JOURNAL OF CELLULAR PHYSIOLOGY 145~162-172(1990)
Effects of Differentiation-Inducing Agents on Maturation of Human MCF-7 Breast Cancer CelIs NICOLAS F. GUILBAUD,* NICOLE GAS, MARIE ANGE DUPONT, A N D ANNIE VALETTE laboratoire de Pharmacologie et Toxicologie Fondamentales, CNRS, 3 1077 Toulouse CEDEX, France (N.F.G., A.V.); Centre de Recherche de Biochimie et de Cenetique Cellulaires, CNRS, 3 1062 Toulouse CEDEX, France (N.C., M.A.D.)
The effectsof the differentiation inducing agents (DIAs), sodium butyrate (NaBu), retinoic acid (RA), dimethylformamide (DMF), hexamethylene bisacetamide (HMBA), forskolin, and 12-O-tetradecanoylphorbol-13-acetate (TPA), on the growth, morphology, and estrogen receptor (ER) content and epithelial membrane antigen (EMA) expression on a serumless human breast cancer cell line (MCF-7) were compared. All these agents reversibly caused a concentration-dependent growth inhibition in monolayers and markedly reduced colony-formingefficiency in soft agar. A twofold increase in doubling time was obtained with RA (1 FM), but cell replication ceased with NaBu (1 mM), forskolin (50 (IM), DMF (1 %), H M B A (5 mM), and TPA (8 nM). Total growth arrest induced by these last compounds was preceded by an accumulation of cells in GO/Gl phase observed at 24 h by flow cytometry and accompanied by a change in cell morphology as seen by light and electronic microscopy. An increase in cell volume and the presence of lipid droplets was noted in treated cells that were spread out, as compared with controls. The acquisition of a more mature phenotype was confirmed by an increased expression of EMA monitored by flow cytometry. A specific reduction in the number of ER without any constant dissociation (Kd) modification was also observed after treatment with the 5 DIAs. No modification of morphological or biochemical characteristics, including EMA expression and ER binding, were observed for RA (1 p,M)-treated cells. All these results suggest that induction of a more differentiated phenotype is associated with a block in G1 cell cycle phase, resulting in total growth arrest. Apparently, RA (1 FM)-treated cells did not fulfill these criteria, since only a slight accumulation in G1 and a slowed growth rate were evaluated. Since cancer can be regarded as a disorder of cell differentiation, the use of drugs that induce the matu: ration of tumor cells has been envisaged as a treatment strategy. However, evaluation of such drugs requires model systems that are representative of the most common clinical tumor types. Although there are several well-characterized models for leukemic cell differentiation (Reiss et al., 1986), there are few in vitro systems for the study of differentiation of malignant human epithelial cells. However, several reports indicate that maturation of adenocarcinoma cells can be induced chemically both in vitro and in vivo (Calabresi et al., 1979; Freshney, 1985; Tsao et al., 1982; Langdon et al., 1988). Estrogen receptor-positive ER(+) and estrogen receptor negative ER(-) breast cancer cell lines provide in vitro models for investigations on hormonally responsive and unresponsive breast cancer (Dickson and Lippman, 1987).The effects of differentiation-inducing agents (DIAs) have been studied on a panel of ER(+) or ER(-) breast cancer cells. DIAs include retinoids (vit A and its synthetic analogues), derivatives of vit D, cyclic adenosine monophosphate (CAMP)and its analogues, and certain small polar 0 1990 WILEY-LISS, INC.
planar molecules, such as dimethylformamide (DMF) and hexamethylene bisacetamide (HMBA).These compounds inhibit cell proliferation in ER( -) breast cancer cell lines, whereas their antiproliferative effects are more often associated with a maturation process in E R ( t ) breast cancer cells. However, vit D3 has been reported to induce a maturation process in ER( -1 BT20 cells (Falette et al., 1988). The maturation process of ER(+) cells is characterized by morphological change and an increase in the expression of milk fat globule antigens, such as DF3 (Abe and Kufe, 1984a). It was shown that tumor promoting agents such as 12-O-tetradecanoylphorbol-13-acetate (TPA) inhibit the proliferation of MCF-7 cells (Osborne et al., 1981; Darbon et al., 1986) and that this inhibition of proliferation results from a specific block in the G1 phase of the cell cycle (Valette et al., 1987). After 2 days of TPA treatment, there is a specific reduction in ER number, accompanied by an induction of the morphological Received December 29, 1989; accepted June 11, 1990. *To whom reprint requestsicorrespondence should be addressed.
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
characteristics of a secretory cell type, a more mature phenotype of MCF-7 cells (Valette et al., 1987; Guilbaud et al., 1988). There is a good correlation between the potency of these phorbol ester derivatives to bind to a high affinity receptor (probably C kinase, PKC) and their ability to inhibit MCF-7 proliferation (Darbon et al., 1986) and reduce ER concentration (Guilbaud et al., 1988). Moreover, the ER of MCF-7 : RPh4 cells (which are insensitive to the antiproliferative and morphological effects of TPA) is not affected by exposure to TPA. Several possibilities could account for this decrease in ER content. ER could be a specific substrate for PKC. The decrease in estrogen receptor could be due to the decrease in growth rate observed after TPA treatment of MCF-7 cells. Yet another explanation is that there is a relationship between the decrease in expression of ER and the maturation process. In order to address this issue, we compared the effects of various DIAs with those induced by TPA. We attem ted to establish relationships between the antiproli erative effect (cell cycle progression), the maturation effect (change in morphology, expression of milk fat globule antigen) and change in ER content after treatment of MCF-7 cells with DIAs that act by different molecular mechanisms.
P
163
underlayer in 4% FCS growth medium. The number of colonies present after 15 days were determined after staining with 3,4,5-dimethylthiazo1-2-yl-2,5-diphenyltetrazolium bromide (MTT) using a colony analyzer. Colonies >50 Fm in diameter were scored as positive. Flow cytometry To determine the cell cycle distribution of treated MCF-7 cells, the cells were removed from the culture surface with trypsin, fixed, stained with propidium iodide, and analyzed using a Coulter Epics fluorescence-activated cell sorter as previously described (Valette et al., 1987). To measure EMA induction, MCF-7 cells were grown for 6 days in the presence or absence of different drugs as indicated. Cells were washed once with phosphate-buffered saline (PBS) and then incubated with rabbit antihuman EMA antibody or rabbit control serum (diluted 1 : 200) for 12 h at 4°C. Cells were then washed three times with PBS and incubated with FITC-conjugated goat antirabbit immunoglobulin diluted 1 : 50 in PBS for 1 h at 4°C. Cells were then washed three times with PBS and fixed in 70% ethanol PBS and analyzed in a Coulter Epics fluorescenceactivated cell sorter. The peak fluorescence was taken as the fluorescence channel with the highest number of cells. Peak fluorescence was then converted from channel units of logarithmic fluorescence to linear units. Electron microscopy After exposure to different drugs, MCF-7 cells were fixed for 60 min with 3% glutaraldehyde in 0.1 M sodium cacodylate buffer and postfixed in 1%osmium tetroxide in 0.1 M sodium cacodylate buffer, dehydrated in graded ethanol, and embedded in Epon 812. Sections were stained with uranyl acetate and lead citrate, and examined in a JEOL 1200 EX electron microscope. Estrogen receptor measurements All procedures were performed at 4°C as described above. The cells were harvested by scraping with a rubber policeman and pelleted by centrifugation at 800g for 10min. The pellets were washed with 0.9% NaC1, resuspended in 20 mM TrisHCl buffer containing 20 mM sodium molybdate-10 mM monothioglycerol pH 7.4, at 24°C. The cells were sonicated twice with a SCMA Pons Machine type 20200 SV, at a 1.25-W setting for 5 s each, with a 30-s interval for cooling. The homogenate was adjusted to 0.4 M KCl and centrifuged at 105,OOOgfor 60 min. The protein concentration of the supernatant was determined by the method of Lowry et al. (1951), with serum albumin as standard. The DNA content of the pellet was measured by the method of Burton (1956). For measurements of ER, samples of 100 p1 of supernatant were incubated at 4°C in triplicate in the presence of 0.1-20 nM [3Hl-17p-estradiol. Nonspecific binding was determined by parallel incubation in the presence of 1 pM of nonradioactive 17pestradiol. Bound and unbound ligand were separated by LH20 chromatography.
MATERIALS AND METHODS Chemicals TPA, forskolin, retinoic acid (RA) (all trans), hexamethylene bisacetamide (HMBA),dimethylformamide (DMF),n-butyric acid (sodium salt), 17p-estradiol were obtained from Sigma Chemical Co., [3Hl-17p-estradiol (sp. act. 80 Cilmmol) was provided by Amersham. Antisera to epithelial membrane antigen (EMA) were kindly provided from Dr. Hendrick and Dr. Collette (Institute of Pathology, University of Liege). Cells MCF-7 cells previously adapted to grow in serumfree medium (Jozan et al., 1982) were maintained in RPMI 1640 medium supplemented with sodium bicarbonate (2 g/L), 2 mM glutamine, 1 pM insulin, and 0.1 FM transferrin in humidified 5% C02/95% air at 37°C. Growth curves in monolayer cultures MCF-7 cells growing in logarithmic phase were plated in Petri dishes at a density of lo5 cellddish in RPMI medium containing 0.5% fetal calf serum (FCS). The following day, the medium was replaced by RPMI serum-free medium containing the different drugs. The culture medium was changed every 3 days. After treating the cells with 0.05% try sin-0.02% EDTA for 1min, the trypsin was removeg and the cells were harvested in RPMI medium containing 0.5% FCS. Cell numbers were determined in a Coulter counter (Coultronics). Anchorage-independent assay RESULTS Clonal growth assay were performed as described by Effects of inducers on cell growth Knabbe et al. (1987). Briefly, 10,000 cells/35-mm dish in RPMI medium supplemented with 4% FCS, 0.4% The effects of NaBu, DMF, forskolin, and RA on the agar, and different drugs were plated on a 0.6% agarose growth of serum-free MCF-7 are illustrated in Figure 1.
