186,250-256 (1990)
EXPERIMENTALCELLRESEARCH
/n Vitro and in Viva Growth and Casein Gene Expression of Mouse Mammary Tumor Epithelial Cells in Response to Hormones JIA-LING CHOU,*~~ZHI-XIANG SHEN,* IRENE J. TAN,* ROBERT L. STOLFI,? DANIEL S. MARTIN,? STEVEN H. DIKMAN,$ AND SAMUEL WAXMAN*,’ *Chemotherapy
Foundation Laboratory, Division of Medical Oncology, Mount Sinai Medical Center, New York, New York 10029; tCatholic Medical Center of Brooklyn and Queens, Inc., Woodhauen, New York, New York 11241; and *Department of Pathology, Mount Sinai Medical Center, New York, New York 10029
INTRODUCTION Cells from autochthonous mouse mammary carcinomas which display estrogen-independent growth in vivo were studied for their hormonal responses in primary culture. A culture system employing insulin-supplemented, serum-free medium and basement membrane Matrigel as a substratum was used to cultivate tumor cells. The cells did not exhibit in vitro estrogenor prolactin-dependent growth. Primary tumors still displayed a constitutional expression of (Y-, &, and ycasein mRNAs. These messages were dramatically reduced during the culture period. However, seven to eightfold increases in LY-and &casein mRNAs were inducible in the B-day cultures by treatment with prolactin and hydrocortisone. If the hormones were present through a a-week culture period, the levels of a-, ,8-, and y-casein mRNAs in the cells were maintained and displayed in a time-dependent increase with a peak at lo14 days. The accumulation of @-casein mRNA in vitro did not require DNA synthesis. Administration of prolactin directly into the growing tumors in vivo could also enhance ,&casein mRNA levels in the tumor cells. Morphological studies of the cells cultured in the presence of prolactin and hydrocortisone did not reveal visible changes compared with those without hormonal treatment. Transplantation of tumor cells cultured in the presence or absence of hormones resulted in the development of tumors in mice at approximately the same time. The current studies suggest that the autochthonous mammary tumor cells, independent of estrogen for cell growth, were still inducible for casein gene expression in vitro and in vivo by appropriate hormones. The induction and maintenance of casein messages by a single hormonal treatment did not appear to correlate with morphology and DNA synthesis of cells in vitro or with tumor-producing capacities in vivo. 0 isso Academic Press,
Inc.
1 Present address: Department of Genetics, The Hospital for Sick Children, Toronto, Ontario, Canada, M5G 1X8. ’ To whom reprint requests should be addressed at: Division of Medical Oncology, Box 1178, Mount Sinai Medical Center, New York, NY 10029.
0014-4827/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
Understanding the relationship between the hormonal regulation of growth and the induction of expression of tissue-specific function in mammary tumor epithelial cells may provide insights into the mechanism whereby transformation perturbs normal hormonal regulation. Synthesis of casein in the mammary gland is under the control of peptide and steroid hormones and can serve as a differentiation marker which in time may be correlated with the responsiveness of mammary tumor epithelial cells to hormonal regulation of growth. It has been demonstrated that 7,12dimethylbenz(a)anthracene (DMBA)-induced rat mammary tumor cells maintain estrogen- or prolactin-dependent growth properties in uiuo. A small population of the tumor cells is inducible for casein gene expression, both in uiuo [l] and in vitro [2]. N-nitroso-methylurea (NMU)-induced rat mammary tumor cells, demonstrating hormonal dependence for growth, were inducible for casein mRNA in uivo [l] but not inducible for ,&casein by prolactin treatment in vitro despite their constitutive expression of the messages [3]. It is certainly of interest to investigate whether mammary tumor cells independent of estrogen for growth are also unresponsive to hormones for tissue-specific gene expression. An important aspect of the current studies was to determine whether induction of differentiation-associated phenotypes by hormonal manipulation in a proper culture milieu would affect the tumor cell growth potential. In the rat mammary cells, it was demonstrated that pregnancy and lactation can suppress the development of chemical carcinogen-transformed mammary epithelial cells [4, 51. There have been comparatively fewer studies in mouse mammary tumor epithelial cells (MMTCs). Previous studies in our laboratory defined a culture system suitable for the growth of MMTCs in serumfree medium on basement membrane Matrigel (Matrigel), and cells grown in primary cultures were inducible for P-casein mRNA by hormonal treatment [6,7]. In the current studies, cells from autochthonous mammary tumors and an established MMTC line were employed to 250
Inc. reserved.
~~R~~NAL
REGULATION
OF GROWTH
investigate the effects of hormonal manipulation on (Y-, /3-, and y-casein gene expression in vitro as well as the correlation between the inducibility of casein gene expression and tumor cell growth in vitro and in vivo. ~A~E~~ALS
AND
METHODS
Cell line. FUKU cells, an established mouse mammary tumor epitbelial cell line derived from BALB/cfC3H mammary adenocarcinoma [8], were kindly provided by Dr. M. J. Yagi (Mount Sinai Medical Center, New York). They were maintained in media composed of a 1:l mixture (v/v) of Dulbecco’s modified Eagle’s medium (DMEM) and Ham’s F12 (DMEM/FlZ, both from GIBCO) supplemented with 10% beat-inactivated fetal bovine serum (HI-FBS, GIBCO) and 10 pg/ml insulin (Sigma). They were cultured in a humidified incubator at 37°C in 5% CO,/95% air. To assay casein gene expression, FUKU cells were initially cultured in 2% HI-FBS for 3 days. Cells in the logarithmic growth phase then were plated on basement membrane Matrigel (Matrigel, Collaborative Research) or Vitrogen 100 (containing 9598% type I collagen; Collagen Co. Palo Alto, CA) in serum-free, HEPES-buffered DMEM/FlZ media supplemented with insulin (10 pg/ml). One day after seeding, media were changed and the Vitrogen gels were made to float. Ovine prolactin (5 pg/ml, Sigma) and hydrocortisone (1 pg/ml, Sigma) were administered 4 days after culture and the cells were harvested 24 or 48 h after hormonal treatment for Northern blot assays. Primary cultures. Mouse mammary tumor virus (MMTV)-induced mammary tumors were obtained from female CD8Fl (BALB/ cfC3H X DBA/B) mice as described [9]. The tumors were removed on the day of experiment and tissue dissociation was conducted essentially as described [lo]. The dissociated cells were then cultured on Matrigel-coated B-well plates in serum-free DMEM/FlZ media supplemented with 10 pg/mI insulin. The media were changed 24 h after seeding. Some cultures received 5 pg/ml prolactin and 1 wg/ml hydrocortisone for various periods. Assay of cell growth. Primary MMTC cultures or FUKU cells in serum-free media were treated with 17P-estradiol (Sigma) or prolactin. On the day of harvest, the cells were released from gels by dispase (Collaborative Research) treatment for 1 h at 37°C. Aliquots of the cells were assayed for viability by trypan blue exclusion and the viability was found to be greater than 90%. The rest of the cells were subject to lysis with glacial acetic acid to obtain intact nuclei. The nuclei were counted on a Coulter counter as described [ll]. Morphological studies. MMTCs were cultured on Matrigel in the presence or absence of insulin, prolactin, and hydrocortisone for up to 14 days. The cultures were terminated by fixing in Bouin’s solution and washed with 70% ethanol. The cells together with the Matrigel were embedded in paraffin. Cross sections were stained with hematoxyhn and eosin for light microscopy. Immunofluorescence. Tissues from primary mammary tumors and normal mouse lactating mammary gland were frozen in liquid N,. Sections made from these tissues were fixed using a formaldehyde-cold acetone procedure as described [x2]. The sections were stained with a rabbit antiserum against mouse p-casein provided by Dr. J. S. Butel (Baylor College of Medicine, Houston, TX) and subsequently with a fluorescein-conjugated goat antiserum. They were examined under a Zeiss fluorescence microscope. Determination of [“H] thymidine incorporation. MMTCs at 0.6 X 10s cells/well were cultured on Matrigel. They were treated with 1 &i/ml of [3H]tbymidine (sp act 14.3 Ci/mmol, Amersham) for various periods. The cells then were removed from Matrigel by digesting with dispase for 1 hat 37°C. After a wash in PBS to recover the intact cells, aliquots were counted to determine cell number. For determination of [Hlthymidine incorporation, the cells were washed consecutively with perchloric acid, ethanol, and perchloric acid. After hydrolysis in perchloric acid (8O”C), aliquots were counted in a scintillation counter [13].
