Mammary epithelial cell differentiation is a complex process, in which quiescent ductular cells proliferate, forming alveolar structures that express their ...
1031
Journal of Cell Science 107, 1031-1040 (1994) Printed in Great Britain © The Company of Biologists Limited 1994
Laminin and tenascin assembly and expression regulate HC11 mouse mammary cell differentiation Roger Chammas1,*,†, Daniela Taverna2,*, Nathalie Cella1,*,‡, Cecilia Santos1 and Nancy E. Hynes2 1Ludwig Institute for Cancer Research, R. Prof. Antonio Prudente, 109/4, 2Friedrich Miescher-Institut, PO Box 2543, CH 4002 Basel, Switzerland
O1509-010 Sao Paulo, Brazil
*Contributed equally to this paper †Author for correspondence ‡Present adress: Friedrich Miescher-Institut, PO Box 2543, CH 4002 Basel, Switzerland
SUMMARY HC11 is a normal mouse mammary epithelial cell line that requires certain growth factors, such as EGF or bFGF, to respond optimally to lactogenic hormones and produce the differentiation marker β-casein. Growth in insulin (Ins) or PDGF does not produce cells competent to respond to lactogenic hormones. Here we show that competency for differentiation is due at least in part to the modulation of extracellular matrix components. In particular we have studied laminin and tenascin. EGF alters endogenous laminin assembly. In addition, promotion of competency can be partially mimicked by plating HC11 cells on the E8 laminin fragment, which is able to induce lactogenic responsiveness in cells grown in the absence of EGF or bFGF. The production and assembly of tenascin is also
INTRODUCTION Mammary epithelial cell differentiation is a complex process, in which quiescent ductular cells proliferate, forming alveolar structures that express their specialized products, the milk proteins. The differentiation process is driven by the cooperative action of multiple steroid and peptide hormones. In order to study the molecular effects of these hormones in mammary epithelial cell differentiation, an in vitro cultured cell system was developed. The HC11 cell line is a clonal derivative from COMMA-1D cells, isolated from mid-pregnant mammary gland cells of Balb/c mice (Danielson et al., 1984). HC11 cells display a normal phenotype and can produce β-casein when lactogenic hormones are added to confluent cells previously grown in the presence of EGF or bFGF (Taverna et al., 1991; Venesio et al., 1992). Work on mammary epithelial cells has shown that a correct microenvironment must be provided in order for the cells to respond optimally to lactogenic hormones. Either co-cultivation of epithelial cells with fibroblasts or adipocytes (cell-cell interaction) or cultivation of epithelial cells in the presence of extracellular matrix proteins (cell-matrix interactions) has been shown to be necessary (reviewed by Streuli, 1993, and Lin and Bissell, 1993).
dependent upon the growth conditions of the HC11 cells. EGF- or bFGF-grown competent cells produce tenascin but do not assemble it at the extracellular matrix as efficiently as Ins- or PDGF-grown, non-competent cells. This alteration apparently leads to a change in the cellular microenvironment that supports β-casein production. In addition, when competent cells are plated on dishes coated with tenascin, lactogenic hormone induction of β-casein is inhibited. The data suggest that tenascin assembly and βcasein production are opposing features of a coordinated differentiation program of HC11 cells. Key words: mammary epithelial cell differentiation, laminin, tenascin
HC11 cells are apparently different, since they need no cocultivation or addition of exogenous substrates (extracellular matrix proteins) to respond to the lactogenic hormones. In this paper we have examined this apparent contradiction and have observed that HC11 cells produce and assemble their own matrix, which is important for their differentiation. We have concentrated on two extracellular matrix glycoproteins, laminin and tenascin. Laminins represent a family of extracellular matrix glycoproteins. The prototype molecule was originally purified from the Engelbreth-Holm-Swarm (EHS) mouse sarcoma (Timpl et al., 1979). It is a heterotrimer of about 1,000,000 Da, composed of an A chain and two B chains (B1 and B2). Laminin isoforms can vary in the chain composition (reviewed by Engel, 1993). Laminin is a multidomain protein, and these domains can be separated by limited proteolysis, with e.g. elastase. There are generally three elastase-derived fragments: E1 (and E1-4), E3 and E8 fragments. Most of the cellular receptors belonging to the integrin family, described so far, interact with the E8 fragment. The E3 fragment interacts with heparan-sulfate proteoglycans (Beck et al., 1990) and dystroglycan (Gee et al., 1993). The globular ends of the E1-4 fragment interact with type IV collagen. Both E3 and E1-4 fragments are important in the supramolecular structure of basement membranes.