164
GUILBAUD ET AL.
NaBu (1 mM), DMF (l%), HMBA (5mM),TPA (8 nM), and forskolin (50 KM) block MCF-7 proliferation after 6-day treatment. Under these conditions, 90-95% of MCF-7 cells remained viable, as assessed by Trypan blue dye exclusion. A decrease in growth rate was observed after exposure of MCF-7 cells to RA (1p.M) for 6 days, The antiproliferative effects of these compounds were dose dependent (Fig. 2). The maximum inhibition was about 50% for RA concentration of 10-6M; for concentrations above M, significant toxicity was observed (data not shown). Incubation of MCF-7 with the different inducers for 24 h led to an increase in the number of cells in the Gl phase of the cell cycle and a concomitant fall of the number of cells in S phase (Table 1). RA (1 pM) treatment induced the least accumulation in the G1 phase (69% vs. 56% in control cells) and the smallest decrease in the percentage of S-phase cells (16%vs. 24% in control cells) at 24h. This effect of RA almost completely disappeared after treatment for 48 h, in contrast with the continually disrupted cell cycle distribution obtained after 48-hr treatment with HMBA (5 mM), DMF (l%),forskolin (50 pM), TPA (8 nM), or NaBu (1mM). To investigate further the inhibitory properties of DIAs, their effects on anchorage-independent growth of MCF-7 cells were examined (Table 2). MCF-7 cells growing in serum-free medium will not form colonies in soft agar without the addition of 4% FCS. Under these conditions, a marked decrease in colony-forming efficiency was also observed for all the inducers except forskolin.
Cell morphology MCF-7 cells grown on a plastic substratum in serumfree medium display typical epithelioid characteristics; cells are small and polygonal in shape. Their nuclei contain several nucleoli (Fig. 3A). As shown in Figure 3B, after 6-day exposure to RA, the cells form small colonies, but no change in cell morphology is noted. By contrast, the same period treatment with DMF (Fig. 3 0 , HMBA (Fig. 3D),NaBu (Fig. 3E), forskolin (Fig. 3F), or TPA (Fig. 3G) results in an enlargement of the cytoplasm. This phenomenon is more marked after TPA administration (Fig. 3G),while with forskolin only few cells are affected (Fig. 3F). At the ultrastructural level, MCF-7 are found to be polarized; their cytoplasm contains few granular endo lasmic reticulum and numerous stress fibers (Fig. 4Af):After HMBA, DMF, or NaBu treatment, cell polarity is preserved, the cytoplasm presents fewer stress fibers and numerous lipid droplets (Figs. 4B, 5B, 6A,B). Mitochondria are encircled with granular endoplasmic reticulum (Fig. 4B,C) in HMBA-treated cells. Numerous microvilli appear in NaBu and in forskolin-treated cells (Fig. 6A,D). Membrane-like material is observed in the intercellular space of DMF-treated cells (Fig. 5B). Immunofluorescence analysis of EMA antigen in MCF-7cells In an effort to find out whether exposure to DIAs induces a more differentiated status of MCF-7 cells, we measured the EMA content after 6 days of treatment (Fig. 7). This antigen, isolated from milk fat globule membranes, is thought to be an indicator of epithelial
osj,
r 2
4
6 Days
Fig. 1. Growth of MCF-7 cells in the absence or presence of 1 pM of RA, 1mM NaBu, 50 p M forskolin, 8 nMTPA, 5 mM HMBA, 1%DMF. Each cell number plotted is the mean of three determinations; SD was always 40%.
100 1
1001
40 20
-
L
0.16 0.8 1.6
8
12.5
25
50
40
B
20
5 z
0
m
0.1 0.5 1 RETiNOIC ACID IuM)
z q
0.25 0.5
1 BUTYRATE (mM1
2
2
J
Y A
40 20 O
1
2 3 4 HMBA ImM)
5
0.2 0.4 0.6 0.8 1 OMF 1%)
Fig. 2. Dose-dependent inhibition of MCF-7 cells growth by TPA, forskolin,RA, NaBu, HMBA, and DMF. Cells were seeded for 24 h and then treated with various concentrations of the indicated drugs for 6 days. Cell number is expressed as a percentage of control cell number. Each point represents the mean of two experiments in which the measurements were made in triplicate.
differentiation (Sloane and Ormerod, 1981). It is expressed in approximately 90% of untreated MCF-7 cells. No significant modification of EMA expression is observed when MCF-7 cell proliferation is blocked by the addition of insulin-free medium for 6 days. TPA (8 nM) led to a four-fold increase in EMA content (Table
165
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
TABLE 1. Effect of differentiation-inducingagents on the cell cycle distribution of MCF-7 cells'" Cell cycle distribution (%)
24 h
Treatment Cnntrnl - -.._ - --
Forskolin (50 pM) NaBu (1 mM) TPA (8 nM)
48 h
GO/G1
S
G2fM
GO/G1
S
56
24
20
56
23
21
69 71
16 8 8 6
15
59 69 69 73
19 11
22
21 21 24
71 70
10 5
20 21 22
'Cells were plated in 0.5%FCS-RPMI medium at 2.5 105 cells in 60-mm Petri dishes; 24 h after plating, the medium was changed to serum-freemedium,to which drugs were added. The cells were trypsinized.fixed, andstained for flow cytometry 24 or 48 h after the addition of drugs, as described under Materials and Methods. m e coefficientof variation (CV) of the G1 phase was always < 5%. 3Data are the means of triplicate measures (deviation standard values are always < 2%).The experiment reported is representative of two separate studies.
TABLE 2. Effect of differentiation-inducinga ents on anchorage-independentgrowth of MCF-7 cell8V2 Treatment
ment. A 6-day treatment of MCF-7 cells in insulin-free medium did not result in a decrease in ER concentration (data not shown).
100
f!nnt.rnl
HMBA (5 mM) DMF (1%) Forskolin (50 1M) TPA (8 nM) 'lo4 MCF-7 cells were plated in 4% FCS-RPMI medium with the indicated oncentrations of drugs. Colonies measuring >50 Frn were counted after 15 days. *Data are expressed as percentage of colonies-number obtained in control culture (250 + 80) and representthe mean SD of three separate experiments,each done in triplicate.
+
3). This effect, albeit to a lesser degree, was observed after treatment with NaBu, HMBA, DMF, or forskolin. However, RA did not induce EMA significantly in MCF-7 cells.
Estrogen receptor In order to determine whether the reduction in ER after exposure to TPA (Guilbaud et al., 1988) might be attributable to the induction of a more mature phenotype, we examined 17p-estradiol binding on high molarity cytosol extracts after treatment of MCF-7 cells with the different DIAs. We previously showed that there is an alteration in this biochemical parameter after exposure to TPA for 48 h. A concomitant dosedependent inhibition of growth is also observed. We therefore chose a 6-day exposure to the DIAs, which produced a maximal effect on proliferation expected. As shown in Figure 8, a significant reduction of ER bindin from 30% to 50% activity was detected when
The present results show that a more differentiated status of MCF-7 cells can be induced in vitro by maturational agents. Sodium butyrate (Abe and Kufe, 1984a,b), CAMP-like substances (Cho-chung et al., 1981), phorbol diesters (Osborne et al., 1981), polar solvents DMF, HMBA, and arabinoside C (ara C) (Abe and Kufe, 1986), and retinoids (Lacroix and Lippman, 1980, Lotan, 1979) have all been shown to inhibit the proliferation of MCF-7 cells. These compounds have also been shown to induce differentiation in cell systems such as leukemic or colon carcinoma cells (Reiss et al., 1986;Bloch, 1984).The MCF-7 cell line used in this study is able to grow in the total absence of serum. Under these conditions, we observed the same inhibitory effect of these substances as has been described for serum-dependent cells. NaBu, DMF, HMBA, forskolin, and TPA rapidly (24 h) block the entry of MCF-7 cells into the S phase, inducing an early inhibition of DNA synthesis. Prolonged exposure of MCF-7 cells to forskolin, NaBu, DMF, and HMBA causes them to accumulate in G1, but a significant fraction of MCF-7 cells have a G2+M DNA content. We previously showed that TPA induces a G1 block and a delayed passage through G2. In the present work, the kinetic studies (24 h and 48 h) do not permit the conclusion that DIAs induce a G2 delay. Despite a fall in S phase, however, we observe that the proportion of cells in G2 remains elevated (particularly after NaBu or forskolin treatment), suggesting a possible delay in this cell hase. Retinoids
not lead to a loss of their proliferative potential (La(5 mM), DMF (l%), or forskolin (50 pM). As previously croix and Liptman, 1980). An increase in RA concenshown for TPA, it corresponded to a decrease in the tration ( ~ 1 0 -M) led to a cytotoxic effect with a rapid number of binding sites with no change in the dissoci- decrease in the MCF-7 cell population. At the inhibiation constant of the ligand for ER as shown in Figure tory concentration of RA, there is an increase in G I 9 after HMBA or NaBu treatment. The respective Kd phase, but this effect is transient and rapidly reversvalues for ER in MCF-7 treated by forskolin and DMF ible. MCF-7 cells grow in a monolayer in the total were 1.35 nM and 1.2 nM. RA (1FM) did not induce a absence of serum but require the presence of serum (4% fall in ER content. In fact, a small but significant FCS) to form colonies in semisolid medium. Most DIAs increase in ER content was observed after RA treat- that inhibit the monolayer growth of MCF-7 cells also
166
GUILBAUD ET AL.