AND
CASEIN
EXPRESSION
251
IN MMTC
In vivo prolactin administration. &Fl mice were given two prola&in injections (200 pg contained in 0.05 ml) with an interval of 4 days into the center of the tumors. Controls received vehicle injection. Tumors were removed 3 days after the second injection and RNA was extracted. Tumor transplantation. Cells from three primary mammary tumors were cultured in serum-free medium on Matrigel in the presence or absence of insulin, prolactin, and hydrocortisone for 9 days. The cells (1 X 106) after dispase treatment in 0.1 ml of saline were injected S.C.into 6-week-old male CD8Fl mice. Hybridization studies. Totai RNA was extracted with the guanidine thiocyanate method as described [14]. RNA was quantitated and stored in ethanol at -30°C. An equal amount of RNA was electropboresed in 1.2% agarose gel containing formaldehyde, blotted to Genescreen Plus membrane (New England Nuclear), crosslinked to membrane by uv light, hybridized with [a-32P]dCTP-3abeled probes, and washed as recommendedby the manufacturer. The mouse (I, F and y-casein probes [15] were provided by Dr. J. M. Rosen (Baylor College of Medicine, Houston, TX). The hamster actin cDNA probe 1161 was provided by Dr. F. Smith (Mount Sinai Medical Center, New York). Hybridization with the actin probe was used as a control in each Northern blot assay. The inserts were isolated according to the method described [i4] and the random primed labeling method gave a specific activity of greater than 10” dpm/pg. The autoradiographs were scanned using a laser densitometer $XB-2222-020, ultrascan XL) and the peak areas integrated.
Growth
of
TCs in Serum-Free
atium
Mouse mammary tumor epithehal cehs were obtained from ~MT~-induced mammary tumors i Because our aim was to inv siveness of MMTCs in vitro, to culture the cells in serum-free demonstrated that basement rigel), prepared from the Enge mor, allowed a better gr fmouse mammary epithelial cells in serum-free fibronectin, laminin, type I, or type IV co hus, in the current studies the cells were cultured on atrigel in serum-free /F12 media supplemen insulin. The of MMTCs on Matrigel nt a steady increase over a 5-day culture peri en 0.6 X IO6 cells ay 0, a total of I.8 X PO6cells were harvested on Day 5 (the average of eight observations). As both prolactin and estrogen have been reported ta be mitogenic for mammary tumor epitheliali cells [B7], the two hormones were tested in serum-free medium for their infhrence on MT@ growtk. Tumor cells from mice received pro tin or ~7~-estra~~o~ treatment ulture period and the cell number was deterays 2 and 4 after culture. (41:5 pg/ml, bad lit (P > 0.05, data not shown). were tested, only a Iigbt increase (1.45 number was observe at Day 4 when 100 n was used (Fig. 1). Similar experiments were carried out with FUKU cells, an estabhshed TC line [$I- The growth of
CHOU
I-
1
Control 0.1 nM I nM IO nM 100 nM
ET AL.
ment and 1 X lo6 cells from each tumor cultured in the presence or absence of prolactin and hydrocortisone were injected S.C. into duplicate male CD8Fl mice. Tumors developed in all the mice 6-8 weeks after injection. Modulation
,-
-J 2
DAYS FIG. 1. Growth of MMTCs in the presence of estradiol. Tumor cells from four mice (0.5 X lo6 cells/well) were cultured on Day 0 on Matrigel in the presence or absence of 17@-estradiol in serum-free DMEM/FlB media. Cells were collected on Days 2 and 4. The cell number and viability in duplicate wells were determined with >90% viable.
FUKU cells was not affected by the presence of prolactin (5 ,ug/ml) or estradiol (0.1-100 r&f) (data not shown). Morphology of MMTCs Cultured in the Presence of Prolactin and Hydrocortisone Morphological studies were done on cross sections made from MMTCs cultured on Matrigel for 3, 7, 10, and 14 days in the presence or absence of insulin, prolactin, and hydrocortisone. The tumor cells plated in Matrigel formed aggregated clumps in both groups (Figs 2A and 2B). Fibroblasts were not visible in any of the sections examined. No marked morphological differences were observed after 2 weeks in cultures of hormonetreated or control groups (Figs. 2E and 2F). The tumor cells in the center of the clumps started to die with longer culturing (Figs. 2C and 2D), whereas the cells located at the periphery remained viable and possibly continued to divide. In some cases, the dead cells were disintegrated and this resulted in “gland-like” structures which became more marked after about 10 days in culture (Figs. 2E and 2F). It is obvious from the gradual changes with time that the appearance of the open spaces in the center of the clumps was not due to an artifact created during section preparations. Tumor Transplantation MMTCs from three primary tumors were cultured in insulin-supplemented, serum-free medium on Matrigel for 9 days in the presence or absence of prolactin and hydrocortisone. Cells were harvested after dispase treat-
of Casein Gene Expression
We have demonstrated that MMTCs in 59 out of 60 mammary tumors from CD8Fl mice constituitively express detectable levels of @-casein mRNA as detected by Northern blot assays [6,7]. The levels were less than 2% of those in the normal mammary gland from lactating mice. In the current studies variable levels of (Y- and ycasein messages were also shown to be constituitively expressed in all the primary tumors tested (Fig. 3). It has been demonstrated by others and by us that the in vitro induction of @-casein mRNA requires an interaction between mammary epithelial cells and extracellular matrix in the presence of appropriate hormones [ 181. In the current studies, experiments were performed to investigate the induction and maintenance of CY-,y-, and @casein mRNAs in MMTCs in primary culture. MMTCs from CD8Fl mice were cultured on Matrigel in serum-free medium supplemented with insulin. The cultures were treated with prolactin and hydrocortisone on Day 3, as both hormones, in the presence of insulin, are required for optimal induction of casein mRNA [lo, 191. We have previously shown that the ,fScasein mRNA messages could not be demonstrated in s-day culture in the absence of prolactin and hydrocortisone by Northern blot assays [6, 71. However, when prolactin and hydrocortisone were added on Day 3, the P-casein messages were detectable 12 h after hormonal treatment. The accumulation of @-casein mRNA increased to eightfold at 24 h and were reduced to about fivefold at 48 h (Fig. 4). cu-casein mRNA was still weakly detectable on Day 3 even in the absence of hormonal treatment. Its level was increased about fivefold 12 h after prolactin and hydrocortisone treatment and then accumulated slowly within the next 36 h (Fig. 4). The time-dependent changes in P-casein mRNA levels in response to hormones in our studies appeared similar to previous reports, as assayed by RNA dot-blotting, of DMBA-induced rat mammary carcinoma cells in primary culture [2] or of COMMA1D cells, an established mouse mammary epithelial cell line [20]. However, the peak levels for p-casein mRNA in our serum-free cultures appeared about 24-48 h earlier. The delay in ,&casein mRNA appearance may be attributed to the presence of serum in the previous culture systems, as serum contains hormones or other factors which may exert a prolonged effect on casein mRNA accumulation. Additional experiments were performed to assess the maintenance of casein mRNAs during a 2-week culture period. Hormones were present throughout the culture period and the MMTCs were collected on Days 3, 7, 10, and 14. After 3 days in culture, the three casein messages
HORMONAL
REGULATION
OF GROWTH
AND
CASEIN
EXPRESSION
IN MMTC
253
FIG. 2. H-E stain of MMTCs in primary cultures (X125). MMTCs were cultured on Matrigel for 3(A, B), 7(C), 10(D), or 14(E, F) days in the presence (B, C, D, F) or absence (A, E) of prolactin and hydrocortisone. Tumor cells formed clumps and no fibroblasts were observed. Cells in the center of the clumps died and became disintegrated after 1 week in culture. This resulted in gland-like structures on Day 14.