1032 R. Chammas and others Tenascins, which represent another family of high molecular mass extracellular matrix glycoproteins (>1,000,000 Da), are hexameric in structure. The prototype, tenascin C, is derived from chicken embryo fibroblasts (Chiquet and Fambrough, 1984; Bristow et al., 1993). Tenascins are expressed ontogenetically by mesenchymal and primordial epithelial cells (Chiquet-Ehrismann et al., 1986; Prieto et al., 1990; Koch et al., 1991) and have been considered modulators of mesenchymal-epithelial interactions. During embryogenesis it is found in the mesenchyme of teeth, kidneys, bones, cartilage, smooth muscle of the gut, skeletal muscle, heart, central nervous system and mammary gland, presenting a more widespread distribution than suggested previously (Erickson, 1993). Interestingly, tenascin is reexpressed in neoplasia, as mammary gland tumors e.g. (Mackie et al., 1987). We have studied the synthesis and assembly of laminin and tenascin during the growth and differentiation of HC11 cells. In addition, effects of these extracellular matrix proteins on βcasein production were examined. The data reported herein suggest a cellular basis for HC11 independence from exogenous matrix. In the appropriate culture conditions the HC11 cells produce a matrix that promotes their differentiation and allows them to respond optimally to lactogenic hormones. MATERIALS AND METHODS Protein purification and polyclonal antisera Fibronectin was purified from fresh human plasma on a gelatinSepharose affinity column (Pharmacia) (Engvall and Ruoslahti, 1977). Laminin-nidogen complex was purified from mouse Engelbreth-Holm-Swarm (EHS) tumor material (Paulsson et al., 1987), and used in the experiments described below. Antilaminin rabbit antibodies were produced as described (Lopes et al., 1985), and recognize the prototype laminin (EHS-laminin). Antitenascin antiserum and purified chicken tenascin (tenascin C) were both kindly provided by Dr R. Chiquet-Ehrismann (Friedrich-Miescher-Institut, Basel, Switzerland). Antitenascin antiserum recognizes all variants of tenascin-C. E1 and E8 laminin fragments were kindly provided by Dr J. Engel (Biocenter of the University of Basel, Switzerland). Rabbit anti-mouse β-casein serum was provided by Dr E. Reichmann (Vienna, Austria). Cell culture HC11 mammary epithelial cells were clonally derived from the COMMA-1D mouse mammary gland cell line (Danielson et al., 1984). These cells were routinely grown in RPMI 1640 containing 10% heatinactivated FCS, 5 µg/ml bovine insulin (Sigma) and 10 ng/ml murine EGF (purchased from Fluka). In the indicated experiments, EGF was replaced by 10 ng/ml of either human recombinant PDGF-BB or human recombinant bFGF (Bissendorf Biochemicals, Hannover) or was omitted. Two-day confluent cultures of HC11 cells were induced for β-casein production by treating them for 4 days with lactogenic hormone-containing medium. The latter contains RPMI 1640, 10% FCS, 5 µg/ml Ins, 5 µg/ml ovine prolactin (Sigma) and 1 µM dexamethasone (Sigma). All media contained 2 mM L-glutamine (Biochrom KG, Germany) and 50 µg/ml gentamicin (Biochrom KG). Effects of extracellular matrix components during HC11 differentiation The effect of extracellular matrix (ECM) on HC11 cell differentiation was examined by growing cells to confluency under various conditions, then removing them gently with PBS containing 0.025 M EDTA. The ECM-containing plates were washed extensively with
PBS. Cells grown in Ins or Ins+EGF were replated in the same medium onto the ECM-containing dishes and 24 hours later were incubated with lactogenic hormone-containing medium. In order to test the effects of ECM glycoproteins during the HC11 growth phase, cells were plated onto dishes coated with equimolar concentrations of fibronectin (5 µg/ml), laminin (10 µg/ml) and tenascin (10 µg/ml) and laminin fragments E1 (3 µg/ml) and E8 (3 µg/ml), in the absence of EGF. Two days after reaching confluence adherent cells were harvested as mentioned above and plated onto fresh dishes in the presence of induction medium. Effects of ECM during the induction phase was assessed by plating competent HC11 cells on laminin- (10 µg/ml) or tenascin-coated dishes (10 and 50 µg/ml), in the presence of induction medium. β-casein production was examined by washing cells with cold PBS, extracting them in a buffer containing 50 mM Tris (pH 7.5), 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 3 mM PMSF, 50 µg/ml leupeptin and 4 µg/ml pepstatin, for 10 minutes at 4°C. Cell debris was removed by centrifugation at 13,000 g for 5 minutes at 4°C. The concentration of the protein extract was determined using the micromethod of Bradford (Bio-Rad). Samples containing 50 µg of protein were separated by 11% SDS-PAGE (Laemmli, 1970), proteins were transferred onto PVDF membranes (Millipore) and β-casein was detected with a rabbit-specific antiserum, followed by incubation with 125I-Protein A. Autoradiograms were analyzed densitometrically, using an UltroScan-LKB (Pharmacia). Immunofluorescence assays HC11 cells were plated at a low density on LabTek chambers (Nunc) in the presence of various growth factors. Two-day confluent cultures received the lactogenic hormone mix for 3 days. Cells were washed with PBS and fixed for 15 minutes at −20°C in a solution containing 1% formaldehyde, 80% acetone and 19% PBS. Protein expression was assessed by indirect immunofluorescence with specific antisera and appropriate FITC-conjugates (Sigma). Stained cells were observed using an immunofluorescence microscope (Nikon). Analysis of intracellular, secreted and matrix assembled extracellular proteins Cell-conditioned medium (15 ml) was recovered from Ins- or Ins+EGF-grown confluent cultures, centrifuged to remove cell debris and concentrated in a Centricon 100 apparatus (Amicon). Detergent extracts of cells were prepared with 10 mM Tris-HCl, pH 7.4, 1% Triton X-100, 1% sodium deoxycholate, 1 mM PMSF, 2 µg/ml aprotinin and 2 µg/ml leupeptin. Soluble proteins represent the intracellular pool, whereas insoluble material was essentially matrix-associated proteins (Parry et al., 1985). Protein extracts were resolved by 3-10% gradient SDS-PAGE. Proteins were transferred to nitrocellulose filters and either tenascin or laminin were probed with specific antibodies and developed immunochemically with alkaline phosphatase conjugates (Sigma) and the proper substrates (nitro blue tetrazolium and bromo-chloro-indolyl-phosphate, BRL).