Fig. 3. Effect of DIAs on morphology of MCF-7 cells. Phase-contrast microscopy.A Untreated cells. B Treated for 6 days with RA 1 FM. C DMF 1%.D HMBA 5 mM. E: NaBu 1 mM. F Forskolin 50 pM. G TPA (8 nM).
Fig. 4. Ultrastructural MCF-7 morphology after HMBA treatment (5mM, 6 days). A Untreated cells, numerous stress fibers (F)are present in the cytoplasma. X8,OOO. B,C: Exposure to HMBA. Cells have developed lipid droplets (L)and rough endoplasmic reticulum around large mitochondrias (arrows). X5,OOO; ~ 2 4 , 0 0 0Tight . junctions are present (arrowhead).
168
GUILBAUD ET AL.
Fig. 5. Ultrastructural MCF-7 morphology after DMF treatment (l%, 6 days). A Note less stress fiber B Membrane-like material (MIin intercellular space. than in control cells (see Fig. 4A). ~10,000. ~ 5 , 5 0 0C: . Numerous lipid droplets (L). x 15,000. Tight junctions are always present (arrowhead).
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
treatment, cells present numerous microvilli (MV) at their apical Fig. 6 . After NaBu (A,B) and forskolin (C,D) surface, sealed by typical tight junctions (arrows).Lipid droplets (L)are accumulated after NaBu (B). Cells treated by forskolin have large mitochondria. A x 18600. B: x 17000.C x 16000.D x 19000.
169
170
GUILBAUD ET AL.
TABLE 3. Expression of EMA in MCF-7 cells','
Treatment Control HMBA (5 mM) DMF (1%) RA (1'pM) Forskolin (50 pM) NaBu (1 mM) TPA (8nM)
Insulin-free medium
120
EMA fluorescenceintensity (% of control) Exp 1 Exp 2 100 179 244 108 175 209 400 111
-
100 216 179 123 182 193 318
-
'MCF-7 cells were incubated for 6 days in the indicated cell culture conditions and assayed for indirect immunofluorescence. Cell fluorescence is expressed as a perFentage of the maximal value observed in corresponding untreated cells. LForeach experiment reported,the values are means of duplicate determinations.
Control
TPA (8nM)
Fig. 8. Effect of a 6-day treatment of MCF-7 cells with TPA (8nM), NaBu (1mM), HMBA (5 mM), DMF (l%), RA (1pM), and forskolin (50 p,M) on ER content of MCF-7 cells. ER content was measured at a saturating concentration of 10 nM [3Hl-17p-estradioland is expressed as a percentage (mean ?SD of three experiments) of ER content in MCF-7 cells (47.6 2 3.5 fmol/106cells).
\
Log Fluorescence Intensity
Fie. 7. Flow cvtometric analvsis of EMA in MCF-7 cells. To measure EMA induction, MCF-7 cells were grown for 6 days in the presence or absence of different drugs, as indicated. Cells were then labeled with rabbit control serum (-) or rabbit anti-EMA antibody (-),followed by FITC-conjugated goat antirabbit immunoglobulin. Cells were analyzed in a Coulter Epics fluorescence cell sorter.
Fig. 9. Effect of treatment of MCF-7 cells with HMBA (5 mM) or NaBu (1mM), for 6 days, on 17p-estradiol binding as a function of estradiol concentration in high molarity cell extract. Scatchard analyses of the binding data are presented.
decrease clonogenic ability of MCF-7 cells. Forskolin is an exception; MCF-7 cells are sensitive in anchoragedependent growth but are insensitive in anchorageindependent growth. In semisolid medium, only 2-3% of MCF-7 cells are able to form colonies. We hypothesize that these cells could be less sensitive to forskolin than is the all-cell population. However, forskolin, inefficient alone, is able to potentiate the inhibitory effect of RA (A. Valette, unpublished results). This fact indicates that MCF-7 cells still respond to forskolin in these conditions. MCF-7 cells are characterized by a high degree of polarity (Van Deurs et al., 1987). Resnicoff et al. (1987) reported that MCF-7 cells undergo unidirectional differentiation in vitro, with a concomitant loss of mul-
tipotency. These investigators demonstrated that this phenomenon could be enhanced by treating MCF-7 cells with NaBu. The morphological changes observed by Abe and Kufe (1984b) after NaBu treatment or by us after TPA treatment (Valette et al., 1987) argue in favor of the differentiation of MCF-7 cells. Ultrastructural studies show that the cytoplasm of MCF-7 cells treated with HMBA, DMF, or NaBu contains large number of lipid droplets. The measurement of [I4C]acetate incorporation into lipid confirms that these compound increase lipid synthesis on MCF-7 cells (data not shown). Lipid accumulation has been suggested to be a differentiated feature of breast cancer cell (Graham and Buick, 1988; Sidi et al., 1988). The results of the present study confirm and extend at the ultra-
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
structural level previous reports (Abe and Kufe, 1984a,b; Valette et al., 1987) indicating that chemical compounds could induce in MCF-7 cells the morphological characteristics of a more differentiated cell. The presence of microvilli and the abundance of lipid droplets argue in favor of a maturation process of MCF-7 cells along the secretory lineage. Moreover, RA, which partially inhibits MCF-7 proliferation, is unable to reproduce such morphological effects. TPA or NaBu increase the differentiation antigen DF3, whereas retinoids, DMF, and HMBA, which all inhibit MCF-7 proliferation lead t o a reduction in DF3 (Abe and Kufe, 1986). In the present study, we measured the expression of another milk fat globule antigen, epithelial membrane antigen (EMA). Although this antigen is not confined to mammary gland, it is useful in histopathology and cytology as an indicator of epithelial differentiation (Sloane and Ormerod, 1981; To et al., 1981). Our cytometric studies confirm previous immunological studies indicating that breast cancer cells express EMA (Sloane and Ormerod, 1981). Treatment of MCF-7 cells with TPA, DMF, HMBA, or forskolin was found to increase significantly the expression of this antigen. With RA, we observed only a slight increase in the expression of EMA. When cell proliferation was blocked by the addition of insulin-free medium for 6 days, we did not observe an increase of EMA expression despite the blockade of MCF-7 cells in the G1 phase of the cell cycle (Gross et al,, 1984). Thus, the change in EMA expression is not simply a result of cytostasis. Although the increase in MCF-7 protein content (1.8-fold)induced only by TPA treatment, this could not entirely account for EMA antigen induction (4-fold). The expression of the ER in breast cancer cells is one of the most important characteristics of the differentiation status of the breast tumor. The proliferation of MCF-7 cells is controlled either directly or indirectly by ER via secretion of growth factors (Dickson and Lippman, 1987). In a previous study, we found that phorbol diesters decrease ER content (Guilbaud et al., 1988);we were therefore interested to find out whether this decrease was the consequence of a direct interaction between PKC and ER or a more general feature of breast cancer differentiation. The fact that there is a decrease in ER expression after treatment of MCF-7 cells with NaBu and CAMP-like drugs (Stevens et al., 1984; Cho-chung et al., 1981) is in favor of the second possibility. All the DIAs that induced maturation along the secretory cell lineage significantly decreased ER content without modifying estrogen affinity. By contrast, after treatment with RA, we observed a slight increase in ER content, in agreement with the findings of Butler (1989). This did not appear to be a simple consequence of the inhibition of growth of MCF-7 cells in the G1 phase of the cell cycle, since the addition of insulin-free medium did not modify the ER content, although it blocked G1-S transition (Gross et al., 1984). Moreover, Jakesz et al. (1984) reported that ER synthesis takes place mainly in the G1 and G2 phases. The specific reduction of ER by several DIAs observed in the present study confirms the results of Stevens et al. (1984) and Graham and Buick (1988) and argues in favor of a relationship between the maturation of
171
