotherapeutic agents. CancerChem other. Pharmacol., 26: 245-249.54. Edery, M., M. McGrath, L. Larson, and S. Nandi 1984 Correlation between in vitrogrowth and regulation of estrogen and progesterone receptors in rat mammary epithelialcells. Endocrinology, 115: 1691-1697.55. Enami, J., S. Enami, and M. Koga 1983 Growth of normal and neoplastic mousemammary epithelial cells in primary culture: stimulation by conditioned medium frommouse mammary fibroblasts. Gann, 74: 845-853.56. Engle, E.T., 1929 Pituitary-gonadal mechanism and heterosexual ovarian grafts. Am. J.Anat., 44: 121-134.57. Ethier, S.P. 1985 Primary culture and serial passage of normal and carcinogen-treatedrat mammary epithelial cells in vitro. J. Natl. Cancer Inst., 74: 1307-1318.58. Ethier, S.P. 1986 Serum-free culture conditions for the growth of normal rat mammaryepithelial cells in primary culture. In vitro Cell. Dev. Biol., 22: 485-490.59. Ethier, S.P., A. Kudla, and K.C. Cundiff 1987 Influence of hormone and growth factorinteractions on the proliferative potential of normal rat mammary epithelial cells in vitro.J. Cell. Physiol., 132: 161-167.60. Ethier, S.P., R.M. Summerfelt, K.C. Cundiff, and B.B. Asch 1990 The influence of growthfactors on the proliferative potential of normal and primary breast cancer-derived humanbreast epithelial cells. Breast Cancer Res. Treat., 17: 221-230.61. Ferguson, D.J. 1956 Endocrine control of mammary glands in C3H mice. Surgery, 39:30-36.62. Ferguson, D.J.P., and T.J. Anderson 1981 Morphological evaluation of cell turnover inrelation to the menstrual cycle in the "resting" human breast. Br. J. Cancer, 44: 177-181.63. Flux, D.S. 1954 Growth of the mammary duct system in intact and ovariectomized miceof the CHI strain. J. Endocrinol., 11: 223-237.64. Flux, D.S. 1954 The effect of adrenal steroids on the growth of the mammary glands,uteri, thymus, and adrenal glands of intact, ovariectomized and oestrone-treatedovariectomized mice. J. Endocrinol., 12: 57-68.65. Flux, D.S. 1958 Mammary gland growth in the male mice of the CHI strain afterhypophysectomy and castration. J. Endorinol., 17: 300-306.66. Friedberg, S.H., T. Oka, and Y.J. Topper 1970 Development of insulin-sensitivity by themouse mammary in vitro. Proc. Natl. Acad. Sci. USA, 67: 1493-1500.10767. Gardner, R.M., V. Verner, J.L. Kirkland, and G.M. Stancel 1989 Regulation of theepidermal growth factor (EGF) receptor by estrogen in the mature rat and during theestrous cycle. J. Steroid Biochem., 32: 339-343.68. Gardner, W.U. 1935 The effects of ovarian hormones and ovarian grafts upon themammary glands of male mice. Endocrinology, 19: 656-660.69. Ghosh-Dastidar, P., W.A. Coty, R.E. Griest, D.D. Woo, and C.F. Fox 1984 Progesteronereceptor sub-units are high affintiy substrates for phosphorylation by epidermal growthfactor receptor. Proc. Natl. Acad. Sci. USA, 81: 1654-1658.70. Giri, D.D., S.A.C. Dundas, J.F. Nottingham, and J.C.E. Underwood 1989 Oestrogenreceptors in benign epithelial lesions and intraduct carcinomas of the breast: animmunohistochemical study. Histopath., 15: 575-584.71. Gompel, A., C. Malet, P. Spritzer, J-P. Lolardrie, F. Kutten, and P. Mauvais-Jarvis 1986Progestin effect on cell proliferation and 17-beta-hydroxysteroid dehydrogenase activityin normal human breast cells in culture. J. Clin. Endo. Metab., 63: 1174-1180.72. Griffith, D.R., and C.W. Turner 1961 Normal growth of rat mammary glands duringpregnancy and early lactation. Proc. Soc. Exp. Biol. Med., 106: 448-450.73. Griffith, D.R., and C.W. Turner 1962 Effect of estrogen and progesterone upon milksecretion in normal and ovariectomized rats and on mammary gland DNA. Proc. Soc.Exp. Biol. Med., 110: 862-863.74. Grosvenor, C.E., and F. Mena 1974 Neural and hormonal control of milk secretion andejection. In: Lactation: A Comprehensive Treatise, Vol. I. B.L. Larson and V.R. Smith,eds. Academic Press, New York, pp. 227-276.75. Gullino, P.M., H.M. Pettigrew, and F.H. Grantham 1975 N-nitrosomethylurea asmammary gland carcinogen in rats. J. Natl. Cancer Inst., 54: 401-414.76. Hahm, H.A., and M.M. 1p 1990 Primary culture of normal rat mammary epithelial cellswithin a basement membrane matrix. I. Regulation of proliferation by hormones andgrowth factors. In Vitro Cell. Dev. Biol., 26: 791-802.77. Hallowes, R.C., P.S. Rudland, R.A. Hawkins, D.J. Lewis, D. Bennet, and H. Durbin 1977Comparison of the effects of hormones on DNA synthesis in cell cultures ofnonneoplastic and neoplastic mammary epithelium from rats. Cancer Res., 37: 2492-2504.78. Haslam, S.Z., and M.L. Levely 1985 Estrogen responsiveness of normal mousemammary cells in primary cell culture: Association of mammary fibroblast with estrogenicregulation of PgR. Endocrinology, 116: 1835-1844.79. Haslam, S.Z. 1986 Mammary fibroblast influence on normal mouse mammary epithelialcell responses to estrogen in vitro. Cancer Res., 46: 310-316.10880. Haslam, S.Z. 1987 Role of sex steroid hormones in normal mammary gland function. In:The Mammary Gland: Development, Regulation and Function. M.G. Neville and C.W.Daniel, eds., Plenum Press, New York, pp.499-533.81. Haslam, S.Z. 1988a Progesterone effects on deoxyribonucleic acid synthesis in normalmouse mammary glands. Endocrinology, 122: 464-470.82. Haslam, S.Z. 1988b Local vs. systemically mediated effects of estrogen on normalmammary epithelial cell deoxyribonucleic acid synthesis. Endocrinology, 122: 860-867.83. Hoshino, H. 1962 Morphogenesis and growth potential of mammary glands in mice. I.Transplantability and growth potentiality of mammary tissue in virgin mice. J. Natl.Cancer Inst., 29: 835-851.84. Huet-Hudson, Y.M., C. Chakraborty, S.W. Dey, Y. Suzuki, G.K. Andrews, and S.K. Dey1990 Estrogen regulates the synthesis of epidermal growth factor in mouse uterineepithelial cells. Mol. Endocrinol., 4: 510-523.85. Huff, K.K., D. Kaufman, K.H. Gabbay, E. Spencer, M.E. Lippman, and R.B. Dickson1986 Secretion of an insulin-like growth factor-1 related protein by human breast cancercells. Cancer Res., 46: 4613-4619.86. Huff, K.K., C. Knabbe, R. Lindsey, D. Kaufman, D. Bronzert, M.E. Lippman, and R.B.Dickson 1988 Multihormonal regulation of insulin-like growth factor-1-related protein inMCF-7 human breast cancer cells. Mol. Endocrin., 2: 200-208.87. Huggins, C., L.C. Grand, and F.P. Brillantes 1961 Mammary cancer induced by a singlefeeding of polynuclear hydrocarbons, and its suppression. Nature, 189: 204-207.88. Hunter, T., and JA. Cooper 1985 Protein tyrosine-kinases. Annu. Rev. Biochem., 54:897-930.89. Ibbotson, K.J., J. Harrod, M. Gowen, S. D'Souza, D.D. Smith, M.E. Winkler, R. Derynck,and G.R. Mundy 1986 Human recombinant transforming growth factor alpha stimulatesbone resorption and inhibits formation in vitro. Proc. Natl. Acad. Sci. USA, 83: 2228-2232.90. Ichinose, R.R., and S. Nandi 1964 Lobuloalveolar differentiation in mouse mammarytissues in vitro. Science, 145: 496-497.91. Ichinose, R.R., and S. Nandi 1966 Influence of hormones on lobuloalveolardifferentiation of mouse mammary glands in vitro. J. Clin. Endocrinol., 35: 331-340.92. Imagawa, W., Y. Tom ooka, and S. Nandi 1982 Serum-free growth of normal and tumormouse mammary epithelial in primary culture. Proc. Natl. Acad. Sci. USA, 79: 4074-4077.10993. Imagawa, W., Y. Tomooka, S. Hamamoto, and S. Nandi 1985 Stimulation of mammaryepithelial cell growth in vitro: Interaction of epidermal growth factor and mammogenichormones. Endocrinology, 116: 1514-1524.94. Imagawa, W., G.K. Bandyopadhyay, and S. Nandi 1990 Regulation of mammaryepithelial cell growth in mice and rats. Endocrine