Search

Epithelial/mesenchymal interactions begin during embryonic development of the mammary gland and continue throughout mammary gland development into adult life. Stromal and epithelial growth factors that may mediate interactions between these compartments of the mammary gland are reviewed. Since mammogenic hormones are the primary regulators of mammary gland development, special consideration is given to hormonal regulation of growth factors in order to explore the integration of hormones and growth factors in the regulation of mammary gland growth and neoplasia. Examination of hormonal regulation of the fibroblast growth factor (FGF)-7/FGFR2-IIIb receptor system in the mammary gland reveals that mammogenic hormones differentially regulate the synthesis of stromal growth factors and their epithelial receptors. These effects serve to optimize the action of estrogen and progesterone on mammary gland development and illustrate that the ratio of these two hormones is critical in regulating this growth factor axis. The role of stromal/epithelial mitogenic microenvironments in modulating the genotype and phenotype of preneoplastic and neoplastic lesions by chemical carcinogens is discussed. Finally, changes in growth factor expression during mammary tumor progression are described to illustrate the relative roles that stromally-derived and epithelial-derived growth factors may play during progression to hormone independent tumor growth.

Mammary gland development is regulated by complex interactions among mammogenic hormones and locally derived paracrine growth factors. In epithelial tissues, keratinocyte growth factor (KGF or FGF-7) originates in the stroma while its receptor (KGFR or FGFR2-IIIb) is present only in the epithelium. Previous work showed that estrogen but not progesterone could stimulate the synthesis of KGF in mammary stroma in vivo. The effects of 17 beta-estradiol and progesterone on KGFR expression in vivo were examined in these studies. Peripubertal and mature virgin mice received subcutaneous injections of hormone in sesame oil after which KGFR mRNA levels were assayed by ribonuclease protection analysis of mammary gland RNA. Estradiol treatment caused a dose- and time-dependent decrease in KGFR mRNA level in mice from both age groups while stimulating ductal growth after 7 days of treatment. Inhibition of KGFR expression was near maximal at an estradiol dose of 2 microg after 1 day of treatment. Progesterone injection increased KGFR mRNA levels but this effect correlated with the stimulation of ductal growth. However, when progesterone was co-administered with estradiol, KGFR mRNA levels were maintained in the absence of any effect on ductal growth. Thus, estradiol inhibited KGFR mRNA only when elevated unopposed by progesterone. These data show that KGFR expression is determined by the ratio of estradiol and progesterone and suggests a mechanism through which these hormones can co-operate to optimize their growth-promoting effects. Consequences of hormone imbalance are also implicated.

Keratinocyte growth factor (KGF), alone and in synergism with progesterone (P) and prolactin (PRL), is mitogenic for normal mammary epithelium (ME) in vitro. In addition, P can upregulate ME sensitivity to KGF by slowing KGF receptor (KGFR) mRNA turnover in vitro. These hormonal interactions with KGF in vitro raise the possibility that alterations in these interactions can play a role in hormone-dependent mammary tumor growth and progression. The effect of hormones on KGF mitogenesis and the regulation of KGFR expression was examined in pregnancy-dependent (PDT) and ovarian-independent (OIT) mouse mammary tumors. In serum-free, collagen gel cell culture, dose/response (2‐20 ng/ml) and time course studies showed that KGF stimulated the proliferation of PDT (not OIT) cells but synergism with P or PRL was not observed. The level of KGFR mRNA in PDT cells was not significantly different from normal ME but in OIT it was reduced more than 90%. P did not affect KGFR mRNA turnover in cultured PDT cells. However, KGFR mRNA was more stable in PDT cells compared to normal ME; after 6 days culture in basal medium, KGFR mRNA levels declined 40% vs. 85% previously shown for normal ME. Determination of KGF mRNA levels in tissues showed that it was lower in PDT compared to normal mammary gland and not detectable in OIT. These data show that in PDT both KGF-stimulated mitogenesis and the regulation of KGFR expression are independent of hormones. OIT has progressed to independence from any KGF influence. Thus, a subset of hormonally regulated pathways related to epithelial/stromal cell interactions can be lost in hormone-dependent mammary tumors during tumor progression. J. Cell. Physiol. 184:222‐228, 2000. © 2000 Wiley-Liss, Inc.

The expression of the KGF receptor (KGFR) and its stromal ligands, KGF and FGF-10, was compared during mouse mammary gland development. KGFR expression in mammary parenchyma is maximal in mature virgin mice, declines during pregnancy and lactation, but rises after weaning. The rise in KGFR mRNA in the virgin animal corresponds to parenchymal growth. The fall in KGFR expression in pregnancy is driven by hormone-induced alveolar differentiation since the level of KGFR mRNA is 5-fold higher in isolated ductal cells compared to alveolar cells. KGF and FGF-10 expression patterns differ during ductal development. FGF-10 is also expressed at about a 15-fold higher molar level than KGF. During pregnancy and lactation, expression of KGF and FGF-10 decreases in intact fat pads but is unchanged in parenchyma-free fat pads. Thus, the decrease in KGF and FGF-10 expression observed in intact glands during pregnancy and lactation is not a direct consequence of the changing hormonal milieu but more likely reflects an increase in the ratio of epithelium to stroma. Differences in the level and pattern of expression of mRNA for KGF, FGF-10, and the KGFR during postnatal development of the mouse mammary gland are a result of morphological development, changes in the ratio of stroma to epithelium, and hormonal regulation of cell differentiation. These changes suggest that the biological roles that these growth factors play are regulated by fluctuations in both growth factor and growth factor receptor expression and that KGF and FGF-10 may have different regulatory functions.