RESULTS Extracellular matrix influences the response of the HC11 cells to lactogenic hormones Primary mammary epithelial cells require either co-cultivation with mesenchymal cells or with exogenous extracellular matrix (ECM) components to differentiate in the presence of lactogenic hormones and produce milk proteins. HC11 cells are an apparent exception since they produce high quantitaties of βcasein in response to lactogenic hormones without co-cultivation or exogenous addition of ECM. We have observed that the cells require growth in the presence of EGF in order to become
EGF regulates ECM assembly during HC11 differentiation 1033
Fig. 1. Effect of ECM on the lactogenic hormone response of HC11 cells. Western blot analysis for β-casein. HC11 cells were grown in medium containing Ins (1 and 2) or Ins + EGF (3 and 4). Following 2 days at confluency cells were removed from the plates, left in their original medium and the equivalent of 1 (1 and 3) or 2 (2 and 4) confluent plates were replated on plastic, or ECM from cells grown in the absence (−) or presence (+) of EGF. One day later the cells were placed in lactogenic hormones and two days later cells were harvested and the level of β-casein was determined. C, positive control. Each lane contains equal amounts of protein extract.
competent to respond to lactogenic hormones (Taverna et al., 1991). In addition, the kinetics of the EGF effect are slow. The amplitude of the signal is time-dependent, since only cells kept for 3-4 days in the presence of EGF, before the lactogenic hormone induction, produce high levels of β-casein. Moreover, the effect of EGF is long-lasting since cells that are removed from EGF for up to 4 days before the addition of lactogenic hormones still synthesize high levels of β-casein following lactogenic hormone addition (Taverna et al., 1991; Hynes et al., 1992). These results suggest that EGF affects the ECM of the HC11 cells and that the ECM might function as a reservoir of differentiation stimuli. In order to test this hypothesis HC11 cells that had been grown in Ins + EGF or Ins alone for 3-4 days, and then maintained at confluency for 2 additional days, were removed from the cell culture plates using a technique that preserves the ECM on the plate surface. The deposited matrix proteins were analyzed in western blots using an antilaminin antiserum (data not shown). The former cells are referred to as competent while the latter are non-competent. Both types of cells were replated, at two densities, on fresh culture plates, or on plates containing the ECM from HC11 cells grown in Ins + EGF or Ins alone. One day later all cultures were treated with lactogenic hormones for two days, cell extracts were prepared and the expression of β-casein was analyzed using a protein blotting technique. The results are shown in Fig. 1. The replated competent cells (+) produced high or moderate levels of βcasein when plated on ECM produced by, respectively, Ins + EGF-grown cells (ECM:+EGF), or Ins-grown cells (ECM: −EGF). The lowest amount of β-casein was produced by competent cells that were replated on plastic. In all cases cells plated at a higher density expressed higher levels of β-casein, a fact that stresses the role of cell-cell interaction in the lactogenic hormone response. No β-casein was detected in noncompetent cells (−) replated on any type of matrix (Fig. 1). Therefore, although the ECM was necessary for the HC11 cells to respond optimally to the lactogenic hormones, it was not sufficient to replace the effect that EGF had upon the cells. In order to test if growth of cells on a competent matrix would replace EGF in the medium, HC11 cells were plated on competent matrix at low density and allowed to grow to confluency in Ins-containing medium. These cells produced no βcasein when treated with lactogenic hormones (data not shown). Neither growth on competent matrix nor replating high density cultures on competent matrix replaces EGF in the growth medium, although an optimal response to the lactogenic hormones was observed when competent cells were
plated on ECM derived from EGF-treated cells. These results suggest that EGF is important in cis, i.e. on the cells per se, as well as in trans, on the ECM. EGF can be partially replaced by ECM components Although complex ECM was not able to induce responsiveness to lactogenic hormones, we tested whether purified ECM components might have an effect on HC11 cell competence. The cells were grown in medium lacking EGF on culture dishes that had been coated with equimolar concentrations of fibronectin, tenascin, laminin or laminin fragments E1 and E8. In order to test only for the effects that these ECM components have during the growth of the HC11 cells, cells were harvested at two days following confluency and plated onto uncoated culture dishes. The cells were treated for three days with the lactogenic hormone mix and the level of β-casein was deter-
Fig. 2. ECM components partially replace EGF in the growth medium. Western blot analysis for β-casein. HC11 cells were grown in medium lacking EGF on plates coated with equimolar concentrations of fibronectin (FN), tenascin (TN) laminin (LN), and the E1 and E8 fragments of laminin. Two days following confluence the cells were removed, replated on plastic in the presence of lactogenic hormones and 3 days later the cells were harvested and the level of β-casein was determined using equal amounts of protein extract. Duplicate plates were analyzed. The control cells, grown on plastic in the absence (−) or presence (+) of EGF were treated exactly as the other cultures.
1034 R. Chammas and others mined using a protein blotting technique. The results presented in Fig. 2 show that the cells grown on the three ECM glycoproteins (FN, TN and LN) all synthesized low levels of βcasein. As a comparison the cells grown on plastic in medium lacking EGF synthesized no detectable β-casein (−). The effect of laminin could be assigned to its long arm since the E8 fragment but not the E1 fragment increased the response of the cells to lactogenic hormones. The E8 fragment was an even stronger competence promoter than intact laminin, although none of the ECM components were as effective as growth in EGF (+). EGF changes endogenous laminin organization and laminin chain synthesis The observation that the E8 laminin fragment could partially replace EGF during the growth phase suggests that laminin plays a role in the phenomenon of competence given by EGF. Consequently, laminin organization in the extracellular space during the HC11 cell growth phase was assessed by indirect
EGF
−
+
Fig. 3. Laminin organization during the growth of HC11 cells. Cells were grown in the absence (−) or presence (+) of EGF for 3 (A and B), 4 (C and D), or 5 (E and F) days. Cells were fixed, permeabilized and stained with a polyclonal serum specific for laminin. Bar, 50 µm.