Rev., 11: 494-523.95. James, V.H.T., F. James, H. Braunsberg, W.T. Irvine, A.E. Carter, and D. Grose 1971Studies on the uptake of oestrogen by human breast tumors in vivo. In: SomeImplications of Steroid Hormones in Cancer. D.C. Williams and M.H. Briggs, eds., W.Heinemann, London, pp. 20-34.96. Jaspar, J.M., and P. Franchimont 1985 Radioimmunoassay of human epidermal growthfactor in human breast cyst fluid. Eur. J. Cancer Clin. Oncol., 21: 1343-1348.97. Jenness, R., 1974 The composition of milk. In: Lactation: A Comprehensive Treatise,Vol. III. B.L. Larson and V.R. Smith, eds. Academic Press, New York, pp. 3-107.98. Karey, K.P., and D.A. Sirbasku 1988 Differential responsiveness of human breastcancer cell lines MCF-7 and T47D to growth factors and 173-estradiol. Cancer Res., 48:4083-4692.99. Katiyar, V.N., J. Enami, and S. Nandi 1978 Effect of polypeptide hormones onstimulation of casein secretion by mouse mammary epithelial cells grown on floatingcollagen gels. In vitro, 14: 771-774.100.Kelsey J.L. 1979 A review of the epidemiology of breast cancer. Epidemiol. Rev., 1: 74-109.101.Kidwell, W.R., and J. Shaffer 1984 Growth stimulatory effects of unsaturated fatty acidsfor normal and neoplastic breast epithelium. J. Am. Oil Chem. Soc., 61: 1900-1907.102.Kimata, K., T. Sakakura, Y. Inaguma, M. Kato, and Y. Nishizuka 1985 Participation oftwo different mesenchyme in the developing mouse mammary gland: Synthesis ofbasement membrane components by fat pad precursor cells. J. Embryol. Exp. Morphol.,89: 243-257.103.King, C.R., I. Borello, F. Bellot, P. Comoglio, and J. Schlessinger 1988 EGF binding toits receptor triggers a rapid phosphorylation of the erbB-2 protein in the mammary tumorcell line SK-BR-3. EMBO J., 7: 1647-1651.104.King, R.J.B. 1989 Progression from steroid-responsive to unresponsive state in breastcancer. Horm. Res., 32: 254-256.105. Kohmoto, K., S. Tsunasawa, and F. Sakiyama 1984 Complete amino acid sequence ofmouse prolactin. Eur. J. Biochem., 138: 227-237.110106.Korc M., B. Chandrasekar, and G.L. Shah 1991 Differential binding and biologicalactivities of epidermal growth factor and transforming growth factor a in a humanpancreatic cancer cell line. Cancer Res., 51: 6243-6249.107.Korenman, S.G. 1980 Oestrogen window hypothesis of the etiology of breast cancer.Lancet, 1: 700-701.108.Kratochwil, K. 1977 Development and loss of androgen responsiveness in theembryonic rudiment of the mouse mammary gland. Dev. Biol., 61: 358-365.109.Kurachi, H., S. Okamoto, and T. Oka 1985 Evidence for involvement of thesubmandibular gland epidermal growth factor in mouse mammary tumorigenesis. Proc.Natl. Acad. Sci. USA., 82: 5940-5943.110. Kuramitzu, S., and L. Loeb 1921 The effect of suckling and castration on the lactatingmammary gland in rat and guinea pig. Am. J. Physiol., 56: 40-59.111.Lasfargues, E.Y. 1957 Cultivation and behavior in vitro of the normal mammaryepithelium of the adult mouse. Anat. Rec., 127: 117-129.112.Leiss, H.P. Jr., 1978 Epidemiology of breast cancer. In: The Breast. H.S. Gallager, H.P.Leiss, R.K. Snyderman, and J.Urban, eds. Mosby, St. Louis, pp. 37-48.113.Levay-Young, B.K., S. Hamamoto, W. Imagawa, and S. Nandi 1990 Caseinaccumulation in mouse mammary epithelial cells after growth stimulated by differenthormonal and nonhormonal agents. Endocrinology, 126: 1173-1182.114.Levine, J.F. and F.E. Stockdale 1984 3T3-L1 adipocytes promote the growth ofmammary epithelium. Exp. Cell Res., 151: 112-122.115. Lippman, M.E., G. Bolan, and K. Huff 1976 The effects of estrogens and antiestrogenson hormone responsive human breast cancer cells in long term culture. Cancer Res.,36: 4595-4601.116.Lippman, M.E., R.B. Dickson, E.P. Gelmann, N. Rosen, C. Knabbe, S. Bates, E.Valverius, D. Bronzed, K. Huff, and A. Kasid 1988 Human