Stromally derived keratinocyte growth factor (KGF) can play an important role in mammary gland development as a mesenchymal/stromal mediator of epithelial growth and morphogenesis. However, the possible coordinate regulation of mammary gland development by mammogenic hormones and KGF is unexplored. In these studies, the direct effect of mammogenic hormones on KGF-mediated mammary epithelial mitogenesis and expression of the KGF receptor was examined using primary mouse mammary epithelium growing in serum-free, collagen gel cell culture. Addition of KGF produced an average 7-fold increase in cell number after 10 days of culture. This effect of KGF was further increased in the presence of PRL (9-fold) or progesterone (P; 15-fold), with the combination of P and PRL (22-fold) producing the strongest synergistic stimulation. Estrogen did not show any additional stimulation of growth either alone or in combination with PRL and/or P. Ribonuclease protection analysis showed that epithelial cells grown in medium s...

Polyunsaturated fatty acids enhance the proliferation of mouse mammary epithelial cells stimulated by epidermal growth factor (EGF) by modulating the post-receptor signaling pathways. The growth stimulatory effect of these fatty acids is completely inhibited by pertussis toxin, whereas the inhibition of EGF and insulin stimulated growth is only partial. The treatment of cell cultures with 12-O-tetradecanoyl-phorbol-13 acetate (TPA) reverses the growth inhibitory effect of pertussis toxin and fully restores the growth as was in the control cultures untreated with the toxin suggesting a role for PKC in this reversal. It appears that the functions of Gi-proteins are required in the mediation of fatty acid effect on growth. The predominant types of Giα in mammary epithelial cells are Giα 1, Giα 2 and Giα 3. Among these, the levels of Giα 1 and 2 appears to be regulated by steroid hormones. Linoleic acid raises the level of GTP-bound Ras in the cells above the levels induced by EGF. Pertussis toxin reduces the level of Ras-GTP and inhibits phosphorylation of MAP kinase by EGF. It has been speculated that Gi-proteins interact with the receptor bound nucleotide exchange factor and the membrane anchored Raf kinase and constitute two sites for pertussis toxin action. The phosphorylation by PKC may uncouple Gi-protein interaction with these effectors and enable the agonist-induced signals to bypass the inhibitory action of PT on growth. It appears that the role of growth stimulatory fatty acids is to sustain a high level of Ras-GTP and amplify the signal passing through Raf-MAP kinase pathway in a Gi-protein dependent manner.

The effect of lysophosphatidic acid (LPA) on the proliferation of normal and tumor mouse mammary epithelial cells in primary, serum-free, collagen gel cell culture was evaluated. LPA stimulated the growth of normal mammary epithelial cells from mature virgin mice. The growth of pregnancy-dependent tumors (PDT) was generally stimulated, although the response was attenuated in some of these tumors compared to normal cells. In contrast, the growth of 70% of ovarian-independent tumors (OIT) was inhibited by LPA; the remainder were unaffected. LPA stimulated cAMP accumulation and phosphoinositide (PI) hydrolysis in normal, PDT, and OIT. Thus, the regulation of adenylyl cyclase and PI-specific phospholipase C by LPA is similar in normal and tumor cells. Pertussis toxin (PT) partially inhibited LPA-stimulated growth in normal cells but did not affect LPA-stimulated PI hydrolysis or cAMP accumulation. Thus, PT-sensitive and -insensitive proliferative pathways are activated. PT also inhibited LPA-stimulated growth of PDT but generally had no effect on the growth of OIT. These results show that the mitogenic response to LPA is attenuated in the hormone-dependent phenotype and switches to growth inhibition in hormone-independent tumors. Furthermore, LPA stimulates multiple signal transduction pathways mediated by PT-sensitive and -insensitive G proteins. The PT-sensitive pathways are not tightly coupled to the proliferative response to LPA in tumor cells. These data suggest that alterations in G protein function may occur during tumor progression.

The growth, morphogenesis and differentiation of milk producing epithelial tissues in the developing mammary glands require interaction with extracellular matrices and stimulation by hormones, growth factors and essential fatty acids. In primary culture, the proliferation of mammary epithelial cells (MEC), induced by epidermal growth factor (EGF), is enhanced and sustained by linoleate and its eicosanoid metabolites. Since a combination of linoleic acid (18:2ω6) and prostaglandin E 2 or cAMP has synergistic effect on EGF-stimulated growth, it is suggested that additional cAMP-dependent protein kinase A (PK-A) independent pathways may also contribute to the linoleate effect on EGF action. Possible involvement of Ca2+phospholipid-dependent protein kinase C (PK-C) is explored. Both linoleate and arachidonate can activate Type-II and Type-III protein kinase-C in MEC and a PK-C inhibitor can block growth stimulation by EGF and fatty acids. Like 12-O-Tetradecanoly phorbol-13-acetate (TPA), a PK-C activator which also enhances EGF-stimulated growth of MEC, linoleate can phosphorylate a 40–42 KD protein. EGF itself can stimulate transient phosphorylation of the same protein in MEC cultures but when supplemented with linoleate, which does not influence the ligand binding affinity of EGF-receptors, the transient phosphorylation signal in 40–42 KD protein is sustained. Based on our results and other published reports, it is proposed that although both EGF and essential fatty acid (EFA) evoke multiple pathways in cells, serine/threonine kinase activity of PK-C and PK-A and tyrosine kinase activity of EGF-receptors may converge in the phosphorylation of 40–42 KD type of protein which could be a downstream messenger of EGF and EFA action.

I. Introduction THE mammary gland is a unique feature of mammals. Its chief biological function is to synthesizeand secrete milk, which is essential for the nourishment of mammalian young. The parenchymal portion of the mammary gland is derived embryologically from ectoderm, while the surrounding stroma comes from mesoderm. Mammary gland growth takes place during both fetal and postnatal life; the gland is the site of milk production and secretion, and in several species (including humans) it is also a site of occurrence of preneoplastic and neoplastic lesions. Thus the mammary gland is an important organ of study by biologists from various disciplines, including endocrinologists, tumor biologists, and developmental biologists. During the last 30 yr, the primary interest of our laboratory has been the understanding of the neoplastic transformation process in mammary epithelial cells. One fundamental characteristic of tumor cells vis a vis their normal counterparts is deregulation of growth. Since cell pro...