could still be maintained although the levels showed different degrees of reduction compared to those prior to culture. The highest levels of y-casein mRNA were
detected in MMTCs on Day 14 (Fig. 5) while those of CY-and ,&casein mRNA around dent that the accumulated levels of casein mRNAs in
254
CHOU
ET AL.
4-
12345678 + (I - casein + Y-casein
3-
+ actin 2FIG. 3. Northern blot assay of 01- and y-casein mRNA levels in MMTCs. Total RNA was extracted from FUKU cells cultured in the presence of serum (lane 2) or from freshly prepared tumor samples before culture (lanes 3-8). In lane 1,50 ng of RNA from 7-day lactating mammary gland of normal mouse was loaded as a comparison to the other lanes where 5 pg of RNA was loaded. The RNA was hybridized with the 01-and y-casein probes. The actin probe was hybridized to the blots to quantitate the levels of mRNA in each lane.
l-
O-
IO DAYS
MMTCs showed a time-dependent increase over the 14day incubation period. The peak levels for a-casein mRNA were 3.5-fold greater than those before culture and appeared earlier than those for y-casein messages. Nevertheless, the actin message levels in MMTCs harvested at different times remained the same (data not shown). Immunofluorescence studies using a specific antibody against @casein did not detect /3-casein in primary mammary tumors expressing ,&casein mRNA before culture. In contrast, positive staining of lactating mammary tissues was readily detectable (data not shown). Therefore, characterization of cultured MMTCs using the same antibody was not pursued.
4
Incorporation
To determine whether the induction of the @-casein messages required DNA synthesis, the MMTCs in Day3 cultures received 24 h treatment of prolactin, hydrocortisone, or both, and then [3H]thymidine incorpora-
2 o-o-o 12
In order to detect in viuo response to prolactin treatment, prolactin was injected into the tumor twice with a 4-day interval. Eight days after the first injection, the tumors were removed to extract RNA. Data from six mice indicated that in vivo treatment with prolactin in two out of three mice resulted in a higher level of accumulated P-casein mRNA (Fig. 6, lanes 4, 5) than that seen in three control mice which retained low levels of p-casein mRNA (that received vehicle injections). When FUKU cells were cultivated on Matrigel or floating type I collagen gels, they formed aggregated clumps just as MMTCs from CD8Fl mice in primary cultures. However, when FUKU cells were treated with prolactin and hydrocortisone, no ,8-casein mRNA was detectable by Northern blot assays (data not shown). Thymidine
./
b 01 0
FIG. 5. Maintenance of casein mRNA levels during a 2-week culture. MMTCs from three tumors were cultured separately in serumfree medium supplemented with insulin, prolactin, and hydrocortisone. Total RNA from MMTCs prior to culture (control) or cultured for 3, 7, 10, and 14 days was analyzed by Northern blotting. The blots were hybridized and the results analyzed as described in Fig. 4 except that y-casein probe was also used in rehybridization.
24
36
48
TIME (hr) FIG. 4. Time-dependent induction of LY-and fl-casein mRNAs in MMTCs. MMTCs from CDSFl mice were cultured on Matrigel in serum-free medium supplemented with insulin. Prolactin and hydrocortisone were added on Day 3. Total RNA in three experiments was extracted from cells on Day 3 before hormonal treatment (0 time control) or 12 h, 24 h, and 48 h following hormonal administration. The Northern blots were hybridized with a- and fl-casein and actin probes. The bands were scanned using a laser densitometer. The curves represent relative changes in the three message levels after hormonal treatment to those before hormonal exposure which were arbitrarily defined as one unit.
123456 * actifl M/3-casein FIG. 6. In vivo changes of p-casein mRNA. Three mice were injected twice with prolactin into the tumors. Tumors from three mice receiving two vehicle injections were used as a control. Two micrograms of total RNA was blotted and hybridized with /3-casein and actin probes. Lanes 1-3, control; lanes 4-6, prolactin-treated.
HORMONAL
TABLE Effects of Prolactin, [3H]Thymidine
REGULATION
OF GROWTH
1
Hydrocortisone, Incorporation
Treatment” Experiment 1 None Prolactin Hydrocortisone Prolactin and hydrocortisone Experiment 2 Prolactin and bydrocortisone Ara-C plus prolactin and hydrocortisone
and Ara-C into MMTCs
on
cpm/106 cells ( X10m3)b
27.5 f 2.4 32.6 t 2.6 28.12 4.5
27.2 f 3.5 35.2 -i 0.9 0.5 + 0.3
e MMTCs were cultured on Day 0 at 0.5 X lo6 cells/well on Matrigel. In Experiment 1, prolactin (5 pg/ml), hydrocortisone (1 pg/ml), or both were added on Day 3, and [3H]thymidine was then added. In Experiment 2, MMTC cultures, initially treated with or without Ara-C (30 pg/ml), were exposed to prolactin and hydrocortisone for 24 h. In all cases, the cells were harvested on Day 4 after exposure to [3H]tbymidine for 23 b to determine cell number and thymidine incorporation. b The mean (iSD) of four observations are shown.
tion was determined. As shown in Table 1, none of the three different kinds of treatment resulted in any significant changes in 13H] thymidine incorporation (P > 0.05). The cell number in each experiment remained the same in the different treatment groups (data not shown). In contrast, cytosine arabinoside (Ara-C) at 30 more than 95% of the [3H]thymidine inle 1). However, MMTCs in the Ara-Ctreated and nontreated cultures displayed similar levels of the /3-casein messages after 24-b exposure to prolactin and hydrocortisone (Fig. 7). This demonstrates that the hormonal induction of ,&casein gene expression in the MMTCs does not require DNA synthesis, which agrees with previous observations on casein synthesis in preneoplastic mouse mammary tissues [21].