immunofluorescence. Cells were grown for 3, 4 or 5 days either in medium containing Ins or medium containing Ins + EGF, fixed and stained with laminin-specific serum. The results are shown in Fig. 3. Ins-grown cells (Fig. 3A,C and E) progressively organized laminin in the extracellular space. Sparse cells presented laminin within cytoplasmic granules (Fig. 3A). As the cells grew (Fig. 3C) and reached confluence (Fig. 3E) laminin was deposited in the extracellular space organized in fibrils that outlined cell boundaries. Under these conditions the HC11 cells assumed a fibroblast-like morphology. HC11 cells grown for three days in the presence of EGF (Fig. 3B) expressed laminin in intracellular granules. Focal expression of laminin at the cell surface with formation of thin fibrils was observed on the fourth day (Fig. 3D). However, this pattern was transient since it was not observed after 5 days of growth (Fig. 3F). Laminin staining at confluence had essentially a cytoplasmic pattern, suggesting either synthesis of an isoform that is poorly assembled or an active process of laminin endocytosis. The differential pattern of laminin assembly upon EGF treatment led us to analyze the laminin chains secreted by HC11 cells in the presence or absence of EGF. To address this, supernatants of either Ins or EGF-treated cells were recovered, partially concentrated in a Centricon 100 apparatus and resolved by SDS-PAGE under reducing conditions. Laminin was identified with a protein blotting technique (Fig. 4). It is generally thought that only heterotrimeric laminins are secreted (reviewed by Engel, 1993). Supernatants of cells grown with Ins displayed 200 kDa laminin chains (A variant, B1 and B2), while supernatants of EGF-treated cells displayed two additional chains of 300 and 400 kDa, which may be A chain variants (Fig. 4, − vs +). These data suggest that either
Fig. 4. Effect of EGF on HC11 laminin secretion. Supernatants of HC11 cells grown in the absence (−) or in the presence (+) of EGF were collected, concentrated and separated by SDS-PAGE (gradient from 3 to 10%) under reducing conditions. Proteins were transferred onto nitrocellulose filters and a western blot using antilaminin antiserum was performed.
EGF regulates ECM assembly during HC11 differentiation 1035
Fig. 5. Tenascin organization during the growth of HC11 cells. Sparse (A and C) and confluent (B and D) HC11 cells, grown in the absence of EGF (A and B) or in its presence (C and D) were examined for their tenascin organization by means of indirect immunofluorescence. Bar, 50 µm.
the decrease in laminin deposition in the matrix network or the increase in endocytic activity of EGF-treated HC11 cells is associated with the synthesis of different A chains. Tenascin assembly correlates negatively with β-casein production Another ECM protein expressed by HC11 is tenascin. Organization of tenascin was examined in the HC11 cells grown under different conditions. Sparse cells grown both in the presence and absence of EGF displayed tenascin in intracytoplasmic granules (Fig. 5A,C). As cells reached confluence an apparent decrease in tenascin production was observed in the EGF-treated cultures (Fig. 5D), whereas thin fibrils of tenascin were observed in confluent Ins-grown cells (Fig. 5B). The cells grown in EGF appeared to contain less tenascin as they reached confluency. We examined tenascin expression during the lactogenic hormone induction of competent and non-competent HC11 cells. Fig. 6C shows that cells grown in EGF and treated for 3 days with lactogenic hormones did not assemble tenascin in the ECM and the cytoplasmic levels were barely detectable. As expected, these cells produce high levels of β-casein (Fig. 6D). In contrast non-competent, Ins-grown cells display thick fibrils of tenascin in the extracellular space (Fig. 6A) and essentially no β-casein (Fig. 6B). These data suggest that tenascin assembly correlates negatively with β-casein production.
It has been shown previously that growth of HC11 cells in bFGF (Venesio et al., 1992), but not in PDGF (Taverna et al., 1991), renders the cells competent to respond to lactogenic hormones. Therefore, we examined the pattern of tenascin distribution in cells grown either in bFGF or in PDGF. Similar to the results observed with EGF-grown cells, the bFGF-treated cells displayed very low staining for cytoplasmic tenascin, no tenascin was observed in the extracellular space (Fig. 7A), and high levels of β-casein were detected (Fig. 7B). In contrast, the PDGF-grown cells have a similar pattern of staining for tenascin as do the Ins-grown cells, i.e. the tenascin was present in the ECM (Fig. 7C) and these cells produce very low levels of β-casein (Fig. 7D). Results from immunofluorescence assays suggested that either there is a decrease in tenascin production or in its assembly during differentiation. Tenascin expression was assessed by western blots using culture supernatants, cell extracts and matrix protein-enriched extracts from HC11 cells (Fig. 8). Extracts were prepared from either competent or noncompetent cultures treated with the lactogenic hormones. No apparent differences were observed in the intracellular pool of tenascin from lactogenic hormone-treated competent (+) or non-competent (−) cultures. No soluble tenascin was observed (secreted). However, β-casein producing cells do not assemble tenascin in the extracellular space (matrix) as well as non-differentiated cells.
1036 R. Chammas and others
Fig. 6. Tenascin synthesis and assembly correlates negatively with β-casein production. Lactogenic hormone treated non-competent (Insgrown, A and B) or competent (EGF-grown, C and D) cells examined for tenascin (A and C) or β-casein (B and D) synthesis, by means of indirect immunofluorescence. In the conditions illustrated in B, 11.6±1.4% of the cells produced β-casein, whereas in the conditions illustrated in D, 31.03±2.6% of the cells synthesized β-casein. Bar, 50 µm.