Proliferation and differentiation of mammary epithelia are regulated by the combined action of systemic hormones and locally derived paracrine growth factors. Keratinocyte growth factor (KGF) is a potential candidate stromal factor that may participate in the hormonal control of stromal/epithelial interactions. In this study, we have examined the in vivo effect of 17beta-estradiol (E) treatment on KGF expression in mammary glands of peripubertal (5-week-old) and mature (11-week-old) mice. Mice received subcutaneous injections of hormone after which KGF mRNA levels were assayed by ribonuclease protection analysis of mammary gland RNA. E treatment caused a dose- and time-dependent increase in KGF mRNA levels in intact mice from both age groups. Neither 17alpha-estradiol nor progesterone injection affected KGF mRNA levels. Comparison of the relative expression of KGF in parenchyma-free fat pads and in intact glands demonstrated that the basal and E-dependent KGF mRNA levels did not require the presence of mammary epithelium. ELISA assay of KGF tissue content demonstrated that concomitantly with an up-regulation of mRNA, E treatment also increased KGF protein in mammary glands from peripubertal and mature mice. These data show that E treatment stimulates both KGF mRNA and protein expression in mammary stroma in vivo and raises the possibility that KGF has a role in E-regulated mammary gland development.

A method for culturing mouse mammary epithelial cells (MECs) within Type I collagen gels is described. This method is permissive for multifold growth, “ductlike” morphogenesis, and hormonal responsiveness. Subcutaneous mouse mammary glands are harvested, digested with collagenase, and epithelial cells obtained after Percoll gradient centrifugation. MECs are then mixed with neutralized, isosmotic collagen, which is pipetted into culture dishes and overlaid with culture medium after gel formation. Preparation of all serum-free medium and stocks required for the procedure and termination of cultures are described, and comments concerning critical procedural steps are given.

A method is described for the chemical carcinogen induced transformation of mouse mammary epithelial cells cultured in three-dimensional collagen gels with defined mitogens. Transformation is assayed in this system by the transplantation of the carcinogen treated cells to the parenchyma free mammary fat pads of syngeneic female mice. Resultant mammary outgrowths are scored as to the development of morphologically distinct preneoplastic and neoplastic lesions. The phenotype and the nature of genetic lesions is greatly dependent on the mitogen used at the time of carcinogen exposure.

An elevation in total MAP kinase activity and expression has been observed in breast cancer tissue. However, the mechanisms underlying these changes in kinase activity and regulation by growth factors are not well characterized. In these studies, the effect of the potent mammary mitogen, epidermal growth factor (EGF), on the activation of the mitogen-activated protein kinases, ERK1 and ERK2 (extracellular regulated protein kinases 1 and 2, respectively), was compared in primary cultures of normal mouse mammary epithelial cells and in a hormone-responsive mouse mammary tumor. In normal epithelium, EGF stimulated an early rise in ERK activity at 4 min followed by a rapid decline, whereas a sustained (1 h) elevation of ERK activity was observed in the tumor cells. The time course of ERK activity in both cell types coincided with the phosphorylation state of the EGF receptor, suggesting that altered regulation of EGF receptor phosphorylation or EGF receptor turnover produces an enhanced ERK response to EGF in tumor cells. The MEK inhibitor, PD 098059 inhibited EGF-stimulated proliferation and ERK activity in a parallel, dose-dependent manner showing that ERK activation is at least permissive for the proliferative response to EGF. In addition, tumor cells showed a 4-fold elevation in basal (or ligand-independent) activity over normal cells without an increase in total enzyme level, and a preferential activation of ERK1 by EGF. These EGF-dependent and -independent changes in ERK regulation in the hormone-responsive mammary tumor underscore how multiple alterations in the regulation of this pathway may play a role in mammary tumorigenesis.

In this report, we have investigated placental lactogens (placental lactogen-I, PL-I; PL-I variant, PL-Iv; PL-II) expressed by differentiated Rcho-1 trophoblast cells. A complementary DNA (cDNA) library to differentiated Rcho-1 trophoblast cells was constructed and screened with probes to detect PL-I and PL-II. Sequence analysis of three independent Rcho-1 PL-I cDNAs indicated that they significantly differed from the previously reported PL-I sequence but more closely resembled a related cDNA referred to as PL-I mosaic (PL-Im). Upon further analysis, Rcho-1 PL-I/PL-Im transcripts could be detected in Rcho-1 trophoblast cells and normal developing placental tissue; however, the previously reported PL-I transcript could not be identified from the same sources. Given these results, we examined the original PL-I cDNA by PCR and nucleotide sequence analyses. The sequence differed from the original report and was found to be identical to the Rcho-1 PL-I and PL-Im cDNA clones. Thus, PL-I, Rcho-1 PL-I, and PL-Im are equivalent and should be referred to as PL-I. The PL-I gene was localized to chromosome 17 of the rat genome, similar to other PRL family members. Rcho-1 PL-II cDNAs were identical to the published PL-II sequence. PL-Iv cDNAs were isolated from differentiated Rcho-1 cells via an RT-PCR strategy and found to be identical to previously isolated PL-Iv cDNAs. Rcho-1 PL-I and PL-II cDNAs were subcloned into the pcDNA3 expression vector and recombinant protein produced in HRP-1 cells. Both recombinant Rcho-1 PL-I and PL-II proteins significantly stimulated the proliferation of lactogen-dependent rat Nb2 lymphoma cells and mouse mammary epithelial cells. In summary, we show that the Rcho-1 PL-I corresponds to PL-Im and Rcho-1 PL-Iv and PL-II are identical to their previously described placental counterparts. Additionally, both recombinant Rcho-1 PL-I and PL-II proteins are biologically active.