The aim of the current studies was to observe hormonal modulation of mammary-specific gene expression in MMTCs in primary culture and to look for a correlation between the induced differentiated phenotype and tumor cell growth potential in vitro and in uiuo. Autochthonous mammary tumors from CD8Fl mice were employed in the current investigation. The tumor cells are known to be independent of ovarian steroids for growth in viva from the following observations: the tumors continue to grow after ovariectomy; the transplanted tumor cells display the same growth rate in male or female recipients; and tamoxifen can not inhibit the growth of transplanted tumors (unpublished data). In the current in vitro studies, it was further proven that estradiol, up to 100 niW, did not stimulate the growth of MMTCs in serum-free medium supplemented with insu-
AND
CASEIN
EXPRESSION
IN MMTC
255
lin. Furthermore, when the same tumor cells were used, prolactin did not increase the cell number or dine incorporation at the ~on~e~tratio~s that sein gene expression The unresponsiveness of th ture to prolactin or estrogen trasted by the in vitro indueib mRNAs in the cells. The levels of casein dramatically reduced on Day 3 after cult free medium on Matrigel in the absence of hormones. However, 24 h treatment with pro sone in the presence of ins n was sufficient to elicit a marked increase in tbe site -state levels of CY-and flcasein mRNAs. This indi presence of insulin and hydroc at least a subpopucumulation of the casein messages lation of the tumor cells. If the three hormones were continuous1 ing a 2-week culture, the levels of mRNAs could be maintained and s dent increase after 10 days in cul whether this increase with time was d 50 am accumulation of casein messages in a subs or due to an expansion of easein. ue to the death of tively, this increase may simply those cells which synthesized lower levels sages. Tbis possibility is mor servations that the cells in the die after 1 week in culture while those closer trigel were still viable. In this regard in robes will be require verify the possibility. Although appropriate and maintained casein m alter the rate of cell grow tion. Cells cultured in the presence or absence of hormones for 2 weeks did not look different morphologically. The in vluo data from t~rnorige~~~ studies employresence of insulin, ing MMTCs cultured in the rocortisone for prolactin, and those from con groups, while c mone-treated primary culture increased with time.
1
2
FIG. 7. Effect of Ara-C treatment on @-casein mRNA levels. MMTCs in primary cultures received 24-b treatment ofprolactin and hydrocortisone in the absence (lane 1) or the presence (lane 2) of AraC (30 pg/ml). Two micrograms of total RNA was loaded in each lane and hybridized with p-casein and actin probes.
256
CHOU
These results suggest that the induction and maintenance of casein mRNAs in the tumor cells by a single lactogenic hormone treatment was not correlated with tumor cell growth in vitro or tumor-producing capacities in uivo. Consistent with this, a lactogenic stimulus of mouse mammary preneoplastic cells did not appear to change the tumor potential in two out of three outgrowth lines but inhibited the tumorigenic potential in one line [22]. It remains to be determined whether a prolonged lactogenic stimulus in our tumor model will alter in ho tumor growth potential. A number of agents known to affect growth and differentiation in established tumor cell lines can be studied in this primary MMTC culture system since it should provide a proliferating tumor cell culture with differentiation-associated phenotypes. We have studied several agents in this system including HMBA, retinoids, and transforming growth factors (TGFs). Thus far, only TGF-a! was shown to modulate ,&casein mRNA levels in the presence of prolactin and hydrocortisone (manuscript in preparation). We thank Dr. G. Coppola and Ms. C. Schreiber for their excellent technical assistance. We also thank R. Hromada for preparing the manuscript. The work was supported by grants from the Chemotherapy Foundation, the Samuel Waxman Cancer Research Foundation and the Herman Goldman Foundation; and by USPHS Grant ROICA 25842 to RLS and DSM.
REFERENCES 1.
Suposit, 1360.
S. C., and Rosen, J. M. (1982) Cancer Res. 42, 1355-
2.
Johnson, Dev.Biol.
M. L., Levy, J., and Rosen, J. M. (1985) In Vitro Cell. 2 1,439-444.
3.
Edery, M., Houdebine, L. M., Djiane, J., and Kelly, Mol. Cell. Endocrinol. 34, 145-151.
Received May 24,1989 Revised version received August 21, 1989
P. A. (1984)
ET AL. 4.
Dao, T. L., Bock, F. G., and Greiner, J. J. (1960) J. Natl. Cancer Inst. 25,991-1002.
5.
Grubbs, C. J., Hill, D. L., McDonough, K. C., and Peckham, J. C. (1983) J. Natl. Cancer Inst. 71,625-628.
6.
Chou, J. L., Coppola, G., Shen, Z. X., Schreiber, C., and Waxman, S. (1988) Proc. Amer. Assoc. Cancer Res. 29,42 [No. 1671
7.
Chou, J.-L., Shen, Z.-X., Stolfi, R. L., Martin, man, S. (1989) Cancer Res. 49,5371-5376.
8.
Yagi, M. J. (1973) J. Natl. Cancer Inst. 51,1849-1855.
9.
Stolfi, R. L., Martin, D. S., and Fugmann, Chemother. Rep. 55,239-251.
D. S., and Wax-
R. A. (1971) Cancer,
10.
Emerman, J. T., Enami, J., Pitelka, Proc. Natl. Acad. Sci. 74,4466-4470.
11.
McGrath, M., Palmer, 1255,182-191.
12.
Asch, B. B., Medina, D., Kretzer, F., Connolly, ley, B. R. (1979) Cancer Res. 39,893-907.
13.
Shamay, A., and Gertler,
D. R., and Nandi,
S., and Nandi,
S. (1977)
S. (1985) J. Cell. Physiol. J. L., and Brink-
A. (1986) Cell Biol. Int. Rep. 10, 923-
929. 14.
Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
15.
Gupta, P., Rosen, J. M., D’Eustachio, (1982) J. Cell Biol. 93, 199-204.
16.
Dodemont, H. J., Soriano, P., Quax, W. J., Ramaekers, F., Johannes, A. L., Groenen, M. A. M., Bernardi, G., and Bloemendal, H. (1982) EMBO J. 1,167-171.
P., and Ruddle,
F. H.
17.
Welsch, C. W. (1985) Cancer Res. 45,3415-3443.
18.
Wicha, M. S., Lowrie, G., Kohn, E., Bagavandoss, T. (1982) Proc. N&l. Acad. Sci. 79,3213-3217.
19.
Hobbs, A. A., Richards, D. A., Kessler, D. J., and Rosen, J. M. (1982) J. Biol. Chem. 25,3598-3605.
20.
Rosen, J. M., Rodgers, J. R., Couch, C. H., Bisbee, C. A., DavidInouye, Y., Campbell, S. M., and Yu-Lee, L.-Y. (1986) Ann. N. Y. Acad. Sci. 478,63-76.
21.
Smith, G. H., Vonderhaar, B. K., Graham, D. E., and Medina, D. (1984) Cancer Res. 44,3426-3437.
22.
Medina, D., Schwartz, M., Taha, M., Oborn, C. J., and Smith, G. H. (1987) Cancer Res. 47,4686-4693.