breast carcinoma cellsproduce growth regulatory peptides. Prog. Breast Cancer Res. Ther., 35: 203-213.117.Lippman, M.E., and R.B. Dickson 1989 Mitogenic Regulation of Normal and MalignantBreast Epithelium. Yale J. Biol. Med., 62: 459-480.118.Liscia, D.S., G. Merlo, C. Fortunato, N. Kim, G.H. Smith, R. Callahan, and D.S. Salomon1990 Transforming growth factor-a messenger RNA localization in the developing ratand human mammary gland by in situ hybridization. Dev. Biol., 140: 123-129.119. Liu, S.C., B. Sanfilippo, I. Perroteau, R. Derynck, D.S. Salomon, and W.R. Kidwell 1987Expression of transforming growth factor a (TGF-a) in differentiated mammary tumors:estrogen induction of TGF-a production. Mol. Endocrinol., 1: 683-692.111120. Longman S.M., and G.C. Beuhring 1987 Oral contraceptives and breast cancer: in vitroeffect of contraceptive steroids on human mammary cell growth. Cancer, 59: 281-287.121.Lyons, W.R.C., C.H. Li, and R.E. Johnson 1958 The hormonal control of mammarygrowth and lactation. Recent Prog. Horm. Res., 14: 219-250.122.MacMahon, B., D. Trichopoulos, J. Brown, A.P. Anderson, K. Aoki, and P. Cole 1980Age at menarche, probability of ovulation and breast cancer. Int. J. Cancer, 29: 13-16.123. Malet, C., A. Gompel, P. Spritzer, N. Bricout, H. Yaneva, I. Mowszowicz, F. Kuttenn, andP. Mauvais-Jarvis 1988 Tamoxifen and hydroxytamoxifen isomers versus estradioleffects on normal human breast cells in culture. Cancer Res., 48: 7193-7199.124.Malet, C., A. Gompel, H. Yaneva, H. Cren, N. Fidji, I. Mowszowicz, F. Kuttenn, and P.Mauvais-Jarvis 1991 Estradiol and progesterone receptors in cultured normal humanbreast epithelial cells and fibroblasts: immunocytochemical studies. J. Clin. Endocrin.Metab., 73: 8-17.125.Martial, JA., R.A. Halewell, R.A. Baxter, and H.M. Goodman 1979 Human growthhormone: Complementary DNA cloning and expression in bacteria. Science, 205: 602-607.126.Massague, J. 1983 Epidermal growth factor-like transforming growth factor. II.Interaction with epidermal growth factor receptors in human placenta membranes andA431 cells. J. Biol. Chem., 258: 13614-13620.127. Masters, J.R.W., J.O. Drife, and J.J. Scarisbrick 1977 Cyclic variations of DNA synthesisin human breast epithelium. J. Natl. Cancer Inst., 58: 1263-1265.128.Mauvais-Jarvis, P., F. Kuttenn, A. Gompel, C. Malet, and S. Fournier 1986 Estradiol-progesterone interaction in normal and pathological human breast cells. Ann.Endocrinol. (Paris), 47: 179-187.129.McCormack, J.T., and G.S. Greenwald 1974 Progesterone and oestradiol-17-betaconcentrations in the peripheral plasma during pregnancy in the mouse. J. Endocrinol.,62: 101-107.130. McGrath, C.M. 1983 Augmentation of the responses of normal mammary epithelial cellsto estradiol by mammary stroma. Cancer Res., 43: 1355-1369.131.McGrath, M., S. Palmer, and S. Nandi 1985 Differential response of normal ratmammary epithelial cells to mammogenic hormones and EGF. J. Cell Physiol., 125: 182-191.132.McManus, J.M., and C.W. Welsch 1981 Hormone-induced ductal DNA synthesis ofhuman breast tissues maintained in the athymic mouse. Cancer Res., 41: 3300-3305.112133. McManus, J.M., and C.W. Welsch 1984 The effect of estrogen, progesterone, thyroxine,and human placental lactogen on DNA synthesis of human breast ductal epitheliummaintained in athymic nude mice. Cancer Res., 54: 1920-1927.134.Meyer, S.J. 1977 Cell proliferation in normal human breast ducts, fibroadenomas, andother ductal hyperplasias as measured by tritiated thymidine, effects of menstrual phase,age and oral contraceptive hormones. Human Pathol., 8: 67-81.135. Miller, F.R., D. McEachern, and B.E. Miller 1989 Growth regulation of mouse mammarytumor cells in collagen gel cultures by diffusible factors produced by normal mammarygland epithelium and