We have shown previously that phosphatidic acid esterified to polyunsaturated fatty acids is mitogenic for primary cultures of mouse mammary epithelial cells embedded within collagen gels. We hypothesized that this mitogenic competence resulted from the ability of this phospholipid to activate multiple signal transduction pathways in mammary epithelium. A closer examination of this hypothesis was undertaken by examining the effect of exogenous phosphatidic acid on phosphoinositide (PI) hydrolysis and its intracellular metabolism to diglyceride, an activator of protein kinase C. For assays of phosphoinositide-specific phospholipase C activation, mammary epithelial cells from virgin Balb/c mice were isolated by collagenase dissociation of mammary glands and cultured on the surface of Type I collagen-coated culture dishes. Phosphatidic acid (PA) stimulated a sustained increase in inositol phosphates and caused inositol phospholipid depletion when added to cells in which inositol phospholipids were prelabeled with 3H-myoinositol. This effect was specific for PA among phospholipids tested. Neither lineoleic acid, that can be released from PA, nor prostaglandin E2 affected PI hydrolysis. When mammary epithelial cells were cultured inside collagen gels in the presence of exogenous PA or phosphatidylcholine (PC) radiolabeled with 3H-glycerol, PA was found to persist intracellularly and be dephosphorylated to diglyceride (an activator of protein kinase C) to a greater extent than PC, a nonmitogenic phospholipid. In contrast to PA, epidermal growth factor (EGF) only slightly stimulated PI hydrolysis, showing that these two different growth-promoting factors do not actively couple to the same signal transduction pathways in mammary epithelial cells. These results show that PA may activate multiple pathways in mammary epithelial cells either directly or via its metabolism to diglyceride. © 1995 Wiley-Liss, Inc.

The direct effect of omega-3 and omega-6 fatty acids on the proliferation of mouse mammary tumor cells (MTC) was examined in a serum-free cell culture system. While the EGF-induced proliferation of normal mammary epithelial cells was shown to be enhanced by omega-3 and omega-6 fatty acids and prostaglandins (PGs), a majority (75-80%) of primary mammary tumors were not stimulated by these agents. Compared to normal cells, some MTC cultures showed a higher susceptibility to inhibition by omega-3 fatty acids. The general lack of response of MTC cultures to PGE2 and cyclic adenosine monophosphate (cAMP) suggests some alterations in the cAMP-mediated pathway. However, the PGE2-induced cAMP levels and cAMP-dependent protein kinase (PKA) activities in the tumor cells were comparable to normal cells. We conclude that the proliferation of mammary tumor cells either follow a cAMP-PKA-independent pathway or have some alterations in the serine/threonine kinase mediated signaling pathway.

Mammary epithelial cells derived from the entire mammary parenchyma or only end buds were isolated by collagenase digestion of mammary glands from virgin mice. Cells were cultured within collagen gels in serum-free medium containing insulin. Keratinocyte growth factor (KGF or FGF-7) and acidic fibroblast growth factor (aFGF or FGF-1) stimulated multifold proliferation when added alone to this medium. Growth occurred as three-dimensional colonies within the collagen gel matrix. KGF stimulated growth was unaffected by adding heparin. Conversely, multifold growth stimulation by acidic FGF required heparin. Since end buds are the actively proliferating cell population of ductal glands, organ cultures of these structures were prepared. KGF stimulated 3H-thymidine incorporation in these end buds in the absence and presence of epidermal growth factor. These data suggest that acidic FGF and KGF may represent in vivo stromal factors capable of regulating mammary gland development.

Mammary epithelial cells obtained from virgin mice were cultured in collagen gel with linoleic acid-containing serum-free growth medium supplemented with hormonal (PRL and progesterone, epidermal growth factor, somatomedin-C) or nonhormonal (lithium ion, phosphatidic acid containing phospholipid liposomes) growth stimulating agents. The phenotypes of the resulting progeny cells were compared by examining the ultrastructure, immunohistochemical staining for luminal epithelial and myoepithelial cells and casein, and assessing the quantity of biochemically detectable alpha- and beta-casein. Although there are some differences in ultrastructure and immunostaining in the progeny cell populations induced by different growth-promoting agents, all the cultures were able to accumulate alpha- and beta-casein on subsequent stimulation by PRL and linoleic acid in the second phase of culture. Since, in vivo, luminal epithelial cells of the mammary gland are the only cell type capable of synthesizing milk products, these results indicate that all the different growth stimulants, hormonal and nonhormonal, result in the predominant proliferation of luminal-type epithelial cells. These results have important implications for studies of the mechanism of growth control in and transformation of mammary epithelial cells.

Epithelial cells obtained by collagenase digestion of mammary glands from virgin BALB/c mice were cultured in collagen gels in serum-free basal medium containing insulin (10 micrograms/ml), to which lipids or growth factors were added. Synthetic phospholipids were added as liposomes. Dilinoleoyl phosphatidic acid or phosphatidylserine or epidermal growth factor stimulated multifold growth. The optimum mitogenic effect of the phospholipids was dependent upon the presence of a polyunsaturated fatty acid esterified to the sn-2 position of the glycerol moiety. Dilinoleoyl phosphatidylcholine also stimulated growth but was generally less stimulatory than phosphatidylserine or phosphatidic acid, and phosphatidylethanolamine did not stimulate growth. Studies using phospholipids radiolabeled in either the sn-2 fatty acyl group or the glycerol backbone showed that the relative effect of phospholipids on growth did not correlate directly with the extent of their incorporation into cellular lipid, indicating that phospholipid turnover was the more important determinant for mitogenesis. Analysis of phosphatidic acid-stimulated growth suggested that both cAMP-dependent and cAMP-independent pathways were involved. Thus, mitogenic phospholipids stimulate proliferation by activating (directly or indirectly) multiple growth-regulatory pathways in mammary epithelial cells.