P., and Mahn,
EXPERIMENTALCELLRESEARCH
/n Vitro and in Viva Growth and Casein Gene Expression of Mouse Mammary Tumor Epithelial Cells in Response to Hormones JIA-LING CHOU,*~~ZHI-XIANG SHEN,* IRENE J. TAN,* ROBERT L. STOLFI,? DANIEL S. MARTIN,? STEVEN H. DIKMAN,$ AND SAMUEL WAXMAN*,’ *Chemotherapy
Foundation Laboratory, Division of Medical Oncology, Mount Sinai Medical Center, New York, New York 10029; tCatholic Medical Center of Brooklyn and Queens, Inc., Woodhauen, New York, New York 11241; and *Department of Pathology, Mount Sinai Medical Center, New York, New York 10029
INTRODUCTION Cells from autochthonous mouse mammary carcinomas which display estrogen-independent growth in vivo were studied for their hormonal responses in primary culture. A culture system employing insulin-supplemented, serum-free medium and basement membrane Matrigel as a substratum was used to cultivate tumor cells. The cells did not exhibit in vitro estrogenor prolactin-dependent growth. Primary tumors still displayed a constitutional expression of (Y-, &, and ycasein mRNAs. These messages were dramatically reduced during the culture period. However, seven to eightfold increases in LY-and &casein mRNAs were inducible in the B-day cultures by treatment with prolactin and hydrocortisone. If the hormones were present through a a-week culture period, the levels of a-, ,8-, and y-casein mRNAs in the cells were maintained and displayed in a time-dependent increase with a peak at lo14 days. The accumulation of @-casein mRNA in vitro did not require DNA synthesis. Administration of prolactin directly into the growing tumors in vivo could also enhance ,&casein mRNA levels in the tumor cells. Morphological studies of the cells cultured in the presence of prolactin and hydrocortisone did not reveal visible changes compared with those without hormonal treatment. Transplantation of tumor cells cultured in the presence or absence of hormones resulted in the development of tumors in mice at approximately the same time. The current studies suggest that the autochthonous mammary tumor cells, independent of estrogen for cell growth, were still inducible for casein gene expression in vitro and in vivo by appropriate hormones. The induction and maintenance of casein messages by a single hormonal treatment did not appear to correlate with morphology and DNA synthesis of cells in vitro or with tumor-producing capacities in vivo. 0 isso Academic Press,
Inc.
1 Present address: Department of Genetics, The Hospital for Sick Children, Toronto, Ontario, Canada, M5G 1X8. ’ To whom reprint requests should be addressed at: Division of Medical Oncology, Box 1178, Mount Sinai Medical Center, New York, NY 10029.
0014-4827/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
Understanding the relationship between the hormonal regulation of growth and the induction of expression of tissue-specific function in mammary tumor epithelial cells may provide insights into the mechanism whereby transformation perturbs normal hormonal regulation. Synthesis of casein in the mammary gland is under the control of peptide and steroid hormones and can serve as a differentiation marker which in time may be correlated with the responsiveness of mammary tumor epithelial cells to hormonal regulation of growth. It has been demonstrated that 7,12dimethylbenz(a)anthracene (DMBA)-induced rat mammary tumor cells maintain estrogen- or prolactin-dependent growth properties in uiuo. A small population of the tumor cells is inducible for casein gene expression, both in uiuo [l] and in vitro [2]. N-nitroso-methylurea (NMU)-induced rat mammary tumor cells, demonstrating hormonal dependence for growth, were inducible for casein mRNA in uivo [l] but not inducible for ,&casein by prolactin treatment in vitro despite their constitutive expression of the messages [3]. It is certainly of interest to investigate whether mammary tumor cells independent of estrogen for growth are also unresponsive to hormones for tissue-specific gene expression. An important aspect of the current studies was to determine whether induction of differentiation-associated phenotypes by hormonal manipulation in a proper culture milieu would affect the tumor cell growth potential. In the rat mammary cells, it was demonstrated that pregnancy and lactation can suppress the development of chemical carcinogen-transformed mammary epithelial cells [4, 51. There have been comparatively fewer studies in mouse mammary tumor epithelial cells (MMTCs). Previous studies in our laboratory defined a culture system suitable for the growth of MMTCs in serumfree medium on basement membrane Matrigel (Matrigel), and cells grown in primary cultures were inducible for P-casein mRNA by hormonal treatment [6,7]. In the current studies, cells from autochthonous mammary tumors and an established MMTC line were employed to 250
Inc. reserved.
~~R~~NAL
REGULATION
OF GROWTH
investigate the effects of hormonal manipulation on (Y-, /3-, and y-casein gene expression in vitro as well as the correlation between the inducibility of casein gene expression and tumor cell growth in vitro and in vivo. ~A~E~~ALS
AND
METHODS
Cell line. FUKU cells, an established mouse mammary tumor epitbelial cell line derived from BALB/cfC3H mammary adenocarcinoma [8], were kindly provided by Dr. M. J. Yagi (Mount Sinai Medical Center, New York). They were maintained in media composed of a 1:l mixture (v/v) of Dulbecco’s modified Eagle’s medium (DMEM) and Ham’s F12 (DMEM/FlZ, both from GIBCO) supplemented with 10% beat-inactivated fetal bovine serum (HI-FBS, GIBCO) and 10 pg/ml insulin (Sigma). They were cultured in a humidified incubator at 37°C in 5% CO,/95% air. To assay casein gene expression, FUKU cells were initially cultured in 2% HI-FBS for 3 days. Cells in the logarithmic growth phase then were plated on basement membrane Matrigel (Matrigel, Collaborative Research) or Vitrogen 100 (containing 9598% type I collagen; Collagen Co. Palo Alto, CA) in serum-free, HEPES-buffered DMEM/FlZ media supplemented with insulin (10 pg/ml). One day after seeding, media were changed and the Vitrogen gels were made to float. Ovine prolactin (5 pg/ml, Sigma) and hydrocortisone (1 pg/ml, Sigma) were administered 4 days after culture and the cells were harvested 24 or 48 h after hormonal treatment for Northern blot assays. Primary cultures. Mouse mammary tumor virus (MMTV)-induced mammary tumors were obtained from female CD8Fl (BALB/ cfC3H X DBA/B) mice as described [9]. The tumors were removed on the day of experiment and tissue dissociation was conducted essentially as described [lo]. The dissociated cells were then cultured on Matrigel-coated B-well plates in serum-free DMEM/FlZ media supplemented with 10 pg/mI insulin. The media were changed 24 h after seeding. Some cultures received 5 pg/ml prolactin and 1 wg/ml hydrocortisone for various periods. Assay of cell growth. Primary MMTC cultures or FUKU cells in serum-free media were treated with 17P-estradiol (Sigma) or prolactin. On the day of harvest, the cells were released from gels by dispase (Collaborative Research) treatment for 1 h at 37°C. Aliquots of the cells were assayed for viability by trypan blue exclusion and the viability was found to be greater than 90%. The rest of the cells were subject to lysis with glacial acetic acid to obtain intact nuclei. The nuclei were counted on a Coulter counter as described [ll]. Morphological studies. MMTCs were cultured on Matrigel in the presence or absence of insulin, prolactin, and hydrocortisone for up to 14 days. The cultures were terminated by fixing in Bouin’s solution and washed with 70% ethanol. The cells together with the Matrigel were embedded in paraffin. Cross sections were stained with hematoxyhn and eosin for light microscopy. Immunofluorescence. Tissues from primary mammary tumors and normal mouse lactating mammary gland were frozen in liquid N,. Sections made from these tissues were fixed using a formaldehyde-cold acetone procedure as described [x2]. The sections were stained with a rabbit antiserum against mouse p-casein provided by Dr. J. S. Butel (Baylor College of Medicine, Houston, TX) and subsequently with a fluorescein-conjugated goat antiserum. They were examined under a Zeiss fluorescence microscope. Determination of [“H] thymidine incorporation. MMTCs at 0.6 X 10s cells/well were cultured on Matrigel. They were treated with 1 &i/ml of [3H]tbymidine (sp act 14.3 Ci/mmol, Amersham) for various periods. The cells then were removed from Matrigel by digesting with dispase for 1 hat 37°C. After a wash in PBS to recover the intact cells, aliquots were counted to determine cell number. For determination of [Hlthymidine incorporation, the cells were washed consecutively with perchloric acid, ethanol, and perchloric acid. After hydrolysis in perchloric acid (8O”C), aliquots were counted in a scintillation counter [13].