stromal fibroblasts. Cancer Res., 49: 6091-6093.136.Mills, E.S., and Y.J. Topper 1970 Some ultrastructural effects of insulin, hydrocortisoneand prolactin on mammary gland explants. J. Cell Biol., 44: 310-328.137.Mumford, R.E. 1957 The effect of cortisol acetate on oestrone-induced mammary glandgrowth in immature ovariectomized albino mice. J. Endocrinol., 16: 72-79.138. Murphy, L.C., L.J. Murphy, D. Dubik, G.I. Bell, and R.P.C. Shiu 1988 Epidermal growthfactor gene expression in human breast cancer cells: regulation of expression byprogestins. Cancer Res., 48: 4555-4560.139.Musgrove, E.A., C.S.L. Lee, and R.L. Sutherland 1991 Progestins both stimulate andinhibit breast cancer cell cycle progression while increasing expression of transforminggrowth factor a, epidermal growth factor receptor, c-fos, and c-myc genes. Mol. Cell.Biol., 11: 5032-5043.140. Nandi, S. 1958 Endocrine control of mammary-gland development and function in theC3H/He Crgl mouse. J. Natl. Cancer Inst., 21: 1039-1063.141.Nandi, S. 1959 Hormonal control of mammogenesis and lactogenesis in the C3H/HeCrgl mouse. Univ. Calif. Berkeley Publ. Zool., 65: 100-128.142.Nicholson, S., J.R.C. Sainbury, G.K. Needham, P. Chambers, J.R. Farndon, and A.L.Harris 1991 Quantitative assays of epidermal growth factor receptor in human breastcancer: cut-off of clinical relevance. Int. J. Cancer, 42: 36-41.143. Nicholson, S., J. Richardson, J.R.C. Sainbury, P. Halcrow, P. Kelly, B. Angus, C. Wright,J. Henry, J.R. Farndon, and A.L. Harris 1988 Epidermal growth factor receptor (EGFR);results of a 6 year follow-up study inoperable breast cancer with emphasis on the nodenegative subgroup. Br. J. Cancer., 63: 146-150.144.Nickell, K.A., J. Harper, and H.L. Moses 1983 Transforming growth factors in solidhuman malignant neoplasms. Cancer Res., 43: 1966-1971.145.Norgren, A. 1968 Modifications of mammary development of rabbits injected withovarian hormones: effects of cortisone acetate, thyroid hormones and insulin afterhypophysectomy. Acta Univ. Lund. Sect. 2, 4: 6-41.113146.Oka, T., and Y.J. Topper 1972 Is prolactin mitogenic for mammary epithelium? Proc.Natl. Acad. Sci. USA, 69: 1693-1696.147.Oka, T., J.W. Perry, and Y.J. Topper 1974 Changes in insulin responsiveness duringdevelopment of mammary epithelium. J. Cell Biol., 62: 550-556.148.Oka, T., 0. Tsutsumi, H. Kurachi, and S. Okamoto 1988 The role of epidermal growthfactor in normal and neoplastic growth of mouse mammary epithelial cells. In: BreastCancer: Cellular and Molecular Biology. M.E. Lippman and R. Dickson, eds. KluwerAcademic Publishers, pp. 343-362.149.Okamoto, S., and T. Oka 1984 Evidence for the physiological function of epidermalgrowth factor: Pregestational sialoadenectomy of mice decreases milk production andincreases offspring mortality during lactation period. Proc. Natl. Acad. Sci. USA, 81:6059-6063.150. Pekonen, F., S. Partanen, T. Makinen, and E.M. Rutanen 1988 Receptors for epidermalgrowth factor and insulin-like growth factor 1 and their relation to steroid receptors inhuman breast cancer. Cancer Res., 48: 1343-1347.151.Petersen, 0.W., P.E. Hoyer, and B. van Deurs 1987 Frequency and distribution ofestrogen receptor-positive cells in normal, nonlactating human breast tissue. CancerRes., 47: 5748-5751.152.Purnell, D.M., and P. Kopen 1976 A quantitative histological analysis of the variation inlobulo-alveolar mitotic activity in the Lewis/Mai rat mammary gland during the estrouscycle. Anat. Rec., 186: 39-48.153.Richards, J., and S. Nandi 1978 Primary culture of rat mammary epithelial cells. I. Effectof plating density, hormones, and serum on DNA synthesis. J. Natl. Cancer Inst., 61:765-771.154.Richardson, F.L. 1955 The relative growth of the mammary glands in normal,ovariectomized and adrenalectomized female