The results obtained to date from studies dealing with the role of hormones, including estrogen, on growth of mammary epithelial cells inside the collagen gel are described. The collagen gel matrix culture system appears to be a suitable system to obtain in vivo-like effects of hormones on mammary cell in vitro. The results thus far indicate that prolactin along with progesterone or cortisol can stimulate mammary cell proliferation. Thus far, estrogen has not been found to be mitogenic in our in vitro system.

Normal mammary epithelial cells from BALB/cfC3H midpregnant mice were freed from stromal cell types by Percoll density gradient centrifugation after collagenase digestion and were then embedded within collagen gels. Sustained growth leading to an increase in cell number was accomplished in response to cholera toxin and high concentrations of horse serum. The extent of growth was found to be dependent on the horse serum concentration, the maximum growth being attained at 50%. A serum concentration of 12.5% horse serum and 2.5% fetal calf serum, along with cholera toxin at 0.01 μg/ml, allowed maintenance but failed to cause any significant increase in cell number during the experimental period of 2 weeks. This same maintenance medium was used to determine the effects of various exogenously added steroids, protein hormones, and organ extracts on the proliferation of mammary epithelial cells in culture. Hormones failed to elicit any proliferative response, but extracts of kidney, brain, uterus, and spleen produced proliferative responses equal to or greater than the response obtained with 50% horse serum and cholera toxin. Kidney extracts prepared from midpregnant mice, virgin mice, and virgin mice given pituitary isografts all showed comparable activities, suggesting that the concentration of stimulatory factor(s) was not influenced by the hormonal status of the donor. Normal mammary epithelial cells that had undergone a 10- to 15-fold increase in cell number over initial values during 2-3 weeks in culture were passaged to secondary gel cultures. Outgrowth similar to those seen in primary culture were seen again in secondary culture. The present system provides a method for sustaining growth in culture of primary mammary epithelial cells from normal tissues.

We investigated the effect of somatomedin C (SM-C) on the growth of mouse mammary ductal epithelial cells in collagen gel culture. Epithelial cells, isolated by collagenase digestion of whole glands, were placed into primary serum-free collagen gel cell culture for 10-12 days, during which SM-C was added alone or in combination with other growth-promoting factors. Previous work has shown that these cells require a superphysiological concentration of insulin (10 micrograms/ml) for optimum growth in serum-free medium (a 1:1 mixture of Ham's F-12 and Dulbecco's Modified Eagle's medium) containing epidermal growth factor (EGF). When SM-C (1-250 ng/ml) alone was added to serum-free basal medium containing EGF, it stimulated growth (at concentrations greater than 25 ng/ml) to at least the same extent as insulin at 10 micrograms/ml. There was no additive stimulation of growth when optimal concentrations of insulin and SM-C were added together. The nonadditive stimulation at optimal concentrations of these hormones may indicate that the previous requirement for a superphysiological concentration of insulin for maximum growth was due to low affinity binding of insulin to the SM-C receptor. Rat insulin-like growth factor II (Collaborative Research) at 50-200 ng/ml did not stimulate growth in the presence or absence of insulin. SM-C could not stimulate growth alone. The presence of EGF or mammogenic hormones (progesterone and PRL) was required.

Normal and neoplastic mammary cells from both human and rodent sources grow in culture in response to a number of hormones and growth factors. However, with the exception of a few human tumor lines, a consistent growth-promoting effect of estrogens on mammary cells has not been observed. Mammary cells can be shown to respond to other hormones and factors, such as PRL, hydrocortisone, and epidermal growth factor. Recent observations suggest that the pH indicator dye phenol red, found in most media, may be masking any exogenous estrogenic effects by acting as a weak estrogen. To test this possibility, we reexamined the effects of estradiol (E2) and the antiestrogen keoxifene on the growth of normal human, mouse, and rat mammary cells in the absence of phenol red. Primary cultures of these mammary cells were grown within a rat tail collagen gel matrix in a serum-free medium made up of Ham's F-12 and Dulbecco's Modified Eagles' medium (1:1) with and without phenol red. The medium was supplemented with various hormones and growth factors. These supplements were selected for each cell type to produce a variety of conditions from nongrowing to rapidly growing. The effects of E2 (10(-10)-10(-8) M, keoxifene (10(-9)-10(-6) M), and phenol red on growth under these various conditions were examined. Phenol red had no effect on growth, and its absence did not restore a response to E2. Keoxifene, in the presence or absence of E2, also had no effect on growth. Although E2 had no effect on growth, it was able to induce a 150% increase in the progesterone receptor levels in normal mouse mammary cells in culture, indicating that the cells retain their capacity to respond to E2. This work supports the idea that the effect of estrogens on growth in vivo may be mediated through some other factor(s).

The effects of basic fibroblast growth factor (bFGF) on the growth and differentiation of mouse mammary epithelial cells in serum-free collagen gel culture were examined. Epithelial cells obtained from virgin or midpregnant mice grew when bFGF was added to medium containing either insulin at a concentration greater than or equal to 1 ftg/ml or somatomedin-C (Sm-C) at 150 ng/ml. This growth-promoting effect is of the same magnitude as, and additive with, the growth-promoting effect of epidermal growth factor (EGF) or mammogenic hormones. The sensitivity of the cells to EGF or mammogenic hormones was not altered by exposure to bFGF. The progeny cells resulting from growth stimulation by bFGF are capable of accumulating casein upon subsequent stimulation by prolactin (PRL), but accumulate less casein than cells grown in response to EGF. bFGF also appears to reduce casein accumulation if it is added to the cultures at the same time as PRL.