AND
CASEIN
EXPRESSION
251
IN MMTC
In vivo prolactin administration. &Fl mice were given two prola&in injections (200 pg contained in 0.05 ml) with an interval of 4 days into the center of the tumors. Controls received vehicle injection. Tumors were removed 3 days after the second injection and RNA was extracted. Tumor transplantation. Cells from three primary mammary tumors were cultured in serum-free medium on Matrigel in the presence or absence of insulin, prolactin, and hydrocortisone for 9 days. The cells (1 X 106) after dispase treatment in 0.1 ml of saline were injected S.C.into 6-week-old male CD8Fl mice. Hybridization studies. Totai RNA was extracted with the guanidine thiocyanate method as described [14]. RNA was quantitated and stored in ethanol at -30°C. An equal amount of RNA was electropboresed in 1.2% agarose gel containing formaldehyde, blotted to Genescreen Plus membrane (New England Nuclear), crosslinked to membrane by uv light, hybridized with [a-32P]dCTP-3abeled probes, and washed as recommendedby the manufacturer. The mouse (I, F and y-casein probes [15] were provided by Dr. J. M. Rosen (Baylor College of Medicine, Houston, TX). The hamster actin cDNA probe 1161 was provided by Dr. F. Smith (Mount Sinai Medical Center, New York). Hybridization with the actin probe was used as a control in each Northern blot assay. The inserts were isolated according to the method described [i4] and the random primed labeling method gave a specific activity of greater than 10” dpm/pg. The autoradiographs were scanned using a laser densitometer $XB-2222-020, ultrascan XL) and the peak areas integrated.
Growth
of
TCs in Serum-Free
atium
Mouse mammary tumor epithehal cehs were obtained from ~MT~-induced mammary tumors i Because our aim was to inv siveness of MMTCs in vitro, to culture the cells in serum-free demonstrated that basement rigel), prepared from the Enge mor, allowed a better gr fmouse mammary epithelial cells in serum-free fibronectin, laminin, type I, or type IV co hus, in the current studies the cells were cultured on atrigel in serum-free /F12 media supplemen insulin. The of MMTCs on Matrigel nt a steady increase over a 5-day culture peri en 0.6 X IO6 cells ay 0, a total of I.8 X PO6cells were harvested on Day 5 (the average of eight observations). As both prolactin and estrogen have been reported ta be mitogenic for mammary tumor epitheliali cells [B7], the two hormones were tested in serum-free medium for their infhrence on MT@ growtk. Tumor cells from mice received pro tin or ~7~-estra~~o~ treatment ulture period and the cell number was deterays 2 and 4 after culture. (41:5 pg/ml, bad lit (P > 0.05, data not shown). were tested, only a Iigbt increase (1.45 number was observe at Day 4 when 100 n was used (Fig. 1). Similar experiments were carried out with FUKU cells, an estabhshed TC line [$I- The growth of
CHOU
I-
1
Control 0.1 nM I nM IO nM 100 nM
ET AL.
ment and 1 X lo6 cells from each tumor cultured in the presence or absence of prolactin and hydrocortisone were injected S.C. into duplicate male CD8Fl mice. Tumors developed in all the mice 6-8 weeks after injection. Modulation
,-
-J 2
DAYS FIG. 1. Growth of MMTCs in the presence of estradiol. Tumor cells from four mice (0.5 X lo6 cells/well) were cultured on Day 0 on Matrigel in the presence or absence of 17@-estradiol in serum-free DMEM/FlB media. Cells were collected on Days 2 and 4. The cell number and viability in duplicate wells were determined with >90% viable.
FUKU cells was not affected by the presence of prolactin (5 ,ug/ml) or estradiol (0.1-100 r&f) (data not shown). Morphology of MMTCs Cultured in the Presence of Prolactin and Hydrocortisone Morphological studies were done on cross sections made from MMTCs cultured on Matrigel for 3, 7, 10, and 14 days in the presence or absence of insulin, prolactin, and hydrocortisone. The tumor cells plated in Matrigel formed aggregated clumps in both groups (Figs 2A and 2B). Fibroblasts were not visible in any of the sections examined. No marked morphological differences were observed after 2 weeks in cultures of hormonetreated or control groups (Figs. 2E and 2F). The tumor cells in the center of the clumps started to die with longer culturing (Figs. 2C and 2D), whereas the cells located at the periphery remained viable and possibly continued to divide. In some cases, the dead cells were disintegrated and this resulted in “gland-like” structures which became more marked after about 10 days in culture (Figs. 2E and 2F). It is obvious from the gradual changes with time that the appearance of the open spaces in the center of the clumps was not due to an artifact created during section preparations. Tumor Transplantation MMTCs from three primary tumors were cultured in insulin-supplemented, serum-free medium on Matrigel for 9 days in the presence or absence of prolactin and hydrocortisone. Cells were harvested after dispase treat-
of Casein Gene Expression
We have demonstrated that MMTCs in 59 out of 60 mammary tumors from CD8Fl mice constituitively express detectable levels of @-casein mRNA as detected by Northern blot assays [6,7]. The levels were less than 2% of those in the normal mammary gland from lactating mice. In the current studies variable levels of (Y- and ycasein messages were also shown to be constituitively expressed in all the primary tumors tested (Fig. 3). It has been demonstrated by others and by us that the in vitro induction of @-casein mRNA requires an interaction between mammary epithelial cells and extracellular matrix in the presence of appropriate hormones [ 181. In the current studies, experiments were performed to investigate the induction and maintenance of CY-,y-, and @casein mRNAs in MMTCs in primary culture. MMTCs from CD8Fl mice were cultured on Matrigel in serum-free medium supplemented with insulin. The cultures were treated with prolactin and hydrocortisone on Day 3, as both hormones, in the presence of insulin, are required for optimal induction of casein mRNA [lo, 191. We have previously shown that the ,fScasein mRNA messages could not be demonstrated in s-day culture in the absence of prolactin and hydrocortisone by Northern blot assays [6, 71. However, when prolactin and hydrocortisone were added on Day 3, the P-casein messages were detectable 12 h after hormonal treatment. The accumulation of @-casein mRNA increased to eightfold at 24 h and were reduced to about fivefold at 48 h (Fig. 4). cu-casein mRNA was still weakly detectable on Day 3 even in the absence of hormonal treatment. Its level was increased about fivefold 12 h after prolactin and hydrocortisone treatment and then accumulated slowly within the next 36 h (Fig. 4). The time-dependent changes in P-casein mRNA levels in response to hormones in our studies appeared similar to previous reports, as assayed by RNA dot-blotting, of DMBA-induced rat mammary carcinoma cells in primary culture [2] or of COMMA1D cells, an established mouse mammary epithelial cell line [20]. However, the peak levels for p-casein mRNA in our serum-free cultures appeared about 24-48 h earlier. The delay in ,&casein mRNA appearance may be attributed to the presence of serum in the previous culture systems, as serum contains hormones or other factors which may exert a prolonged effect on casein mRNA accumulation. Additional experiments were performed to assess the maintenance of casein mRNAs during a 2-week culture period. Hormones were present throughout the culture period and the MMTCs were collected on Days 3, 7, 10, and 14. After 3 days in culture, the three casein messages
HORMONAL
REGULATION
OF GROWTH
AND
CASEIN
EXPRESSION
IN MMTC
253
FIG. 2. H-E stain of MMTCs in primary cultures (X125). MMTCs were cultured on Matrigel for 3(A, B), 7(C), 10(D), or 14(E, F) days in the presence (B, C, D, F) or absence (A, E) of prolactin and hydrocortisone. Tumor cells formed clumps and no fibroblasts were observed. Cells in the center of the clumps died and became disintegrated after 1 week in culture. This resulted in gland-like structures on Day 14.