MCF-7 cells and the decrease in the expression of ER. The expression of ER is characteristic of a more differentiated status of breast carcinoma cells. ER concentrations in normal mammary epithelial cells are lower than in benign breast tumor, while higher concentrations are found in malignant tumor (Isotalo et al., 1983). Since the carcinogenic process in breast is accompanied by an elevation of ER number (Nenci et al., 1988), the decrease in ER content may represent a differentiation criterion. The results of the present study define some characteristics of the maturation of MCF-7 cells. A rapid block of the G 1 S transition with a complete inhibition of DNA synthesis is observed with the DIAs, which induce a maturational rocess (change in morphology and expression of E A). Inhibition of the proliferation of MCF-7 cells seems to be the primary effect of DMF, HMBA, TPA, NaBu, and forskolin. RA, which induces a decrease in growth rate, did not modify EMA expression and MCF-7 morphology, indicating that RA is not a DIA in MCF-7 cells. The major finding is the relationship between the decrease in ER expression and the induction of a more differentiated status of MCF-7 cells. The mechanism by which DMF, HMBA, TPA, and NaBu induce differentiation and decrease ER remains unclear. It now is of interest to determine the molecular mechanisms leading to the ER decrease observed during the MCF-7 maturation process. After treatment of MCF-7 cells with TPA, induction of the progesterone receptor by the ER is as effective as in untreated control cells (Guilbaud et al., 1988), suggesting that the remaining ER is active. ACKNOWLEDGMENTS This work was supported by Federation Nationale des Centres de Lutte Contre le Cancer. We are grateful to J.C. Hendrick and J. Colette for generously providing antibody to epithelial membrane antigen and to F. Roubinet and G. Cassar for their efficient assistance in performing flow cytometry experiments. LITERATURE CITED
s
Abe, M., and Kufe, D.W. (1984a) Sodium butyrate induction of milk related antigens in human MCF-7 breast carcinoma cells. Cancer Res., 44:45144577. Abe, M., and Kufe, D.W. (1984b) Effect of sodium butyrate on human breast carcinoma (MCF-7) cellular proliferation, morphology, and CEA production. Breast Cancer Res. Treatm., 41269-274. Abe, M., and Kufe, D.W. (1986) Effect of maturational agents on expression and secretion of two partially characterized high molecular weight milk related glycoproteins in MCF-7 breast carcinoma cells. J. Cell. Physiol., 126:12&132. Bloch, A. (1984)Induced cell differentiation in cancer therapy. Cancer Treatm. Rep., 681199-205. Burton, K. (1956) A study of the conditions and the mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem. J., 621315423. Butler, W.B. (1989) Responses to retinoic acid of tamoxifen-sensitive and tamoxifen-resistant sublines of the human cancer cell line MCF-7. Proc. Am. Assoc. Cancer Res., 30:305. Calabresi, P., Dexter, D.L., and Hepner, G.H. (1979) Clinical and pharmacological implications of cancer cell differentiation and heterogeneity. Biochem. Pharmacol., 2811933-1941. Cho-chung, Y.S., Clair, T., Bodwin, J.S., and Berghoffer, B. (1981) Growth arrest and morphological change of human breast cancer cells by dibutyryl cyclic AMP and L-arginine. Science, 214177-79. Chouvet, C., Vicard, E., Devonec, M., and Saez, S. (1986) 1,25Dihydroxyvitamin D3 inhibitory effect on the growth of two human
172
GUILBAUD ET AL
breast cancer cell lines (MCF-7, BT20). J. Steroid Biochem., 24:373376. Darbon, J.M., Valette, A,, and Bayard, F. (1986)Phorbol esters inhibit the proliferation of MCF-7 cells: Possible implication of protein kinase C. Biochem. Pharmacol., 35:2683-2686. Dickson, R.B., and Lippman, M.E. (1987) Estrogenic regulation of growth and polypeptides growth factor secretion in human breast carcinoma. Endocrine Rev., 8:29-43. Falette, N., Frappart, L., Lefebvre, M.F., and Saez, S. (1988) 1.25(OH)ZVit D3 treatments results in morphological changes, growth reduction and increase in epidermal growth factor receptors in a human breast adenocarcinoma cell line (BT 20). Proc. Am. Assoc. Cancer Res., 2951. Freshnev. R.I. (1985) Induction of differentiation in neodastic cells. Anticancer Res., 5:111-130. Graham, K.A.. and Buick, R.N. (1988) Sodium butyrate induces differentiation in breast cancer cell lines expressing the estrogen receptor. J. Cell. Physiol., 136r63-71. Gross, G.E., Boldt, D.H., and Osborne, C.K. (1984) Perturbation by insulin of human breast cancer cell cycle kinetics. Cancer Res., 44:3570-3575. Guilbaud, N., Pichon, M.F., Faye, J.C., Bayard, F., and Valette, A. (1988) Modulation of estrogen receptors by phorbol diesters in human breast cancer MCF-7 cell line. Mol. Cell. Endocrinol., 56:157-163. Isotalo, H., Tryggvason, K., Vierikko, P., Kauppila, A., and Vihko, R. (1983) Plasminogen activators and steroid receptor concentrations in normal, benign, and malignant breast and ovarian tissues. Anticancer Res., 3:331336. Jakesz, R., Smith, C.A.,Aitken, S., Huff, K., Schuette, W., Schackney, S., and Lippman, M. (1984) Influence of cell proliferation and cell cycle phase on expression of estrogen receptor in MCF-7 breast cancer cells. Cancer Res., 44:619-625. Jozan, S., Tournier, J.F., Tauber, J.P., and Bayard, F. (1982) Adap tation of the human breast cancer cell line MCF-7 to serum free medium culture on extracellular matrix. Biochem. Biophys. Res. Commun., 107:1566-1570. Knabbe, C., Lippman, M.E., Wakefield, L.M., Flanders, K.C., Kasid, A., Derynk, R., and Dickson, R.B. (1987) Evidence that transforming growth factor-p is a hormonally regulated negative growth factor in human breast cancer cell. Cell, 48:417-428. Lacroix, A., and Lippman, M.E. (1980) Binding of retinoids to human breast cancer cell lines and their effects on cell growth. J. Clin. Invest., 65:58&591. Langdon, S.P., Hawkes, M.M., Hay, F.G., Lowrie, S.S., Schol, D.J., Hilgers, J., Leonard, R.C.F., and Smyth, J.F. (1988) Effect of sodium butyrate and other differentiation inducers on poorly differentiated human ovarian carcinoma cell lines. Cancer Res., 48:6161-6165. Lotan, R. (1979) Different susceptibilities of human melanoma and breast carcinoma cell lines to retinoic acid induced growth inhibition. Cancer Res., 39:101&1019. Lowry, D.H., Rosebrough, N.H., Farr, A.L., and Randall, R.J. (1951) ~
Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193:265-275. Nenci, I., Machetti, E., and Querzoli, P. (1988) Comment on human mammary preneoplasia. The estrogen receptor-promotion hypothesis. J. Steroid Biochem., 30:105-106. Osborne, C.K., Hamilton, B., Nover, M., and Ziegler, J . (1981) Antagonism between epidermal growth factor and Phorbol ester tumor promoters in human breast cancer cell. J . Clin. Invest., 67:943-95 1. Reiss, M., Gamba-Vitalo, C., and Sartorelli, A.C. (1986) Induction of tumor cell differentiation as therapeutic approach: Preclinical models for hematopoietic and solid neoplasms. Cancer Treatm. Rep., 70:20 1-2 18. Resnicoff, M., and Medrano, E.E. (1989) Growth factors and hormones which affect survival, growth, and differentiation of the MCF-7 stem cells and their descendants. Exp. Cell Res., 181:116-125. Resnicoff, M., Medrano, E.E., Podhajcer, O.J., Bravo, A.I., Bover, L., and Mordoh, J. (1987) Subpopulations of MCF-7 cells separated by percoll gradient centrifugation: A model to analyse the heterogeneity of human breast cancer. Proc. Natl. Acad. Sci. USA, 84:72957299. Sidi, Y., Panet, C., Wasserman, L., Cyjon, A., Novogrodsky, A., and Nordenberg, J. (1988) Growth inhibition and induction of phenotypic alteration in MCF-7 breast cancer cells by IMP deshydrogenase inhibitor. Br. J. Cancer, 58:61-63. Sloane, J.P., and Ormerod, M.G. (1981) Distribution of epithelial membrane antigen in normal and neoplastic tissues and its value in diagnostic tumor pathology. Cancer, 47:1786-1795. Stevens, M.S., Aliabadi, Z., and Moore, M.R. (1984) Associated effects of sodium butyrate on histone acetylation and estrogen receptor in the human breast cancer cell line. Biochem. Biophys. Res. Commun., 119:132-138. To, A., Coleman, D.V., Dearnaly, J.P., Ormerod, M.G., Steele, K., and Neville, A.M. (1981) Use ofantisera to epithelial membrane antigen for the cytodiagnosis of malignancy in serous effusions. J . Clin. Pathol., 34:132&1332. Tsao, D., Morita, A., Bella, A.J.R., Liu, P., and Kim, Y.S. (1982) Differential effects of sodium butyrate, dimethyl sulfoxide, and retinoic acid on membrane associated antigen, enzymes and glycoproteins of human rectal adenocarcinoma cells. Cancer Res., 42:1052-1058. VAlette, A., Gas, N. Josan, S., Roubinet, F., Dupont, M.A., and Bayard, F. (1987) Influence of 12-O-tetradecanoylphorbol-13acetate on proliferation and maturation of human breast cancer cells (MCF-7): Relationship to cell cycle events. Cancer Res., 47:1615-1623. Van Deurs, B., Zou, Z., Briand, P., Balslev, Y., and Petersen, O.W. (1987) Epithelial membrane polarity: A stable, differentiated feature of an established human breast carcinoma cell line MCF-7. J. Histochem. Cytochem., 33:461-469.