Kinetics of (45)Ca efflux and insulin release were studied in collagenase-isolated rat islets during 2-h perifusions with calcium-depleted (0.05 mM) bicarbonate-phosphate buffer containing 2.2 mM glucose. Addition of glucose (16.7 mM) suppressed (45)Ca efflux by 30%. Removal of glucose caused an "off response" of insulin release. The perifusion of a normal concentration of Ca (2.3 mM) greatly stimulated (45)Ca efflux, indicating Ca (45)Ca exchange. When Ca and glucose were superimposed, the effects on (45)Ca efflux and insulin release depended upon the order of presentation of the stimuli: when Ca was added to an ongoing 16.7-mM glucose perifusion, biphasic patterns of (45)Ca and insulin release were seen; when glucose was superimposed on a Ca perifusion, an inhibition of the Ca-stimulated (45)Ca efflux occurred, and a reduced but clearly biphasic insulin response was seen. The subsequent insulin off response after with-drawal of the glucose was also reduced. Mathematical "peeling" of (45)Ca efflux curves from unstimulated islets suggests that there are at least two, and probably three, different intracellular Ca compartments (not including the extracellular sucrose space). At the beginning of perifusion, these three compartments (I, II, III) contain 25, 56, and 19% of the intracellular (45)Ca, and their rates of efflux are 6.7, 1.2, and 0.1%/min, respectively. Glucose appears to suppress efflux from the largest compartment (II); Ca appears to exchange with (45)Ca from a more inert compartment (III). The relationship between insulin and (45)Ca release is not stoichiometric.

Mammary epithelial cells from adult virgin mice have been cultured within collagen gels in totally serum-free medium containing either epidermal growth factor or the mammogenic hormones, progesterone and prolactin, or prolactin alone. The cellular organization, differentiation and cell-type composition of the colonies from the three culture conditions were assessed by transmission electron microscopy and light-microscope immunocytochemistry. The epithelial cells form branching duct-like structures and, when exposed to mammogenic hormones, assume a secretory morphology (including casein micelles) similar to that seen in the early to mid-pregnant mouse.

Mammary glands are enzymatically dissociated and the resulting tissue digest enriched for epithelial cells by isopycnic banding on a density gradient of Percoll. The cells are embedded within a rat tail collagen gel matrix and fed with the appropriate medium. Growth and differentiation are superior in such a system when compared to culture on plastic, using identical media.

Sustained growth of normal mouse mammary epithelial cells in primary culture, leading to an increase in cell number, in response to growth factors [epidermal growth factor (EGF) and fibroblast growth factor (FGF)] or cholera toxin has been achieved by embedding the cells inside collagen cells. Inclusion of agents known to increase the level of cellular cAMP have been found to be favorable for mammary epithelial cell proliferation. Cholera toxin is by far the best of all of the agents tested (prostaglandins E1 and E2, isoproterenol, theophylline, and dibutyryl cAMP). When growth factors (EGF or FGF) are added with cholera toxin, a synergistic effect resulting in a response much greater than with either of them alone is seen. This synergism was best seen in normal mammary epithelial cells from nonpregnant mice. The extent of this synergistic effect was found to be less in normal cells from pregnant mice, suggesting that these cells may be less responsive to EGF during pregnancy. Tumor cells were found to be rather inconsistent in their responses to EGF and cholera toxin, ranging from a minimal response, similar to that of normal cells from pregnant animals, to a maximal response, similar to that of normal cells from nonpregnant animals.

Linoleate metabolism via the cyclooxygenase pathway enhances the proliferation of mammary epithelial cells in serum-free culture in the presence of epidermal growth factor and insulin (Bandyopadhyay, G.K., Imagawa, W., Wallace, D., and Nandi, S. (1987) J. Biol. Chem. 262, 2750-2756). Prostaglandin E2 (PGE2) can fully substitute for linoleic acid provided endogenous hydroxyeicosatetraenoic acids (HETEs, lipoxygenase metabolites) are available. The PGE2 effect is partial if lipoxygenase activity is inhibited by nordihydroguaiaretic acid. Any combination of two HETEs out of three tested (5-, 12-, and 15-HETEs) stimulates growth synergistically with PGE2; and together (i.e. PGE2 + HETEs), they completely substitute for linoleate. In the absence of PGE2, maximal stimulation cannot be attained with HETEs. Exogenous 5-HETE, compared with 12- or 15-HETE, is preferentially incorporated by the mammary epithelial cells, and about 25-30% of it is retained esterified in phospholipids. The cellular level of nonesterified, free HETE is low. Radioimmunoassay revealed that the concentrations of 12- and 15-HETEs in the culture media (with or without added linoleate) were always higher than that of 5-HETE. Both intra- and extracellular free HETEs are rapidly metabolized by the cells. Since these cells are capable of producing eicosanoids from linoleate, periodic supplementation of the cultures with linoleate allows maintenance of higher HETE and PGE2 levels. Thus, it appears that not only are HETEs short-lived in the cell cultures, but cells handle 5-HETE differently than 12- and 15-HETEs. Whatever may be the pathways of interaction, synergism between HETEs and PGE2 seems to explain how linoleate stimulates the growth of mammary epithelial cells in the presence of epidermal growth factor and insulin.

Metabolism of 1-14C-[18:3(n-3)] and 1-14C-[22:6(n-3)] were investigated in the primary cultures of normal and tumor mouse mammary epithelial cells. Analysis of endogenous fatty acid composition indicated a decreased proportion of total (n-6) PUFA in the cultured tumor cells compared to normal cells. These cells can synthesize significant amount of 20:5 (n-3) and 22:5 (n-3) but not 22:6 (n-3), from 18:3 (n-3). There was very little or no retroconversion of 22:6 (n-3) by these cells. It has been concluded that mammary epithelial cells may be deficient in 4-desaturase activity and also that exogenous 22:6 (n-3), instead of serving as a source of 20:5 (n-3), may actually counter the effects of both 20:4 (n-6) and 20:5 (n-3) in the mammary tissue.