Exogenous tenascin, but not laminin, inhibits β-casein production Results presented in the preceding section showed that tenascin assembly is different in differentiated versus non-differentiated cells, while laminin assembly was similar in the two conditions (data not shown). In order to examine the effect of these two proteins during the actual lactogenic hormone induction phase, competent HC11 cells, grown to confluency in EGF-containing medium, were plated on 10 µg/ml laminin- or on 10 and 50 µg/ml tenascin-coated dishes. The results are shown in Fig. 9. In comparison with the level of β-casein produced by the cells replated on laminin (LN) or on plastic (ctl) the cells replated on 10 or 50 µg/ml tenascin (TN 10 and 50) expressed, respectively, 69% and 78% less β-casein. Therefore, the inhibitory effect of exogenous tenascin was dose-dependent. These data suggest that the negative correlation between tenascin assembly and β-casein is not only an epiphenomenon.
DISCUSSION HC11 mammary epithelial cells, in contrast to most mammary cell models studied (Lin and Bissell, 1993), which require addition of ECM protein or cocultivation with mesenchymal elements, depend only on growth in EGF in order to respond
to lactogenic hormones. We have shown that the ECM deposited by the HC11 cells enhances their ability to respond to lactogenic hormones. EGF affects laminin organization and assembly, which appears to modulate lactogenic hormone responsiveness. In fact, EGF in the medium could be partially replaced by plating the cells on the E8 laminin fragment. In addition, we have shown that tenascin assembly is modulated during HC11 cell growth and differentiation and there is a negative correlation between the presence of tenascin and βcasein production. EGF and the E8 laminin fragment induce lactogenic responsiveness in HC11 cells We have shown previously that the EGF effect on the competence of the HC11 cells is long lasting (Taverna et al., 1991; Hynes et al., 1992). In this paper we show that competent cells produce greater amounts of β-casein when plated on matrix assembled by EGF-cultured cells. These observations provide the first evidence suggesting that in HC11 cells EGF affects matrix production or assembly and that a particular matrix is involved in the differentiation process. Purified ECM proteins, in particular the E8 laminin fragment, were able to induce β-casein production in HC11 cells cultured without EGF. Heterologous displacement of EGF binding to its receptor by laminin and its fragments was not observed (data not shown). This rules out the possibility
EGF regulates ECM assembly during HC11 differentiation 1037
Fig. 7. bFGF and PDGF effects on tenascin and β-casein synthesis. Lactogenic hormone-treated non-competent (PDGF-grown, C and D) or competent (bFGF-grown, A and B) cells examined for tenascin (A and C) or β-casein (B and D) synthesis, by means of indirect immunofluorescence. In the conditions illustrated in B, 31.9±7.02% of the cells produced β-casein, whereas in the conditions of D, 7.7±2.1% of the cells were positive for casein staining. Bar, 50 µm.
that laminin or the E8 fragment act via the EGF receptor or that this effect was mediated by contaminating EGF, which may reside in the EHS matrix, our source of laminin. Panayotou et al. (1989) have also shown that none of the laminin fragments can inhibit EGF binding to its receptor. The possibility of indirect EGF receptor activation through the interaction of the E8 fragment with its receptor cannot be excluded. The observation that laminin does not exert as strong an effect as E8 could be due to the conformation of both molecules adsorbed to culture dishes. In vitro, laminin forms an independent network (Yurchenco et al., 1992), and in this network the E8 domain might be less available for cell interaction. Furthermore, E8-mediated adhesion and spreading depends on its secondary and tertiary structure (Deutzmann et al., 1990), which is probably different when this domain is in the intact molecule. During in vitro kidney tubule development, the E8 laminin fragment directly induces epithelial cell polarization (Klein et al., 1988). This effect is associated with expression of the α6 integrin chain (Sorokin et al., 1990). HC11 cells also express α6-containing integrins, both α6β1 and α6β4 (N. Cella and R. Chammas, unpublished results). The effect of EGF on the α6 expression at the cell surface is now being addressed. It is possible that the observed competence of E8 plated cells is due to the induction of cell polarization. It has
been shown for 31E mammary epithelial cells that the maintenance of a polarized phenotype is enough to induce lactogenic hormone responsiveness (Strange et al., 1991). Recently, it has also been shown that anti-β1-integrins could inhibit β-casein production in primary culture of mammary epithelial cells (Streuli et al., 1991). The amount of β-casein produced by cells plated on E8 was far less than that produced by EGF-grown cells, indicating that E8 replaces EGF only partially. EGF could also be acting in an indirect manner, inducing secretion of cellular factors that bind to the ECM and also affect HC11 lactogenic responsiveness. Streuli et al. (1992) have suggested that laminin can induce mammary epithelial cell differentiation in C1D-9 cells, which, like HC11 were derived from COMMA-1D. Interestingly, laminin also leads to accumulation of α-casein mRNA, by increasing its stability (Zeigler and Wicha, 1992). In the context of a complex ECM produced by HC11 cells, the role of laminin in the differentiation process is less clear. EGF modifies the assembly of laminin in the extracellular space, causing it to change from a fibrillar organized pattern to a granular pattern. We have shown that different laminin chains are present in the supernatant of EGF-grown cells. A 200 kDa molecule, which can be either laminin B chains or a smaller A chain (Engel, 1993) appears in the supernatant of both Ins- and EGF-treated cells. Two other laminin chains of
1038 R. Chammas and others
Fig. 8. Tenascin assembly, but not synthesis, is modulated in differentiated HC11 cells. Western blot analysis for tenascin. Conditioned medium, intracellular pool and matrix assembled proteins of either non-differentiated HC11 cells (β-casein negative) or differentiated cells (β-casein positive) were analyzed for tenascin expression. Size standards (in kDa) are shown on the left.