Effects of Differentiation-Inducing Agents on Maturation of Human MCF-7 Breast Cancer CelIs NICOLAS F. GUILBAUD,* NICOLE GAS, MARIE ANGE DUPONT, A N D ANNIE VALETTE laboratoire de Pharmacologie et Toxicologie Fondamentales, CNRS, 3 1077 Toulouse CEDEX, France (N.F.G., A.V.); Centre de Recherche de Biochimie et de Cenetique Cellulaires, CNRS, 3 1062 Toulouse CEDEX, France (N.C., M.A.D.)
The effectsof the differentiation inducing agents (DIAs), sodium butyrate (NaBu), retinoic acid (RA), dimethylformamide (DMF), hexamethylene bisacetamide (HMBA), forskolin, and 12-O-tetradecanoylphorbol-13-acetate (TPA), on the growth, morphology, and estrogen receptor (ER) content and epithelial membrane antigen (EMA) expression on a serumless human breast cancer cell line (MCF-7) were compared. All these agents reversibly caused a concentration-dependent growth inhibition in monolayers and markedly reduced colony-formingefficiency in soft agar. A twofold increase in doubling time was obtained with RA (1 FM), but cell replication ceased with NaBu (1 mM), forskolin (50 (IM), DMF (1 %), H M B A (5 mM), and TPA (8 nM). Total growth arrest induced by these last compounds was preceded by an accumulation of cells in GO/Gl phase observed at 24 h by flow cytometry and accompanied by a change in cell morphology as seen by light and electronic microscopy. An increase in cell volume and the presence of lipid droplets was noted in treated cells that were spread out, as compared with controls. The acquisition of a more mature phenotype was confirmed by an increased expression of EMA monitored by flow cytometry. A specific reduction in the number of ER without any constant dissociation (Kd) modification was also observed after treatment with the 5 DIAs. No modification of morphological or biochemical characteristics, including EMA expression and ER binding, were observed for RA (1 p,M)-treated cells. All these results suggest that induction of a more differentiated phenotype is associated with a block in G1 cell cycle phase, resulting in total growth arrest. Apparently, RA (1 FM)-treated cells did not fulfill these criteria, since only a slight accumulation in G1 and a slowed growth rate were evaluated. Since cancer can be regarded as a disorder of cell differentiation, the use of drugs that induce the matu: ration of tumor cells has been envisaged as a treatment strategy. However, evaluation of such drugs requires model systems that are representative of the most common clinical tumor types. Although there are several well-characterized models for leukemic cell differentiation (Reiss et al., 1986), there are few in vitro systems for the study of differentiation of malignant human epithelial cells. However, several reports indicate that maturation of adenocarcinoma cells can be induced chemically both in vitro and in vivo (Calabresi et al., 1979; Freshney, 1985; Tsao et al., 1982; Langdon et al., 1988). Estrogen receptor-positive ER(+) and estrogen receptor negative ER(-) breast cancer cell lines provide in vitro models for investigations on hormonally responsive and unresponsive breast cancer (Dickson and Lippman, 1987).The effects of differentiation-inducing agents (DIAs) have been studied on a panel of ER(+) or ER(-) breast cancer cells. DIAs include retinoids (vit A and its synthetic analogues), derivatives of vit D, cyclic adenosine monophosphate (CAMP)and its analogues, and certain small polar 0 1990 WILEY-LISS, INC.
planar molecules, such as dimethylformamide (DMF) and hexamethylene bisacetamide (HMBA).These compounds inhibit cell proliferation in ER( -) breast cancer cell lines, whereas their antiproliferative effects are more often associated with a maturation process in E R ( t ) breast cancer cells. However, vit D3 has been reported to induce a maturation process in ER( -1 BT20 cells (Falette et al., 1988). The maturation process of ER(+) cells is characterized by morphological change and an increase in the expression of milk fat globule antigens, such as DF3 (Abe and Kufe, 1984a). It was shown that tumor promoting agents such as 12-O-tetradecanoylphorbol-13-acetate (TPA) inhibit the proliferation of MCF-7 cells (Osborne et al., 1981; Darbon et al., 1986) and that this inhibition of proliferation results from a specific block in the G1 phase of the cell cycle (Valette et al., 1987). After 2 days of TPA treatment, there is a specific reduction in ER number, accompanied by an induction of the morphological Received December 29, 1989; accepted June 11, 1990. *To whom reprint requestsicorrespondence should be addressed.
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
characteristics of a secretory cell type, a more mature phenotype of MCF-7 cells (Valette et al., 1987; Guilbaud et al., 1988). There is a good correlation between the potency of these phorbol ester derivatives to bind to a high affinity receptor (probably C kinase, PKC) and their ability to inhibit MCF-7 proliferation (Darbon et al., 1986) and reduce ER concentration (Guilbaud et al., 1988). Moreover, the ER of MCF-7 : RPh4 cells (which are insensitive to the antiproliferative and morphological effects of TPA) is not affected by exposure to TPA. Several possibilities could account for this decrease in ER content. ER could be a specific substrate for PKC. The decrease in estrogen receptor could be due to the decrease in growth rate observed after TPA treatment of MCF-7 cells. Yet another explanation is that there is a relationship between the decrease in expression of ER and the maturation process. In order to address this issue, we compared the effects of various DIAs with those induced by TPA. We attem ted to establish relationships between the antiproli erative effect (cell cycle progression), the maturation effect (change in morphology, expression of milk fat globule antigen) and change in ER content after treatment of MCF-7 cells with DIAs that act by different molecular mechanisms.
P
163
underlayer in 4% FCS growth medium. The number of colonies present after 15 days were determined after staining with 3,4,5-dimethylthiazo1-2-yl-2,5-diphenyltetrazolium bromide (MTT) using a colony analyzer. Colonies >50 Fm in diameter were scored as positive. Flow cytometry To determine the cell cycle distribution of treated MCF-7 cells, the cells were removed from the culture surface with trypsin, fixed, stained with propidium iodide, and analyzed using a Coulter Epics fluorescence-activated cell sorter as previously described (Valette et al., 1987). To measure EMA induction, MCF-7 cells were grown for 6 days in the presence or absence of different drugs as indicated. Cells were washed once with phosphate-buffered saline (PBS) and then incubated with rabbit antihuman EMA antibody or rabbit control serum (diluted 1 : 200) for 12 h at 4°C. Cells were then washed three times with PBS and incubated with FITC-conjugated goat antirabbit immunoglobulin diluted 1 : 50 in PBS for 1 h at 4°C. Cells were then washed three times with PBS and fixed in 70% ethanol PBS and analyzed in a Coulter Epics fluorescenceactivated cell sorter. The peak fluorescence was taken as the fluorescence channel with the highest number of cells. Peak fluorescence was then converted from channel units of logarithmic fluorescence to linear units. Electron microscopy After exposure to different drugs, MCF-7 cells were fixed for 60 min with 3% glutaraldehyde in 0.1 M sodium cacodylate buffer and postfixed in 1%osmium tetroxide in 0.1 M sodium cacodylate buffer, dehydrated in graded ethanol, and embedded in Epon 812. Sections were stained with uranyl acetate and lead citrate, and examined in a JEOL 1200 EX electron microscope. Estrogen receptor measurements All procedures were performed at 4°C as described above. The cells were harvested by scraping with a rubber policeman and pelleted by centrifugation at 800g for 10min. The pellets were washed with 0.9% NaC1, resuspended in 20 mM TrisHCl buffer containing 20 mM sodium molybdate-10 mM monothioglycerol pH 7.4, at 24°C. The cells were sonicated twice with a SCMA Pons Machine type 20200 SV, at a 1.25-W setting for 5 s each, with a 30-s interval for cooling. The homogenate was adjusted to 0.4 M KCl and centrifuged at 105,OOOgfor 60 min. The protein concentration of the supernatant was determined by the method of Lowry et al. (1951), with serum albumin as standard. The DNA content of the pellet was measured by the method of Burton (1956). For measurements of ER, samples of 100 p1 of supernatant were incubated at 4°C in triplicate in the presence of 0.1-20 nM [3Hl-17p-estradiol. Nonspecific binding was determined by parallel incubation in the presence of 1 pM of nonradioactive 17pestradiol. Bound and unbound ligand were separated by LH20 chromatography.
MATERIALS AND METHODS Chemicals TPA, forskolin, retinoic acid (RA) (all trans), hexamethylene bisacetamide (HMBA),dimethylformamide (DMF),n-butyric acid (sodium salt), 17p-estradiol were obtained from Sigma Chemical Co., [3Hl-17p-estradiol (sp. act. 80 Cilmmol) was provided by Amersham. Antisera to epithelial membrane antigen (EMA) were kindly provided from Dr. Hendrick and Dr. Collette (Institute of Pathology, University of Liege). Cells MCF-7 cells previously adapted to grow in serumfree medium (Jozan et al., 1982) were maintained in RPMI 1640 medium supplemented with sodium bicarbonate (2 g/L), 2 mM glutamine, 1 pM insulin, and 0.1 FM transferrin in humidified 5% C02/95% air at 37°C. Growth curves in monolayer cultures MCF-7 cells growing in logarithmic phase were plated in Petri dishes at a density of lo5 cellddish in RPMI medium containing 0.5% fetal calf serum (FCS). The following day, the medium was replaced by RPMI serum-free medium containing the different drugs. The culture medium was changed every 3 days. After treating the cells with 0.05% try sin-0.02% EDTA for 1min, the trypsin was removeg and the cells were harvested in RPMI medium containing 0.5% FCS. Cell numbers were determined in a Coulter counter (Coultronics). Anchorage-independent assay RESULTS Clonal growth assay were performed as described by Effects of inducers on cell growth Knabbe et al. (1987). Briefly, 10,000 cells/35-mm dish in RPMI medium supplemented with 4% FCS, 0.4% The effects of NaBu, DMF, forskolin, and RA on the agar, and different drugs were plated on a 0.6% agarose growth of serum-free MCF-7 are illustrated in Figure 1.