Freshly isolated normal and tumor mouse mammary epithelial cells embedded within a collagen gel matrix undergo sustained growth when cultured for as long as 3 wk in a serum-free medium composed of a 1:1 (vol/vol) mixture of Hepesbuffered Ham's F12 and Dulbecco's modified Eagle's medium supplemented with insulin, epidermal growth factor (EGF), transferrin, bovine serum albumin fraction V, and cholera toxin. Of these additives, only insulin, EGF, and albumin are required for the growth of most normal cells. Albumin is not always an absolute requirement for growth but greatly enhances it. Lithium has been found to stimulate the growth of normal cells and can replace EGF. The collagen matrix culture system allows sustained growth of primary cultures of both normal and neoplastic mammary epithelium in serum-free conditions. This serum-free system will be useful in identifying and investigating the role of hormones, growth factors, and nutritional factors in regulating the growth of mammary epithelial cells.

Net 45Ca uptake (in excess of the extracellular 3H-sucrose space) and insulin release were measured under low (2 mM) and high (20 mM) glucose conditions in collagenase-isolated rat islets. 45Ca uptake curves were mathematically fitted and subjected to compartmental analysis. Within the first 60 seconds after addition of trace 45Ca, islets showed a similar rapid uptake of 45Ca regardless of the glucose concentration or length of time of prior exposure to glucose. Net 45Ca uptake continued to increase for 30-60 minutes, and the islets in high glucose showed approximately twofold greater maximum uptake than islets in low glucose. Islets preincubated in low glucose and then incubated in high glucose showed a 5-15-minute delay in net 45Ca uptake as compared with islets that had been preincubated in high glucose. Insulin release was detectable by 10 minutes of incubation with high glucose. Mathematical modeling of the low- and high-glucose net 45Ca uptake curves suggests that there are at least two calcium “compartments” within the β-cell. One compartment is small, rapidly filled, and insensitive to glucose, while the other, larger compartment, is slowly filled and fills to a much greater extent in the presence of high glucose. A major proportion of the glucose-stimulated uptake is at the level of influx.

A serum-free primary cell culture system was used to examine the direct effects and interactions of mammogenic hormones and epidermal growth factor (EGF) on the growth of mouse mammary epithelial cells. Epithelial cells were isolated by collagenase dissociation followed by Percoll gradient centrifugation and cultured within collagen gels in a mixture of Ham's F-12-Dulbecco's Minimum Essential Medium (1:1) containing insulin (10 micrograms/ml), crude soybean lecithin, trace elements, trypsin inhibitor, and antioxidants. Progesterone (P; 10(-6) - 10(-8) M) or ovine PRL (1 microgram/ml), in the absence of EGF, stimulated the growth of cells from mature virgin mice 2- to 4-fold over that of controls cultured in basal medium only. P and PRL synergized in stimulating growth 3- to 17-fold. 17 beta-Estradiol (10(-7) - 10(-10) M) alone did not stimulate growth or synergize with P and/or PRL. This lack of growth stimulation by 17 beta-estradiol was also observed in medium containing a low concentration of insulin (0.1 microgram/ml). EGF (10 ng/ml) alone stimulated growth to the same extent as the combination of P and PRL. EGF at 1, but not 10, ng/ml when combined with P and PRL could additively stimulate growth. Cells from midpregnant mice were less responsive than cells from virgin mice to the growth-stimulating effects of the combination of P and PRL (2-fold stimulation at most), but not to EGF (3- to 6-fold stimulation). Corticosterone, deoxycorticosterone, and aldosterone, but not cortisol, could synergize with PRL in stimulating the growth of cells from mature virgin mice. However, only deoxycorticosterone could stimulate growth in the absence of PRL. These results suggest that PRL, P, and adrenal corticoids may directly stimulate the growth of mouse mammary epithelial cells. The physiologically relevent adrenal corticoids, corticosterone and aldosterone, only potentiate the stimulatory effect of PRL. The hormonal stimulation of growth in vitro can be obscured by an optimum concentration (10 ng/ml) of EGF. The relative growth responses to mammogenic hormones and EGF may depend on the degree of differentiation of the cells.

Linoleic acid, arachidonic acid, prostaglandin E1, and prostaglandin E2 stimulated the proliferation of mammary epithelial cells in serum-free primary cultures only in the presence of epidermal growth factor. Linoleate-stimulated growth was manifest later in culture when proliferation, initiated by epidermal growth factor only, reached a plateau while linoleate-supplemented epidermal growth factor cultures continued to proliferate. The cultures in the plateau phase of growth could be restimulated to grow by adding either linoleic acid or prostaglandin E2 to the media. While the linoleate response could be abolished by the cyclooxygenase inhibitor, indomethacin, prostaglandin E2-stimulated growth remained unaffected. Linoleic acid was metabolized to arachidonic acid and prostaglandin E2, both in the growing and resting cultures. Proliferating cells metabolized linoleate and prostaglandin E2 extensively so that neither the fatty acid nor prostaglandin E2 accumulated in large quantities in the proliferating cultures. The concentrations of prostaglandin E2 in growing cultures supplemented with linoleic acid were much higher than in cultures without it. These results suggest that the metabolism of linoleic acid leading to prostaglandin production, not its contribution to membrane polyunsaturation, is necessary for sustained growth of mammary epithelial cells in the presence of epidermal growth factor.