mice. Anat. Rec., 123: 279-289.155.Ruedl, C., V. Cappeletti, G. Coradini, and G. di Fronzo 1990 Influence of cultureconditions on the estrogenic cell growth stimulation of human breast cancer cells. Molec.Biol., 37: 195-200.156. Russo, J., and I.H. Russo 1987 Development of the human mammary gland. In: TheMammary Gland: Development, Regulation, and Function. M.G. Neville and C.W.Daniel, eds. Plenum Press, New York, pp. 67-93.157.Sakakura, T., and Y. Nishizuka 1967 Effect of thymectomy on mammary tumorigenesis,noduligenesis, and mammogenesis in the mouse. Gann, 58: 441-450.158.Salomon, D.S., L.A. Liotta, and W.R. Kidwell 1981 Differential responses to growthfactor by rat mammary epithelium plated on different collagen substrata in serum-freemedium. Proc. Natl. Acad. Sci. USA, 78: 382-386.114159.Salomon, D.S., JA. Zwiebel, M. Bano, I. Losoczy, P. Fehnel, W.R. Kidwell 1984Presence of transforming growth factors in human breast cancer cells. Cancer Res., 44:4069-4077.160.Sankaran, L., P. Qasba, and Y.J. Topper 1984 Effects of estrogen depletion on ratcasein gene expression. Biochem. Biophys. Res. Commun., 125: 682-689.161.Schlessinger, J. 1988 The epidermal growth factor receptor as a multifunctionalallosteric protein. Biochemistry, 27: 3119-3123.162.Schreiber, A.B., M.E. Winkler, and R. Derynck 1986 Transforming growth factor a: amore potent angiogenic mediator than epidermal growth factor. Science, 232: 1250-1253.163.Selye, H. 1940 Effect of chronic progesterone overdosage on the female accesory sexorgans of normal, ovariectomized and hypophysectomized rats. Anat. Rec., 78: 253-265.164. Shaikh, A.A. 1971 Estrone and estradiol levels in the ovarian venous blood from ratsduring the estrous cycle and pregnancy. Biol. Reprod., 5: 297-307.165.Shamay, A., N. Cohen, N. Mineo, and A. Gertler 1988 Effect of insulin-like growth factor1 on deoxyribonucleic acid synthesis and galactopoiesis in bovine undifferentiated andlactating mammary tissue in vitro. Endocrinology, 123: 804-809.166.Sheffield, L.G., C.F. Ayslworth, and C.W. Welsch 1987 Cyclic nucleotide and proteinphosphorylation in mouse mammary glands: effects of estrogen and progesteroneadministered in vivo. Proc. Soc. Exp. Biol. Med., 185: 283-290.167.Sheffield, L.G., and C.W. Welsch 1987 Influence of the submandibular salivary glandson hormone responsiveness of mouse mammary glands. Proc. Soc. Exp. Biol. Med.,186: 368-377.168.Sheth, N.A., S.S. Tikekar, K.J. Ranadive, and A.R. Seth 1978 Influence ofbromoergocryptine on estrogen-induced prolactin receptors of the mouse mammarygland. Mol. Cell. Endocrin., 12: 167-176.169.Shyamala, G., and A. Ferenczy 1984 Mammary fat pad may be a potential site forinitiation of estrogen action in normal mouse mammary glands. Endocrinology, 110:1078-1081.170.Sirbasku, D.A. 1978 Estrogen induction of growth factors specific for hormoneresponsive mammary, pituitary, and kidney cells. Proc, Natl. Acad. Sci. USA, 75: 3786-3790.171.Sirbasku, D.A., T. Ikeda, and D. Danielpour 1985 Endocrine and autocrineestromedins for mammary and pituitary tumor cells. In: Mediators in Cell Growth andDifferentiation. R.J. Ford and A.L. Maizel, eds., Raven Press, New York, pp. 213-232.115172.Slamon, D.J., G.M. Clark, S.G. Wong, W.J. Levin, A. Ullrich, and W.L. McGuire 1987Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science, 235: 177-182.173.Soule, H.D. and C.M. McGrath 1980 Estrogen responsive proliferation of clonal humanbreast carcinoma cells in athymic mice. Cancer Lett., 10: 177-189.174.Stampfer, M., R.C. Hallowes, and A.J. Hacket 1980 Growth of normal mammary cells inculture. In vitro, 16: 415-425.175.Stockdale, F.E., and Y.J. Topper 1966 The role of DNA synthesis and mitosis inhormone-dependent differentiation. Proc. Natl. Acad. Sci. USA, 56: 1283-1289.176. Stoker, M.G., D. Pigot, and