164
GUILBAUD ET AL.
NaBu (1 mM), DMF (l%), HMBA (5mM),TPA (8 nM), and forskolin (50 KM) block MCF-7 proliferation after 6-day treatment. Under these conditions, 90-95% of MCF-7 cells remained viable, as assessed by Trypan blue dye exclusion. A decrease in growth rate was observed after exposure of MCF-7 cells to RA (1p.M) for 6 days, The antiproliferative effects of these compounds were dose dependent (Fig. 2). The maximum inhibition was about 50% for RA concentration of 10-6M; for concentrations above M, significant toxicity was observed (data not shown). Incubation of MCF-7 with the different inducers for 24 h led to an increase in the number of cells in the Gl phase of the cell cycle and a concomitant fall of the number of cells in S phase (Table 1). RA (1 pM) treatment induced the least accumulation in the G1 phase (69% vs. 56% in control cells) and the smallest decrease in the percentage of S-phase cells (16%vs. 24% in control cells) at 24h. This effect of RA almost completely disappeared after treatment for 48 h, in contrast with the continually disrupted cell cycle distribution obtained after 48-hr treatment with HMBA (5 mM), DMF (l%),forskolin (50 pM), TPA (8 nM), or NaBu (1mM). To investigate further the inhibitory properties of DIAs, their effects on anchorage-independent growth of MCF-7 cells were examined (Table 2). MCF-7 cells growing in serum-free medium will not form colonies in soft agar without the addition of 4% FCS. Under these conditions, a marked decrease in colony-forming efficiency was also observed for all the inducers except forskolin.
Cell morphology MCF-7 cells grown on a plastic substratum in serumfree medium display typical epithelioid characteristics; cells are small and polygonal in shape. Their nuclei contain several nucleoli (Fig. 3A). As shown in Figure 3B, after 6-day exposure to RA, the cells form small colonies, but no change in cell morphology is noted. By contrast, the same period treatment with DMF (Fig. 3 0 , HMBA (Fig. 3D),NaBu (Fig. 3E), forskolin (Fig. 3F), or TPA (Fig. 3G) results in an enlargement of the cytoplasm. This phenomenon is more marked after TPA administration (Fig. 3G),while with forskolin only few cells are affected (Fig. 3F). At the ultrastructural level, MCF-7 are found to be polarized; their cytoplasm contains few granular endo lasmic reticulum and numerous stress fibers (Fig. 4Af):After HMBA, DMF, or NaBu treatment, cell polarity is preserved, the cytoplasm presents fewer stress fibers and numerous lipid droplets (Figs. 4B, 5B, 6A,B). Mitochondria are encircled with granular endoplasmic reticulum (Fig. 4B,C) in HMBA-treated cells. Numerous microvilli appear in NaBu and in forskolin-treated cells (Fig. 6A,D). Membrane-like material is observed in the intercellular space of DMF-treated cells (Fig. 5B). Immunofluorescence analysis of EMA antigen in MCF-7cells In an effort to find out whether exposure to DIAs induces a more differentiated status of MCF-7 cells, we measured the EMA content after 6 days of treatment (Fig. 7). This antigen, isolated from milk fat globule membranes, is thought to be an indicator of epithelial
osj,
r 2
4
6 Days
Fig. 1. Growth of MCF-7 cells in the absence or presence of 1 pM of RA, 1mM NaBu, 50 p M forskolin, 8 nMTPA, 5 mM HMBA, 1%DMF. Each cell number plotted is the mean of three determinations; SD was always 40%.
100 1
1001
40 20
-
L
0.16 0.8 1.6
8
12.5
25
50
40
B
20
5 z
0
m
0.1 0.5 1 RETiNOIC ACID IuM)
z q
0.25 0.5
1 BUTYRATE (mM1
2
2
J
Y A
40 20 O
1
2 3 4 HMBA ImM)
5
0.2 0.4 0.6 0.8 1 OMF 1%)
Fig. 2. Dose-dependent inhibition of MCF-7 cells growth by TPA, forskolin,RA, NaBu, HMBA, and DMF. Cells were seeded for 24 h and then treated with various concentrations of the indicated drugs for 6 days. Cell number is expressed as a percentage of control cell number. Each point represents the mean of two experiments in which the measurements were made in triplicate.
differentiation (Sloane and Ormerod, 1981). It is expressed in approximately 90% of untreated MCF-7 cells. No significant modification of EMA expression is observed when MCF-7 cell proliferation is blocked by the addition of insulin-free medium for 6 days. TPA (8 nM) led to a four-fold increase in EMA content (Table
165
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
TABLE 1. Effect of differentiation-inducingagents on the cell cycle distribution of MCF-7 cells'" Cell cycle distribution (%)
24 h
Treatment Cnntrnl - -.._ - --
Forskolin (50 pM) NaBu (1 mM) TPA (8 nM)
48 h
GO/G1
S
G2fM
GO/G1
S
56
24
20
56
23
21
69 71
16 8 8 6
15
59 69 69 73
19 11
22
21 21 24
71 70
10 5
20 21 22
'Cells were plated in 0.5%FCS-RPMI medium at 2.5 105 cells in 60-mm Petri dishes; 24 h after plating, the medium was changed to serum-freemedium,to which drugs were added. The cells were trypsinized.fixed, andstained for flow cytometry 24 or 48 h after the addition of drugs, as described under Materials and Methods. m e coefficientof variation (CV) of the G1 phase was always < 5%. 3Data are the means of triplicate measures (deviation standard values are always < 2%).The experiment reported is representative of two separate studies.
TABLE 2. Effect of differentiation-inducinga ents on anchorage-independentgrowth of MCF-7 cell8V2 Treatment
ment. A 6-day treatment of MCF-7 cells in insulin-free medium did not result in a decrease in ER concentration (data not shown).
100
f!nnt.rnl
HMBA (5 mM) DMF (1%) Forskolin (50 1M) TPA (8 nM) 'lo4 MCF-7 cells were plated in 4% FCS-RPMI medium with the indicated oncentrations of drugs. Colonies measuring >50 Frn were counted after 15 days. *Data are expressed as percentage of colonies-number obtained in control culture (250 + 80) and representthe mean SD of three separate experiments,each done in triplicate.
+
3). This effect, albeit to a lesser degree, was observed after treatment with NaBu, HMBA, DMF, or forskolin. However, RA did not induce EMA significantly in MCF-7 cells.
Estrogen receptor In order to determine whether the reduction in ER after exposure to TPA (Guilbaud et al., 1988) might be attributable to the induction of a more mature phenotype, we examined 17p-estradiol binding on high molarity cytosol extracts after treatment of MCF-7 cells with the different DIAs. We previously showed that there is an alteration in this biochemical parameter after exposure to TPA for 48 h. A concomitant dosedependent inhibition of growth is also observed. We therefore chose a 6-day exposure to the DIAs, which produced a maximal effect on proliferation expected. As shown in Figure 8, a significant reduction of ER bindin from 30% to 50% activity was detected when
The present results show that a more differentiated status of MCF-7 cells can be induced in vitro by maturational agents. Sodium butyrate (Abe and Kufe, 1984a,b), CAMP-like substances (Cho-chung et al., 1981), phorbol diesters (Osborne et al., 1981), polar solvents DMF, HMBA, and arabinoside C (ara C) (Abe and Kufe, 1986), and retinoids (Lacroix and Lippman, 1980, Lotan, 1979) have all been shown to inhibit the proliferation of MCF-7 cells. These compounds have also been shown to induce differentiation in cell systems such as leukemic or colon carcinoma cells (Reiss et al., 1986;Bloch, 1984).The MCF-7 cell line used in this study is able to grow in the total absence of serum. Under these conditions, we observed the same inhibitory effect of these substances as has been described for serum-dependent cells. NaBu, DMF, HMBA, forskolin, and TPA rapidly (24 h) block the entry of MCF-7 cells into the S phase, inducing an early inhibition of DNA synthesis. Prolonged exposure of MCF-7 cells to forskolin, NaBu, DMF, and HMBA causes them to accumulate in G1, but a significant fraction of MCF-7 cells have a G2+M DNA content. We previously showed that TPA induces a G1 block and a delayed passage through G2. In the present work, the kinetic studies (24 h and 48 h) do not permit the conclusion that DIAs induce a G2 delay. Despite a fall in S phase, however, we observe that the proportion of cells in G2 remains elevated (particularly after NaBu or forskolin treatment), suggesting a possible delay in this cell hase. Retinoids
not lead to a loss of their proliferative potential (La(5 mM), DMF (l%), or forskolin (50 pM). As previously croix and Liptman, 1980). An increase in RA concenshown for TPA, it corresponded to a decrease in the tration ( ~ 1 0 -M) led to a cytotoxic effect with a rapid number of binding sites with no change in the dissoci- decrease in the MCF-7 cell population. At the inhibiation constant of the ligand for ER as shown in Figure tory concentration of RA, there is an increase in G I 9 after HMBA or NaBu treatment. The respective Kd phase, but this effect is transient and rapidly reversvalues for ER in MCF-7 treated by forskolin and DMF ible. MCF-7 cells grow in a monolayer in the total were 1.35 nM and 1.2 nM. RA (1FM) did not induce a absence of serum but require the presence of serum (4% fall in ER content. In fact, a small but significant FCS) to form colonies in semisolid medium. Most DIAs increase in ER content was observed after RA treat- that inhibit the monolayer growth of MCF-7 cells also
166
GUILBAUD ET AL.