The effect of collagenase dissociation of virgin mouse mammary glands on the level of mammary epithelial cytosolic estrogen receptors (ER) and progesterone receptors (PR) was assessed. After cell dissociation, ER was present in mammary epithelial cells at concentrations similar to those found in the whole gland. However, PR appeared to be affected by the collagenase treatment. The regulation of ER and PR in mouse mammary epithelial cells isolated by collagenase dissociation and grown within collagen gels was then determined. After 7 days in culture under serum-free conditions inside a collagen gel, PR and, to a lesser extent ER, as characterized by high affinity binding and specificity, were present in the epithelial cells. Although at a low level, the ER were determined to be functional, since estradiol (E2) was able to promote nuclear accumulation of ER and to induce PR. PRL was able to increase cytosolic ER and PR concentrations. The combination of progesterone (P) and PRL was more effective than PRL or P alone in increasing PR. The induction of PR by P and PRL was inhibited when epidermal growth factor was present in the culture medium. Previous studies have shown that P, PRL, and epidermal growth factor, but not E2 (either alone or in combination with these factors) are able to stimulate cell proliferation in vitro. We conclude that the effects of E2 on protein synthesis and proliferation are dissociated in vitro. The difference between the effect of E2 and PRL or P on growth may be related either to the initial concentrations of their respective receptors or estrogen may stimulate growth indirectly.

The importance of hormones in the growth, morphogenesis, differentiation and neoplastic transformation of the mammary gland is well recognized due to a large number of experimentsin vivo (for review see 1). However, the precise nature of hormonal involvement is not clearly understood because of the difficulties in analyzing complex in vivo interactions. The role of growth factors in any of these processes is, perhaps, even less well understood, due in part to the complexity ofin vivo interactions and in part to the general lack of knowledge of the role of growth factors in anyin vivo process. The recent advances relating some viral oncogene products to growth factor receptor variants or growth factor sub-units makes it important to understand more fully the role of these molecules in the processes listed above. Culture systems of various types have been used to try and address these questions and are, theoretically, of great advantage since they allow observations and manipulations of cells and tissues in isolation, where experimental variables can be controlled or minimized. The goal of these investigations is to be able to reproduce in culture a system that mimics the in vivo physiological and pathological phemomena of interest. No individual system completely meets this goal, but recent advances allow a better simulation of some of the processes that occurin vivo. Unfortunately, even those systems which closely mimic the in vivo situation in terms of growth and differentiation cannot yet mimic the process of neoplastic transformation in culture.

Islet isografts were injected into the portal veins of rats made diabetic with streptozotocin. The isografts normalized not only plasma glucose and insulin levels but also the elevated plasma immunoreactive glucagon level. The in vitro basal insulin secretion and prompt sensitivity to glucose were shown directly by perfusing isolated livers containing transplanted islets. In vitro glucagon secretion to an arginine stimulus could not be demonstrated, although it would have been expected demonstrated, although it would have been expected in normal islets. Thus, it appears that insulin derived from transplanted islets is capable of correcting endogenous hyperglucagonemia and of ameliorating the effects of experimental diabetes while transplanted islet glucagon secretion is relatively suppressed.

The effect of lithium on the growth of mammary epithelial cells from adult virgin and midpregnant BALB/c or BALB/cfC3H mice was tested in a serum-free collagen gel culture system. The serum-free medium consisted of a 1:1 mixture of Ham's F12 and Dulbecco's Modified Eagle's medium supplemented with insulin, transferrin, cholera toxin, epidermal growth factor (EGF), and bovine serum albumin fraction V (BSA V). A multifold increase in cell number occurred during 10-12 days of culture in this medium. In dose-response studies in which the concentration of each component of this serum-free medium was varied in turn, the addition of LiCL (10 mM) enhanced growth at most concentrations of each factor. However, LiCl could not enhance growth in the absence of insulin or BSA V, but could replace EGF. The optimal concentration of LiCl was 5-10 mM; higher concentrations (20-80 mM) were toxic. KCl (1-10 mM) when added to the serum-free medium slightly stimulated growth; the addition of NaCl to the medium had little effect on growth. LiCl did not enhance the growth of cells from spontaneous mammary tumors of BALB/cfC3H mice.

The fundamental challenge in cancer biology today is to understand the abnormal growth control of tumor cells. Undoubtedly, as recent work on cellular oncogenes has indicated (I), there can be multiple abnormalities in growth regulation. The complexity of abnormal growth regulation seems imposing but, perhaps, it will become less intimidating if normal growth control is better understood. It is probably true that an understanding of the variations in tumor growth control must rest upon a foundation built from sound knowledge of the growth regulation of normal cells.

Mammary epithelial cells from virgin Balb/c mice were isolated by collagenase digestion and cultured within collagen gels in serum-free basal medium containing insulin (10 micrograms/ml). Previous work has shown that linoleate or its metabolite, prostaglandin E2 (PGE2), stimulate the growth of these cells only in the presence of a growth stimulant such as epidermal growth factor (EGF). Since PGE2 can stimulate cyclic AMP (cAMP) production, the role of cAMP in linoleate and EGF-stimulated growth was examined. The cAMP phosphodiesterase inhibitor, IBMX (0.1 mM), was found to augment growth when cells were cultured in the presence of both EGF and linoleate or PGE2, but not either factor alone. These results indicated that EGF does not stimulate proliferation via cyclic AMP mediated events but could synergize with cAMP events if cAMP levels were elevated by PGE2. When assayed in cells plated on top of collagen-coated culture dishes, cellular cyclic AMP levels were stimulated by PGE2, but only marginally by EGF. Although the stimulation of endogenous cAMP by PGE2 and IBMX was insufficient to stimulate growth in the absence of EGF, exogenous dibutyryl-cAMP (greater than 100 micrograms/ml) was able to do so showing that a sustained, and high level of cAMP (greater than 100 micrograms/ml) could stimulate growth in insulin-containing basal medium. EGF was capable of enhancing the cellular sensitivity to dibutyryl-cAMP but the converse was not observed. cAMP stimulation of growth was dependent upon a superphysiological concentration of insulin (10 micrograms/ml) or a physiological concentration of somatomedin-C. These results indicate that the proliferation of mouse mammary epithelial cells can be stimulated separately or in synergism by cAMP-dependent or -independent events.