J. Taylor-Papadimitriou, 1976 Response to epidermal growthfactors of cultured human mammary epithelial cells from benign tumors. Nature, 264:764-767.177.Spyratos, F., J-C. Delarue, C. Andrieu, R. Lidereau, M-H. Champene, K. Hacene, andM. Brunet 1990 Epidermal growth factor receptors and prognosis in primary breastcancer. Breast Cancer Res. Treat., 17: 83-89.178.Taketani Y., and T. Oka 1983 Biological action of epidermal growth factor and itsfunctional receptors in normal mammary epithelial cells. Proc. Natl. Acad. Sci. USA, 80:2647-2650.179.Taylor-Papadimitriou, J., M. Shearer, and M.G. Stoker 1977 Growth requirements ofhuman mammary epithelial cells in culture. Int. J. Cancer, 20: 903-908.180.Taylor-Papadimitriou, J., P. Purkis, and I.S. Fentiman 1980 Cholera toxin and analoguesof cyclic AMP stimulate the growth of cultured human mammary epithelial cells. J. Cell.Physiol., 102: 317-321.181.Thomas, D.B. 1982 Noncontraceptive exogenous estrogens and the risk of breastcancer: a review. Breast Cancer Res. Treat. 2: 203-211.182.Todaro, G.L., C. Fryling, and J.E. De Larco 1980 Transforming growth factors producedby certain human tumor cells: polypeptides that interact with epidermal growth factorreceptors. Proc. Natl. Acad. Sci. USA, 77: 5258-5262.183.Toi, M., Y. Hamada, T. Nakamura, H. Mukaida, S. Suehiro, T. Wada, T. Toge, M.Nimoto, and T. Hattori 1989 lmmunocytochemical and biochemical analysis of epidermalgrowth factor receptor expression in human breast cancer tissues: relationship toestrogen receptor and lymphatic invasion. Int. J. Cancer, 43: 220-225.184.Tonelli, Q.J., and S. Sorof 1980 Epidermal growth factor requirement for development ofthe cultured mammary gland. Nature, 285: 250-252.185.Topper, Y.J., and C.S. Freeman 1980 Multiple hormone interactions in thedevelopmental biology of the mammary gland. Physiol. Rev., 60: 1049-1106.116186.Traurig, H.H. and C.F. Morgan 1964 The effect of ovarian and hypophyseal hormoneson mammary gland epithelial cell proliferation. Anat. Rec., 150: 423-434.187.Trichopoulos, D., B. MacMahon, and P. Cole 1972 Menopause and breast cancer risk.J. Natl. Cancer Inst., 48: 605-613.188.Tucker, H.A., 1974 General endocrinological control of lactation. In: Lactation: AComprehensive Treatise, Vol. I. B.L. Larson and V.R. Smith, eds. Academic Press, NewYork, pp. 277-326.189.Valverius, E., S.E. Bates, D.S. Salomon, M. Stampfer, R. Clark, F. McCormick, M.E.Lippman, and R.B. Dickson 1987 Production of transforming growth factor alpha byhuman mammary epithelial cell and sub-lines. Proc. Endocrine Soc., 69: 97-109.190.van Bogaert, L.J., J. Quinones, and M.P. van Craynest 1980 Estriol-induced dose-dependent DNA synthesis in normal human mammary in vitro. Eur. J. Cell Biol. 21: 234-236.191.van Bogaert, L.J., M.P. van Craynest, and J. Abarca-Quinones 1982 Direct influence ofthe three natural estrogens on human mammary epithelium in vitro. Horm. Metab. Res.,14: 598-601.192.van der Burg, B., G.B. Rutteman, M.A. Blankenstein, S.W. de Laat, and E.J.J. vanZoelen 1988 Mitogenic stimulation of human breast cancer cells in a growth factor-defined medium: synergistic action of insulin and estrogen. J. Cell. Physiol., 134: 101-108.193.van der Burg, B., A.J.P. van Selm-Miltenburg, S.W. de Laat, and E.J.J. van Zoelen 1989Direct effects of estrogen on c-fos and c-myc protooncogene expression and cellularproliferation in human breast cancer cells. Mol. Cell. Endocrinol., 64: 223-228.194.van der Burg, B., L. IsbrOcker, A.J.P. van Selm-Miltenburg, S.W. de Laat, and E.J.J. vanZoelen 1990 Role of estrogen-induced insulin-like growth factors in the proliferation ofhuman breast cancer cells. Cancer Res., 50: 7770-7774.195.Vonderhaar, B.K. 1977 Studies on the mechanism by which thyroid hormones