Fig. 3. Effect of DIAs on morphology of MCF-7 cells. Phase-contrast microscopy.A Untreated cells. B Treated for 6 days with RA 1 FM. C DMF 1%.D HMBA 5 mM. E: NaBu 1 mM. F Forskolin 50 pM. G TPA (8 nM).
Fig. 4. Ultrastructural MCF-7 morphology after HMBA treatment (5mM, 6 days). A Untreated cells, numerous stress fibers (F)are present in the cytoplasma. X8,OOO. B,C: Exposure to HMBA. Cells have developed lipid droplets (L)and rough endoplasmic reticulum around large mitochondrias (arrows). X5,OOO; ~ 2 4 , 0 0 0Tight . junctions are present (arrowhead).
168
GUILBAUD ET AL.
Fig. 5. Ultrastructural MCF-7 morphology after DMF treatment (l%, 6 days). A Note less stress fiber B Membrane-like material (MIin intercellular space. than in control cells (see Fig. 4A). ~10,000. ~ 5 , 5 0 0C: . Numerous lipid droplets (L). x 15,000. Tight junctions are always present (arrowhead).
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
treatment, cells present numerous microvilli (MV) at their apical Fig. 6 . After NaBu (A,B) and forskolin (C,D) surface, sealed by typical tight junctions (arrows).Lipid droplets (L)are accumulated after NaBu (B). Cells treated by forskolin have large mitochondria. A x 18600. B: x 17000.C x 16000.D x 19000.
169
170
GUILBAUD ET AL.
TABLE 3. Expression of EMA in MCF-7 cells','
Treatment Control HMBA (5 mM) DMF (1%) RA (1'pM) Forskolin (50 pM) NaBu (1 mM) TPA (8nM)
Insulin-free medium
120
EMA fluorescenceintensity (% of control) Exp 1 Exp 2 100 179 244 108 175 209 400 111
-
100 216 179 123 182 193 318
-
'MCF-7 cells were incubated for 6 days in the indicated cell culture conditions and assayed for indirect immunofluorescence. Cell fluorescence is expressed as a perFentage of the maximal value observed in corresponding untreated cells. LForeach experiment reported,the values are means of duplicate determinations.
Control
TPA (8nM)
Fig. 8. Effect of a 6-day treatment of MCF-7 cells with TPA (8nM), NaBu (1mM), HMBA (5 mM), DMF (l%), RA (1pM), and forskolin (50 p,M) on ER content of MCF-7 cells. ER content was measured at a saturating concentration of 10 nM [3Hl-17p-estradioland is expressed as a percentage (mean ?SD of three experiments) of ER content in MCF-7 cells (47.6 2 3.5 fmol/106cells).
\
Log Fluorescence Intensity
Fie. 7. Flow cvtometric analvsis of EMA in MCF-7 cells. To measure EMA induction, MCF-7 cells were grown for 6 days in the presence or absence of different drugs, as indicated. Cells were then labeled with rabbit control serum (-) or rabbit anti-EMA antibody (-),followed by FITC-conjugated goat antirabbit immunoglobulin. Cells were analyzed in a Coulter Epics fluorescence cell sorter.
Fig. 9. Effect of treatment of MCF-7 cells with HMBA (5 mM) or NaBu (1mM), for 6 days, on 17p-estradiol binding as a function of estradiol concentration in high molarity cell extract. Scatchard analyses of the binding data are presented.
decrease clonogenic ability of MCF-7 cells. Forskolin is an exception; MCF-7 cells are sensitive in anchoragedependent growth but are insensitive in anchorageindependent growth. In semisolid medium, only 2-3% of MCF-7 cells are able to form colonies. We hypothesize that these cells could be less sensitive to forskolin than is the all-cell population. However, forskolin, inefficient alone, is able to potentiate the inhibitory effect of RA (A. Valette, unpublished results). This fact indicates that MCF-7 cells still respond to forskolin in these conditions. MCF-7 cells are characterized by a high degree of polarity (Van Deurs et al., 1987). Resnicoff et al. (1987) reported that MCF-7 cells undergo unidirectional differentiation in vitro, with a concomitant loss of mul-
tipotency. These investigators demonstrated that this phenomenon could be enhanced by treating MCF-7 cells with NaBu. The morphological changes observed by Abe and Kufe (1984b) after NaBu treatment or by us after TPA treatment (Valette et al., 1987) argue in favor of the differentiation of MCF-7 cells. Ultrastructural studies show that the cytoplasm of MCF-7 cells treated with HMBA, DMF, or NaBu contains large number of lipid droplets. The measurement of [I4C]acetate incorporation into lipid confirms that these compound increase lipid synthesis on MCF-7 cells (data not shown). Lipid accumulation has been suggested to be a differentiated feature of breast cancer cell (Graham and Buick, 1988; Sidi et al., 1988). The results of the present study confirm and extend at the ultra-
DIFFERENTIATING AGENTS ON BREAST CANCER CELLS
structural level previous reports (Abe and Kufe, 1984a,b; Valette et al., 1987) indicating that chemical compounds could induce in MCF-7 cells the morphological characteristics of a more differentiated cell. The presence of microvilli and the abundance of lipid droplets argue in favor of a maturation process of MCF-7 cells along the secretory lineage. Moreover, RA, which partially inhibits MCF-7 proliferation, is unable to reproduce such morphological effects. TPA or NaBu increase the differentiation antigen DF3, whereas retinoids, DMF, and HMBA, which all inhibit MCF-7 proliferation lead t o a reduction in DF3 (Abe and Kufe, 1986). In the present study, we measured the expression of another milk fat globule antigen, epithelial membrane antigen (EMA). Although this antigen is not confined to mammary gland, it is useful in histopathology and cytology as an indicator of epithelial differentiation (Sloane and Ormerod, 1981; To et al., 1981). Our cytometric studies confirm previous immunological studies indicating that breast cancer cells express EMA (Sloane and Ormerod, 1981). Treatment of MCF-7 cells with TPA, DMF, HMBA, or forskolin was found to increase significantly the expression of this antigen. With RA, we observed only a slight increase in the expression of EMA. When cell proliferation was blocked by the addition of insulin-free medium for 6 days, we did not observe an increase of EMA expression despite the blockade of MCF-7 cells in the G1 phase of the cell cycle (Gross et al,, 1984). Thus, the change in EMA expression is not simply a result of cytostasis. Although the increase in MCF-7 protein content (1.8-fold)induced only by TPA treatment, this could not entirely account for EMA antigen induction (4-fold). The expression of the ER in breast cancer cells is one of the most important characteristics of the differentiation status of the breast tumor. The proliferation of MCF-7 cells is controlled either directly or indirectly by ER via secretion of growth factors (Dickson and Lippman, 1987). In a previous study, we found that phorbol diesters decrease ER content (Guilbaud et al., 1988);we were therefore interested to find out whether this decrease was the consequence of a direct interaction between PKC and ER or a more general feature of breast cancer differentiation. The fact that there is a decrease in ER expression after treatment of MCF-7 cells with NaBu and CAMP-like drugs (Stevens et al., 1984; Cho-chung et al., 1981) is in favor of the second possibility. All the DIAs that induced maturation along the secretory cell lineage significantly decreased ER content without modifying estrogen affinity. By contrast, after treatment with RA, we observed a slight increase in ER content, in agreement with the findings of Butler (1989). This did not appear to be a simple consequence of the inhibition of growth of MCF-7 cells in the G1 phase of the cell cycle, since the addition of insulin-free medium did not modify the ER content, although it blocked G1-S transition (Gross et al., 1984). Moreover, Jakesz et al. (1984) reported that ER synthesis takes place mainly in the G1 and G2 phases. The specific reduction of ER by several DIAs observed in the present study confirms the results of Stevens et al. (1984) and Graham and Buick (1988) and argues in favor of a relationship between the maturation of
171