300 and 400 kDa were detected only in the supernatant of EGFgrown cells. Laminin A chains of these sizes have been described in other tissues (Paulsson et al., 1991). EGF may inhibit the assembly of these laminin isoforms. EGF exerts two effects on HC11 cell differentiation. During EGF treatment of the cells specific ECM organization contributes to their acquisition of hormone responsiveness. EGF also has an intracellular effect. Ins-grown cells never produce as much β-casein as their EGF-grown counterpart, suggesting that additional cellular pathways activated by the EGF-receptor are involved. Tenascin can inhibit β-casein expression In contrast to laminin, the ECM protein tenascin has an inhibitory influence on β-casein production. Tenascin is a glycoprotein produced when mesenchymal cells are induced by epithelial cells (Prieto et al., 1990). Tenascin is also expressed by primordial epithelial cells (Koch et al., 1991; Sakai et al., 1993), but has a restricted distribution in adult tissues. The fact that tenascin is expressed by HC11 mammary epithelial cells, suggests that this cell line has characteristics of a stem cell. In fact, a proportion of HC11 cells co-express keratin 6 and 14 (G. H. Smith, personal communication) a characteristic of the presumptive mammogenic stem cell (Smith and Medina, 1988; Smith et al., 1990). It has also been reported that tenascin is expressed in human mammary cells at different stages of the menstrual cycle (Ferguson et al., 1990). Western analysis revealed that the predominant tenascin chain was a 210-220 kDa polypeptide. This size corresponds to the smallest isoform described in primary murine embryonic
Fig. 9. Effect of laminin and tenascin on lactogenic hormone induction of β-casein. Western blot analysis for β-casein. Competent, EGF-grown cells were plated on: plastic (ctl) or laminin (LN 10 µg/ml) and tenascin (TN, 10 and 50 µg/ml)-coated dishes, and incubated with lactogenic hormone containing medium. Duplicate plates were made. Cells were harvested and the level of β-casein was determine using equal amounts of protein extract. Results were evaluated densitometrically and the level of β-casein in comparison to the control is indicated below the lanes.
fibroblasts (Weller et al., 1991). Interestingly, the smallest chicken tenascin variant is only detected in the ECM network (Chiquet-Ehrismann et al., 1991). In HC11 cells this 220 kDa variant is also deposited in the ECM, in agreement with the previous observation. There is no significant difference in the intracellular amounts of tenascin in lactogenic hormoneinduced cells previously grown in Ins or EGF. Extracellular deposition of tenascin was detected by immunofluorescence. Ins-grown cells express high amounts of tenascin, which is organized as fibrils at the extracellular space following treatment of the cells with lactogenic hormones. EGF-grown cells express lower levels of tenascin and following lactogenic hormone induction the tenascin is not assembled. The results from immunofluorescence studies suggest that tenascin assembly and β-casein production are mutually exclusive. EGF treatment seems to guide HC11 cells to an epithelial phenotype, since their morphology and lactogenic responsiveness markedly change. In this regard, the decrease in tenascin assembly might be due to the phenotypic shift of HC11 cells along the epithelial pathway. Tenascin can also abolish βcasein expression, since competent cells do not respond to lactogenic hormones when they are plated on tenascin surfaces. Tenascin assembly and β-casein production seem to be the two opposite sides of a coordinated differentiation program of HC11 cells. We also have studied the effects of two other growth factors on the assembly of tenascin in HC11 cells. bFGF-treated cells
EGF regulates ECM assembly during HC11 differentiation 1039 are competent to produce β-casein in response to lactogenic hormones, and tenascin assembly was suppressed. It is interesting that in mouse fibroblasts, bFGF induces the expression of tenascin mRNA and secretion of tenascin protein (Tucker et al., 1993). On the other hand, cells treated with PDGF, which is mitogenic but does not induce competence (Taverna et al., 1991) organized tenascin in fibrils in the extracellular space. It is interesting that dexamethasone, one of the lactogenic hormones, induces a decrease in the secretion of tenascin in a stromal cell line and in fibroblasts (Ekblom et al., 1993). In the HC11 cells, dexamethasone exerts its inhibitory effect only on EGF- or bFGF-primed cells. Control of β-casein gene expression appears to be mainly at the transcriptional level. Promoter elements that respond to ECM (Schmidhauser et al., 1992) and to lactogenic hormones (Schmitt-Ney et al., 1991) have been described. BCE-1, an ECM-responsive transcriptional enhancer in the bovine βcasein promoter, is located approximately 1.5 kb upstream of the RNA start site. The lactogenic hormone-responsive trancription factor MGF binds between positions −80 and −100 bp in the rat β-casein promoter. MGF activity is present in lactating mammary glands from various species including rodent and bovine (Wakao et al., 1992). In HC11 cells its binding is lactogenic hormone-dependent and absolutely essential for promoter activity (Schmitt-Ney et al., 1991). It is interesting that between nucleotides −1575 and −1565, BCE-1 contains some homology to the consensus binding site for MGF. MGF binding is lost in HC11 cells that do not produce β-casein due to, e.g. expression of the ras or raf oncogenes (Happ et al., 1993). We can now examine the effects of ECM proteins on MGF activity. It would be interesting to see if the negative effects of tenascin are due to its effects on MGF activity. The HC11 cell line will be a valuable tool for studying the effects of ECM proteins on molecular aspects of mammary epithelial cell differentiation. R.C. received an UICC-ICRETT grant for this project. D.T. was supported by the Swiss Cancer League. N.C. was a recipient of a grant from FAPESP (Fundacao de Amparo a Pesquisa do Estado de Sao Paulo).The authors are indebted to Drs R. Chiquet-Ehrismann and J. Engel, for providing some reagents and useful suggestions. Dr E. Reichmann is acknowledged for providing us with anti-β-casein antiserum. The technical assistance of U. Stiefel and L. C. B. de Sousa are greatly acknowledged. We are also indebted to Drs R. Brentani, B. Groner and M. Schmitt-Ney for helpful discussions.