could still be maintained although the levels showed different degrees of reduction compared to those prior to culture. The highest levels of y-casein mRNA were
detected in MMTCs on Day 14 (Fig. 5) while those of CY-and ,&casein mRNA around dent that the accumulated levels of casein mRNAs in
254
CHOU
ET AL.
4-
12345678 + (I - casein + Y-casein
3-
+ actin 2FIG. 3. Northern blot assay of 01- and y-casein mRNA levels in MMTCs. Total RNA was extracted from FUKU cells cultured in the presence of serum (lane 2) or from freshly prepared tumor samples before culture (lanes 3-8). In lane 1,50 ng of RNA from 7-day lactating mammary gland of normal mouse was loaded as a comparison to the other lanes where 5 pg of RNA was loaded. The RNA was hybridized with the 01-and y-casein probes. The actin probe was hybridized to the blots to quantitate the levels of mRNA in each lane.
l-
O-
IO DAYS
MMTCs showed a time-dependent increase over the 14day incubation period. The peak levels for a-casein mRNA were 3.5-fold greater than those before culture and appeared earlier than those for y-casein messages. Nevertheless, the actin message levels in MMTCs harvested at different times remained the same (data not shown). Immunofluorescence studies using a specific antibody against @casein did not detect /3-casein in primary mammary tumors expressing ,&casein mRNA before culture. In contrast, positive staining of lactating mammary tissues was readily detectable (data not shown). Therefore, characterization of cultured MMTCs using the same antibody was not pursued.
4
Incorporation
To determine whether the induction of the @-casein messages required DNA synthesis, the MMTCs in Day3 cultures received 24 h treatment of prolactin, hydrocortisone, or both, and then [3H]thymidine incorpora-
2 o-o-o 12
In order to detect in viuo response to prolactin treatment, prolactin was injected into the tumor twice with a 4-day interval. Eight days after the first injection, the tumors were removed to extract RNA. Data from six mice indicated that in vivo treatment with prolactin in two out of three mice resulted in a higher level of accumulated P-casein mRNA (Fig. 6, lanes 4, 5) than that seen in three control mice which retained low levels of p-casein mRNA (that received vehicle injections). When FUKU cells were cultivated on Matrigel or floating type I collagen gels, they formed aggregated clumps just as MMTCs from CD8Fl mice in primary cultures. However, when FUKU cells were treated with prolactin and hydrocortisone, no ,8-casein mRNA was detectable by Northern blot assays (data not shown). Thymidine
./
b 01 0
FIG. 5. Maintenance of casein mRNA levels during a 2-week culture. MMTCs from three tumors were cultured separately in serumfree medium supplemented with insulin, prolactin, and hydrocortisone. Total RNA from MMTCs prior to culture (control) or cultured for 3, 7, 10, and 14 days was analyzed by Northern blotting. The blots were hybridized and the results analyzed as described in Fig. 4 except that y-casein probe was also used in rehybridization.
24
36
48
TIME (hr) FIG. 4. Time-dependent induction of LY-and fl-casein mRNAs in MMTCs. MMTCs from CDSFl mice were cultured on Matrigel in serum-free medium supplemented with insulin. Prolactin and hydrocortisone were added on Day 3. Total RNA in three experiments was extracted from cells on Day 3 before hormonal treatment (0 time control) or 12 h, 24 h, and 48 h following hormonal administration. The Northern blots were hybridized with a- and fl-casein and actin probes. The bands were scanned using a laser densitometer. The curves represent relative changes in the three message levels after hormonal treatment to those before hormonal exposure which were arbitrarily defined as one unit.
123456 * actifl M/3-casein FIG. 6. In vivo changes of p-casein mRNA. Three mice were injected twice with prolactin into the tumors. Tumors from three mice receiving two vehicle injections were used as a control. Two micrograms of total RNA was blotted and hybridized with /3-casein and actin probes. Lanes 1-3, control; lanes 4-6, prolactin-treated.
HORMONAL
TABLE Effects of Prolactin, [3H]Thymidine
REGULATION
OF GROWTH
1
Hydrocortisone, Incorporation
Treatment” Experiment 1 None Prolactin Hydrocortisone Prolactin and hydrocortisone Experiment 2 Prolactin and bydrocortisone Ara-C plus prolactin and hydrocortisone
and Ara-C into MMTCs
on
cpm/106 cells ( X10m3)b
27.5 f 2.4 32.6 t 2.6 28.12 4.5
27.2 f 3.5 35.2 -i 0.9 0.5 + 0.3
e MMTCs were cultured on Day 0 at 0.5 X lo6 cells/well on Matrigel. In Experiment 1, prolactin (5 pg/ml), hydrocortisone (1 pg/ml), or both were added on Day 3, and [3H]thymidine was then added. In Experiment 2, MMTC cultures, initially treated with or without Ara-C (30 pg/ml), were exposed to prolactin and hydrocortisone for 24 h. In all cases, the cells were harvested on Day 4 after exposure to [3H]tbymidine for 23 b to determine cell number and thymidine incorporation. b The mean (iSD) of four observations are shown.
tion was determined. As shown in Table 1, none of the three different kinds of treatment resulted in any significant changes in 13H] thymidine incorporation (P > 0.05). The cell number in each experiment remained the same in the different treatment groups (data not shown). In contrast, cytosine arabinoside (Ara-C) at 30 more than 95% of the [3H]thymidine inle 1). However, MMTCs in the Ara-Ctreated and nontreated cultures displayed similar levels of the /3-casein messages after 24-b exposure to prolactin and hydrocortisone (Fig. 7). This demonstrates that the hormonal induction of ,&casein gene expression in the MMTCs does not require DNA synthesis, which agrees with previous observations on casein synthesis in preneoplastic mouse mammary tissues [21].