MCF-7 cells and the decrease in the expression of ER. The expression of ER is characteristic of a more differentiated status of breast carcinoma cells. ER concentrations in normal mammary epithelial cells are lower than in benign breast tumor, while higher concentrations are found in malignant tumor (Isotalo et al., 1983). Since the carcinogenic process in breast is accompanied by an elevation of ER number (Nenci et al., 1988), the decrease in ER content may represent a differentiation criterion. The results of the present study define some characteristics of the maturation of MCF-7 cells. A rapid block of the G 1 S transition with a complete inhibition of DNA synthesis is observed with the DIAs, which induce a maturational rocess (change in morphology and expression of E A). Inhibition of the proliferation of MCF-7 cells seems to be the primary effect of DMF, HMBA, TPA, NaBu, and forskolin. RA, which induces a decrease in growth rate, did not modify EMA expression and MCF-7 morphology, indicating that RA is not a DIA in MCF-7 cells. The major finding is the relationship between the decrease in ER expression and the induction of a more differentiated status of MCF-7 cells. The mechanism by which DMF, HMBA, TPA, and NaBu induce differentiation and decrease ER remains unclear. It now is of interest to determine the molecular mechanisms leading to the ER decrease observed during the MCF-7 maturation process. After treatment of MCF-7 cells with TPA, induction of the progesterone receptor by the ER is as effective as in untreated control cells (Guilbaud et al., 1988), suggesting that the remaining ER is active. ACKNOWLEDGMENTS This work was supported by Federation Nationale des Centres de Lutte Contre le Cancer. We are grateful to J.C. Hendrick and J. Colette for generously providing antibody to epithelial membrane antigen and to F. Roubinet and G. Cassar for their efficient assistance in performing flow cytometry experiments. LITERATURE CITED
s
Abe, M., and Kufe, D.W. (1984a) Sodium butyrate induction of milk related antigens in human MCF-7 breast carcinoma cells. Cancer Res., 44:45144577. Abe, M., and Kufe, D.W. (1984b) Effect of sodium butyrate on human breast carcinoma (MCF-7) cellular proliferation, morphology, and CEA production. Breast Cancer Res. Treatm., 41269-274. Abe, M., and Kufe, D.W. (1986) Effect of maturational agents on expression and secretion of two partially characterized high molecular weight milk related glycoproteins in MCF-7 breast carcinoma cells. J. Cell. Physiol., 126:12&132. Bloch, A. (1984)Induced cell differentiation in cancer therapy. Cancer Treatm. Rep., 681199-205. Burton, K. (1956) A study of the conditions and the mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem. J., 621315423. Butler, W.B. (1989) Responses to retinoic acid of tamoxifen-sensitive and tamoxifen-resistant sublines of the human cancer cell line MCF-7. Proc. Am. Assoc. Cancer Res., 30:305. Calabresi, P., Dexter, D.L., and Hepner, G.H. (1979) Clinical and pharmacological implications of cancer cell differentiation and heterogeneity. Biochem. Pharmacol., 2811933-1941. Cho-chung, Y.S., Clair, T., Bodwin, J.S., and Berghoffer, B. (1981) Growth arrest and morphological change of human breast cancer cells by dibutyryl cyclic AMP and L-arginine. Science, 214177-79. Chouvet, C., Vicard, E., Devonec, M., and Saez, S. (1986) 1,25Dihydroxyvitamin D3 inhibitory effect on the growth of two human
172
GUILBAUD ET AL
breast cancer cell lines (MCF-7, BT20). J. Steroid Biochem., 24:373376. Darbon, J.M., Valette, A,, and Bayard, F. (1986)Phorbol esters inhibit the proliferation of MCF-7 cells: Possible implication of protein kinase C. Biochem. Pharmacol., 35:2683-2686. Dickson, R.B., and Lippman, M.E. (1987) Estrogenic regulation of growth and polypeptides growth factor secretion in human breast carcinoma. Endocrine Rev., 8:29-43. Falette, N., Frappart, L., Lefebvre, M.F., and Saez, S. (1988) 1.25(OH)ZVit D3 treatments results in morphological changes, growth reduction and increase in epidermal growth factor receptors in a human breast adenocarcinoma cell line (BT 20). Proc. Am. Assoc. Cancer Res., 2951. Freshnev. R.I. (1985) Induction of differentiation in neodastic cells. Anticancer Res., 5:111-130. Graham, K.A.. and Buick, R.N. (1988) Sodium butyrate induces differentiation in breast cancer cell lines expressing the estrogen receptor. J. Cell. Physiol., 136r63-71. Gross, G.E., Boldt, D.H., and Osborne, C.K. (1984) Perturbation by insulin of human breast cancer cell cycle kinetics. Cancer Res., 44:3570-3575. Guilbaud, N., Pichon, M.F., Faye, J.C., Bayard, F., and Valette, A. (1988) Modulation of estrogen receptors by phorbol diesters in human breast cancer MCF-7 cell line. Mol. Cell. Endocrinol., 56:157-163. Isotalo, H., Tryggvason, K., Vierikko, P., Kauppila, A., and Vihko, R. (1983) Plasminogen activators and steroid receptor concentrations in normal, benign, and malignant breast and ovarian tissues. Anticancer Res., 3:331336. Jakesz, R., Smith, C.A.,Aitken, S., Huff, K., Schuette, W., Schackney, S., and Lippman, M. (1984) Influence of cell proliferation and cell cycle phase on expression of estrogen receptor in MCF-7 breast cancer cells. Cancer Res., 44:619-625. Jozan, S., Tournier, J.F., Tauber, J.P., and Bayard, F. (1982) Adap tation of the human breast cancer cell line MCF-7 to serum free medium culture on extracellular matrix. Biochem. Biophys. Res. Commun., 107:1566-1570. Knabbe, C., Lippman, M.E., Wakefield, L.M., Flanders, K.C., Kasid, A., Derynk, R., and Dickson, R.B. (1987) Evidence that transforming growth factor-p is a hormonally regulated negative growth factor in human breast cancer cell. Cell, 48:417-428. Lacroix, A., and Lippman, M.E. (1980) Binding of retinoids to human breast cancer cell lines and their effects on cell growth. J. Clin. Invest., 65:58&591. Langdon, S.P., Hawkes, M.M., Hay, F.G., Lowrie, S.S., Schol, D.J., Hilgers, J., Leonard, R.C.F., and Smyth, J.F. (1988) Effect of sodium butyrate and other differentiation inducers on poorly differentiated human ovarian carcinoma cell lines. Cancer Res., 48:6161-6165. Lotan, R. (1979) Different susceptibilities of human melanoma and breast carcinoma cell lines to retinoic acid induced growth inhibition. Cancer Res., 39:101&1019. Lowry, D.H., Rosebrough, N.H., Farr, A.L., and Randall, R.J. (1951) ~
Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193:265-275. Nenci, I., Machetti, E., and Querzoli, P. (1988) Comment on human mammary preneoplasia. The estrogen receptor-promotion hypothesis. J. Steroid Biochem., 30:105-106. Osborne, C.K., Hamilton, B., Nover, M., and Ziegler, J . (1981) Antagonism between epidermal growth factor and Phorbol ester tumor promoters in human breast cancer cell. J . Clin. Invest., 67:943-95 1. Reiss, M., Gamba-Vitalo, C., and Sartorelli, A.C. (1986) Induction of tumor cell differentiation as therapeutic approach: Preclinical models for hematopoietic and solid neoplasms. Cancer Treatm. Rep., 70:20 1-2 18. Resnicoff, M., and Medrano, E.E. (1989) Growth factors and hormones which affect survival, growth, and differentiation of the MCF-7 stem cells and their descendants. Exp. Cell Res., 181:116-125. Resnicoff, M., Medrano, E.E., Podhajcer, O.J., Bravo, A.I., Bover, L., and Mordoh, J. (1987) Subpopulations of MCF-7 cells separated by percoll gradient centrifugation: A model to analyse the heterogeneity of human breast cancer. Proc. Natl. Acad. Sci. USA, 84:72957299. Sidi, Y., Panet, C., Wasserman, L., Cyjon, A., Novogrodsky, A., and Nordenberg, J. (1988) Growth inhibition and induction of phenotypic alteration in MCF-7 breast cancer cells by IMP deshydrogenase inhibitor. Br. J. Cancer, 58:61-63. Sloane, J.P., and Ormerod, M.G. (1981) Distribution of epithelial membrane antigen in normal and neoplastic tissues and its value in diagnostic tumor pathology. Cancer, 47:1786-1795. Stevens, M.S., Aliabadi, Z., and Moore, M.R. (1984) Associated effects of sodium butyrate on histone acetylation and estrogen receptor in the human breast cancer cell line. Biochem. Biophys. Res. Commun., 119:132-138. To, A., Coleman, D.V., Dearnaly, J.P., Ormerod, M.G., Steele, K., and Neville, A.M. (1981) Use ofantisera to epithelial membrane antigen for the cytodiagnosis of malignancy in serous effusions. J . Clin. Pathol., 34:132&1332. Tsao, D., Morita, A., Bella, A.J.R., Liu, P., and Kim, Y.S. (1982) Differential effects of sodium butyrate, dimethyl sulfoxide, and retinoic acid on membrane associated antigen, enzymes and glycoproteins of human rectal adenocarcinoma cells. Cancer Res., 42:1052-1058. VAlette, A., Gas, N. Josan, S., Roubinet, F., Dupont, M.A., and Bayard, F. (1987) Influence of 12-O-tetradecanoylphorbol-13acetate on proliferation and maturation of human breast cancer cells (MCF-7): Relationship to cell cycle events. Cancer Res., 47:1615-1623. Van Deurs, B., Zou, Z., Briand, P., Balslev, Y., and Petersen, O.W. (1987) Epithelial membrane polarity: A stable, differentiated feature of an established human breast carcinoma cell line MCF-7. J. Histochem. Cytochem., 33:461-469.