REFERENCES Aufderheide, E. and Ekblom, P. (1988). Tenascin during gut development: appearence in the mesenchyme, shift in molecular forms, and dependence on epithelial-mesenchymal interactions. J. Cell Biol. 107, 2341-2349. Ball, R. K., Friis, R. R., Schoenenberger, C. A., Doppler, W. and Groner, B. (1988). Prolactin regulation of β-casein gene expression and of a cytosolic 120-kDa protein in a cloned mouse mammary epithelial cell line. EMBO J. 7, 2089-2095. Beck, K., Hunter, I. and Engel, J. (1990). Structure and function of laminin: anatomy of a multidomain glycoprotein. FASEB J. 4, 148-160. Bristow, J., Tee, M. K., Gitelman, S. E., Mellon, S. and Miller, W. L. (1993). Tenascin-X: a novel extracellular matrix protein encoded by the human XB gene overlapping P450c21B. J. Cell Biol. 122, 265-278. Chiquet, M. and Fambrough, D. (1984). Chick myotendinous antigen I. A monoclonal antibody as a marker for tendon and muscle morphogenesis. J. Cell Biol. 98, 1926-1936.
Chiquet-Ehrismann, R., Mackie, E., Pearson, C. and Sakakura, T. (1986). Tenascin: an extracellular matrix protein involved in tissue interactions during fetal development and oncogenesis. Cell 47, 131-139. Chiquet-Ehrismann, R., Matsuoka, Y., Horfer, U., Spring, J., Bernasconi, C. and Chiquet, M. (1991). Tenascin variants: differential binding to fibronectin and distinct distribution in cell cultures and tissues. Cell Regul. 2, 927-938. Danielson, K., Oborn, C., Durban, E., Butel, J. and Medina, D. (1984). Epithelial mouse mammary cell line exhibiting normal morphogenesis in vivo and functional differentiation in vitro. Proc. Nat. Acad. Sci. USA 81, 3756-3760. Deutzmann, R., Aumailley, M., Wiedemann, H., Pysny, W., Timpl, R. and Edgar, D. (1990). Cell adhesion, spreading and neurite stimulation by laminin fragment E8 depends on maintenance of secondary and tertiary structure in its rod and globular domain. Eur. J. Biochem. 191, 513-522. Ekblom, M., Fassler, R., Tomasini-Johansson, B., Nillson, K. and Ekblom, P. (1993). Downregulation of tenascin expression by glucocorticoids in bone marrow stromal cells and in fibroblasts. J. Cell Biol. 123, 1037-1045. Engel, J. (1993). Structure and function of laminin. In Molecular and Cellular Aspects of Basement Membranes (ed. D. Rohrbach and R. Timpl), pp. 147176. Academic Press, San Diego. Engvall, E. and Ruoslahti, E. (1977). Binding of soluble form of fibroblast surface protein, fibronectin, to collagen. Int. J. Cancer, 20, 1-7. Erickson, H. P. (1993). Tenascin-C, tenascin-R and tenascin-X: a family of talented proteins in search of functions. Curr. Opin. Cell Biol., 5, 869-876. Ferguson, J. E., Schor, A. M., Howell, A. and Ferguson, M. W. (1990). Tenascin distribution in the normal human breast is altered during the menstrual cycle and in carcinoma. Differentiation 42, 199-207. Gee, S., Blacher, R., Douville, P., Provost, P., Yurchenco, P. and Carbonetto, S. (1993). Laminin-binding protein 120 from brain is closely related to the dystrophin-associated glycoprotein, dystroglycan, and binds with high affinity to the major heparin binding domain of laminin. J. Biol. Chem. 268, 14972-14980. Happ, B., Hynes, N. and Groner, B. (1993). Ha-ras and v-raf oncogenes, but not int-2 and c-myc, interfere with the lactogenic hormone dependent activation of the mammary gland specific transcription factor. Cell Growth Differ. 4, 9-15. Hynes, N. Taverna, D., Cella, N. and Chammas, R. (1992). The effect of extracellular matrix upon HC11 mammary epithelial cell differentiation. Ci. Cultura 44, 253-256. Klein, G., Langegger, M., Timpl, R. and Ekblom, P. (1988). Role of laminin A chain in the development of epithelial cell polarity. Cell 55, 331-341. Koch, M., Wehrlle-Haller, B., Baumgartner, S., Spring, J., Brubacher, D. and Chiquet, M. (1991). Epithelial synthesis of tenascin at tips of growing bronchi and graded accumulation in basement membrane and mesenchyme. Exp. Cell Res. 194, 297-300. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227, 680-685. Lin, C. and Bissell, M. (1993). Multi-faceted regulation of cell differentiation by extracellular matrix. FASEB J. 7, 737-743. Lopes, J. D., Reis, M. and Brentani, R. (1985). Laminin receptor in Staphylococcus aureus. Science 229, 275-277. Mackie, E., Chiquet-Ehrismann, R., Pearson, C., Inaguma, Y., Taya, K., Kawarada, Y. and Sakakura, T. (1987). Tenascin is a stromal marker for epithelial malignancy in the mammary gland. Proc. Nat. Acad. Sci. USA 84, 4621-4625. Panayotou, G., End, P., Aumailley, M., Timpl, R. and Engel, J. (1989). Domains of laminin with growth-factor activity. Cell 56, 93-101. Parry, G., Lee, E., Farson, D., Koval, M. and Bissell, M. (1985). Collagenous substrata regulate the nature and distribution of glycosaminoglycans produced by differentiated cultures of mouse mammary epithelial cells. Exp. Cell Res. 156, 487-499. Paulsson, M., Aumailley, M., Deutzmann, R., Timpl, R., Beck, K. and Engel, J. (1987). Laminin-nidogen complex. Extraction with chelating agents and structural characterization. Eur. J. Biochem. 