enhancealpha-lactalbumin activity in explants from mouse mammary gland. Endocrinology, 100:1423-1431.196.Vonderhaar, B.K., G.H. Smith, R.J. Pauley, J.M. Rosen, and Y.J. Topper 1978 Adifference between mammary epithelial cells from mature virgin and primiparous mice.Cancer Res., 38: 4059-4065.197.Vonderhaar, B.K. 1987 Local effects of epidermal growth factor, transforming growthfactor-alpha and EGF-like growth factor on lobuloalveolar development of the mousemammary gland in vivo. J. Cell. Physiol., 132: 581-584.198.Vonderhaar, B.K. 1987 Prolactin: Transport, function, and receptors in mammary glanddevelopment and differentiation. In: The Mammary Gland: Development, Regulation,117and Function. M.C. Neville and C.W. Daniel, eds. Plenum Press, New York, pp. 383-417.199.Vorherr, H., 1974 Development of the female breast. In: The Breast. H. Vorherr, ed.Academic Press, New York, pp. 1-18.200.Wallis, M. 1984 The molecular evolution of prolactin and related hormones. In: ProlactinSecretion: A Multidisciplinary Approach. F. Mena and C. Valverde-R., eds. AcademicPress, New York, pp. 1-16.201.Wang, D.Y., and V. Amor 1971 A study on the effect of insulin on DNA, RNA and proteinsynthesis in mouse mammary gland tissue in organ culture. J. Endocrinol., 50: 241-249.202.Wasner, G., I. Hennerman, and K. Kratochwil 1983 Ontogeny of mesenchymalandrogen receptors in the embryonic mouse mammary gland. Endocrinology, 113:1771-1778.203.Wiens, D., C.S. Park, and F.E. Stockdale 1987 Milk protein expression and ductalmorphogenesis in the mammary gland in vitro: hormone-dependent and -independentphases of adipocyte-mammary epithelial cell interaction. Dev. Biol., 120: 245-258.204.Williams, J.M., and C.W. Daniel 1983 Mammary ductal elongation: differentiation ofmyoepithelium and basal lamina during branching morphogenesis. Dev. Biol., 97: 214-219.205. Witliff, J.L. 1974 Specific receptors of the steroid hormones. Semin. Oncol., 1: 109-118.206.Wood, B.G., L.L. Washburn, A.S. Mukherjee, and M.R. Banerjee 1975 Hormonalregulation of lobuloalveolar growth, functional differentiation, and regression of wholemammary gland in organ culture. J. Endocrinol., 65: 1-6.207.Yang, J., R. Guzman, J. Richards, N. Imagawa, K. McCormick, and S. Nandi 1980Growth factor and cyclic nucleotide-induced proliferation of normal and malignantmammary epithelial cells in primary culture. Endocrinology, 107: 35-41.208.Yang, J., L. Larson, D. Flynn, J. Elias, and S. Nandi 1982 Serum-free primary culture ofhuman mammary epithelial cells in collagen gel matrix. Cell Biol Int. Reports, 6: 969-975.209.Yang, J., A. Balakrishnan, S. Hamamoto, C.W. Beattie, T.K. Gupta, S.R. Wellings, andS. Nandi 1986 Different mitogenic and phenotypic responses of human mammaryepithelial cells grown in two versus three dimensions. Exp. Cell. Res., 167: 563-569.210.Yarden, Y., and J. Schlessinger 1988 Growth factor receptor tyrosine kinases. Annu.Rev. Biochem. 57: 443-478.211.Zumoff, B., J. Fishman, H.L. Bradlow, and S. Hellman 1975 Hormone profiles inhormone dependant cancer. Cancer Res., 34: 3365-3373.118212.Zwiebel, J.A., M. Bano, E. Nexo, D.S. Salomon, and W.R. Kidwell 1986 Partialpurification of transforming growth factors from human milk. Cancer Res., 46: 933-939.119


Citation Scheme:


Usage Statistics

Country Views Downloads
China 23 1
United States 10 2
United Kingdom 4 0
Canada 2 0
Japan 1 0
India 1 0
City Views Downloads
Beijing 23 0
Unknown 3 2
Ashburn 2 0
Renton 2 0
London 2 0
Buffalo 2 0
Sunnyvale 2 0
Surrey 1 0
Redmond 1 0
Phoenix 1 0
Tokyo 1 0
Bangalore 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items

Admin Tools

To re-ingest this item use button below, on average re-ingesting will take 5 minutes per item.


To clear this item from the cache, please use the button below;

Clear Item cache