166, 11-19. Paulsson, M., Saladin, K. and Engvall, E. (1991). Structure of laminin variants: the 300-kDa chains of murine and bovine heart laminin are related to the human placenta merosin heavy chain and replace the A chain in some laminin variants. J. Biol. Chem. 266, 17545-17551. Prieto, A., Jones, F., Cunningham, B., Crossin, K. and Edelman, G. (1990). Localization during development of alternatively spliced forms of cytotactin mRNA by In Situ hybridization. J. Cell Biol. 111, 685-698. Sakai, T., Kawakatsu, H., Hirota, N., Yokoyama, T., Takaoka, T., Sakakura, T. and Saito, M. (1993). Tenascin expression in vitro and in
1040 R. Chammas and others vivo: comparison between epithelial and nonepithelial rat cell lines. Exp. Cell Res. 206, 244-254. Schmidhauser, C., Casperson, G., Myers, C., Bolten, S. and Bissell, M. (1992). A novel transcriptional enhancer is involved in the prolactin- and extracellular matrix-dependent regulation of β-casein gene expression. Mol. Biol Cell 3, 699-709. Schmitt-Ney, M., Doppler, W., Ball, R. K. and Groner, B. (1991). β-casein gene promoter activity is regulated by the hormone-mediated relief of transcriptional repression and a mammary-gland-specific nuclear factor. Mol. Cell. Biol. 11, 3745-3755. Smith, G. and Medina, D. (1988). A morphologically distinct candidate for an epithelial stem cell in mouse mammary gland. J. Cell Sci. 89, 173-183. Smith, G., Mehrel, T. and Roop, D. (1990). Differential keratin gene expression in developing, differentiating, preneoplastic, and neoplastic mouse mammary epithelium. Cell Growth Diff. 1, 161-170. Sorokin, L., Sonnenberg, A., Aumailley, M., Timpl, R. and Ekblom, P. (1990). Recognition of the laminin cell-binding site by an integrin possessing the α6 subunit is essential for epithelial polarization in developing kidney tubules. J. Cell Biol. 111, 1265-1273. Strange, R., Li, F., Friis, R., Reichmann E., Haenni, B. and Burri, H. (1991). Mammary epithelial differentiation in vitro : minimum requirements for a functional response to hormonal stimulation. Cell Growth Differ. 2, 549-559. Streuli, C., Bailey, N. and Bissell, M. (1991). Control of mammary epithelial differentiation: basement membrane induces tissue-specific gene expression in the absence of cell-cell interaction and morphological polarity. J. Cell Biol. 115, 1383-1395. Streuli, C., Bailey, N., Yurchenco, P. and Bissell, M. (1992). Purified laminin can induce differentiation in mammary epithelial cells. 5th Int. Congr. Cell Biol. 178 (abstract).
Streuli, C. (1993). Extracellular matrix and gene expression in mammary epithelium. Semin. Cell Biol. 4, 203-212. Taverna, D., Groner, B. and Hynes, N. (1991). Epidermal growth factor receptor, platelet-derived growth factor receptor, and c-erbB-2 receptor activation all promote growth but have distinctive effects upon mouse mammary epithelial cell differentiation. Cell Growth Differ. 2, 145-154. Timpl, R., Rohde, H., Robey, P. G., Rennard, S., Foidart, J. and Martin, G. (1979). Laminin - a glycoprotein from basement membranes. J. Biol. Chem. 254, 9933-9937. Tucker, R. P., Hammarback, J. A., Jenrath, D. A., Mackie, E. J., Xu, Y. (1993). Tenascin expression in the mouse: in situ localization and induction in vitro by bFGF. J. Cell Sci. 104, 69-76. Venesio, T., Taverna, D., Hynes, N., Deed, R., MacAllen, D., Ciardiello, F., Valverius, E., Salomon, D. S., Callahan, R. and Merlo, G. (1992). The int2 gene product acts as a growth factor and substitutes for basic fibroblast growth factor in promoting the differentiation of a normal mouse mammary epithelial cell line. Cell Growth Differ. 3, 63-71. Wakao, H., Schmitt-Ney, M. and Groner, B. (1992). Mammary glandspecific nuclear factor is present in lactating rodent and bovine mammary tissue and composed of a single polypeptide of 89 kDa. J. Biol. Chem. 267, 16365-16370. Weller, A., Beck, S. and Ekblom, P. (1991). Amino acid sequence of mouse tenascin and differential expression of two tenascin isoforms during embryogenesis. J. Cell Biol. 112, 355-362. Yurchenco, P., Cheng, Y. and Colognato, H. (1992). Laminin forms an independent network in basement membranes. J. Cell Biol. 117, 1119-1133. Zeigler, M. and Wicha., M. (1992). Posttranscriptional regulation of α-casein mRNA accumulation by laminin. Exp. Cell Res. 200, 481-489. (Received 4 November 1993 - Accepted 20 December 1993)