The aim of the current studies was to observe hormonal modulation of mammary-specific gene expression in MMTCs in primary culture and to look for a correlation between the induced differentiated phenotype and tumor cell growth potential in vitro and in uiuo. Autochthonous mammary tumors from CD8Fl mice were employed in the current investigation. The tumor cells are known to be independent of ovarian steroids for growth in viva from the following observations: the tumors continue to grow after ovariectomy; the transplanted tumor cells display the same growth rate in male or female recipients; and tamoxifen can not inhibit the growth of transplanted tumors (unpublished data). In the current in vitro studies, it was further proven that estradiol, up to 100 niW, did not stimulate the growth of MMTCs in serum-free medium supplemented with insu-
AND
CASEIN
EXPRESSION
IN MMTC
255
lin. Furthermore, when the same tumor cells were used, prolactin did not increase the cell number or dine incorporation at the ~on~e~tratio~s that sein gene expression The unresponsiveness of th ture to prolactin or estrogen trasted by the in vitro indueib mRNAs in the cells. The levels of casein dramatically reduced on Day 3 after cult free medium on Matrigel in the absence of hormones. However, 24 h treatment with pro sone in the presence of ins n was sufficient to elicit a marked increase in tbe site -state levels of CY-and flcasein mRNAs. This indi presence of insulin and hydroc at least a subpopucumulation of the casein messages lation of the tumor cells. If the three hormones were continuous1 ing a 2-week culture, the levels of mRNAs could be maintained and s dent increase after 10 days in cul whether this increase with time was d 50 am accumulation of casein messages in a subs or due to an expansion of easein. ue to the death of tively, this increase may simply those cells which synthesized lower levels sages. Tbis possibility is mor servations that the cells in the die after 1 week in culture while those closer trigel were still viable. In this regard in robes will be require verify the possibility. Although appropriate and maintained casein m alter the rate of cell grow tion. Cells cultured in the presence or absence of hormones for 2 weeks did not look different morphologically. The in vluo data from t~rnorige~~~ studies employresence of insulin, ing MMTCs cultured in the rocortisone for prolactin, and those from con groups, while c mone-treated primary culture increased with time.
1
2
FIG. 7. Effect of Ara-C treatment on @-casein mRNA levels. MMTCs in primary cultures received 24-b treatment ofprolactin and hydrocortisone in the absence (lane 1) or the presence (lane 2) of AraC (30 pg/ml). Two micrograms of total RNA was loaded in each lane and hybridized with p-casein and actin probes.
256
CHOU
These results suggest that the induction and maintenance of casein mRNAs in the tumor cells by a single lactogenic hormone treatment was not correlated with tumor cell growth in vitro or tumor-producing capacities in uivo. Consistent with this, a lactogenic stimulus of mouse mammary preneoplastic cells did not appear to change the tumor potential in two out of three outgrowth lines but inhibited the tumorigenic potential in one line [22]. It remains to be determined whether a prolonged lactogenic stimulus in our tumor model will alter in ho tumor growth potential. A number of agents known to affect growth and differentiation in established tumor cell lines can be studied in this primary MMTC culture system since it should provide a proliferating tumor cell culture with differentiation-associated phenotypes. We have studied several agents in this system including HMBA, retinoids, and transforming growth factors (TGFs). Thus far, only TGF-a! was shown to modulate ,&casein mRNA levels in the presence of prolactin and hydrocortisone (manuscript in preparation). We thank Dr. G. Coppola and Ms. C. Schreiber for their excellent technical assistance. We also thank R. Hromada for preparing the manuscript. The work was supported by grants from the Chemotherapy Foundation, the Samuel Waxman Cancer Research Foundation and the Herman Goldman Foundation; and by USPHS Grant ROICA 25842 to RLS and DSM.
REFERENCES 1.
Suposit, 1360.
S. C., and Rosen, J. M. (1982) Cancer Res. 42, 1355-
2.
Johnson, Dev.Biol.
M. L., Levy, J., and Rosen, J. M. (1985) In Vitro Cell. 2 1,439-444.
3.
Edery, M., Houdebine, L. M., Djiane, J., and Kelly, Mol. Cell. Endocrinol. 34, 145-151.
Received May 24,1989 Revised version received August 21, 1989
P. A. (1984)
ET AL. 4.
Dao, T. L., Bock, F. G., and Greiner, J. J. (1960) J. Natl. Cancer Inst. 25,991-1002.
5.
Grubbs, C. J., Hill, D. L., McDonough, K. C., and Peckham, J. C. (1983) J. Natl. Cancer Inst. 71,625-628.
6.
Chou, J. L., Coppola, G., Shen, Z. X., Schreiber, C., and Waxman, S. (1988) Proc. Amer. Assoc. Cancer Res. 29,42 [No. 1671
7.
Chou, J.-L., Shen, Z.-X., Stolfi, R. L., Martin, man, S. (1989) Cancer Res. 49,5371-5376.
8.
Yagi, M. J. (1973) J. Natl. Cancer Inst. 51,1849-1855.
9.
Stolfi, R. L., Martin, D. S., and Fugmann, Chemother. Rep. 55,239-251.
D. S., and Wax-
R. A. (1971) Cancer,
10.
Emerman, J. T., Enami, J., Pitelka, Proc. Natl. Acad. Sci. 74,4466-4470.
11.
McGrath, M., Palmer, 1255,182-191.
12.
Asch, B. B., Medina, D., Kretzer, F., Connolly, ley, B. R. (1979) Cancer Res. 39,893-907.
13.
Shamay, A., and Gertler,
D. R., and Nandi,
S., and Nandi,
S. (1977)
S. (1985) J. Cell. Physiol. J. L., and Brink-
A. (1986) Cell Biol. Int. Rep. 10, 923-
929. 14.
Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
15.
Gupta, P., Rosen, J. M., D’Eustachio, (1982) J. Cell Biol. 93, 199-204.
16.
Dodemont, H. J., Soriano, P., Quax, W. J., Ramaekers, F., Johannes, A. L., Groenen, M. A. M., Bernardi, G., and Bloemendal, H. (1982) EMBO J. 1,167-171.
P., and Ruddle,
F. H.
17.
Welsch, C. W. (1985) Cancer Res. 45,3415-3443.
18.
Wicha, M. S., Lowrie, G., Kohn, E., Bagavandoss, T. (1982) Proc. N&l. Acad. Sci. 79,3213-3217.
19.
Hobbs, A. A., Richards, D. A., Kessler, D. J., and Rosen, J. M. (1982) J. Biol. Chem. 25,3598-3605.
20.
Rosen, J. M., Rodgers, J. R., Couch, C. H., Bisbee, C. A., DavidInouye, Y., Campbell, S. M., and Yu-Lee, L.-Y. (1986) Ann. N. Y. Acad. Sci. 478,63-76.
21.
Smith, G. H., Vonderhaar, B. K., Graham, D. E., and Medina, D. (1984) Cancer Res. 44,3426-3437.
22.
Medina, D., Schwartz, M., Taha, M., Oborn, C. J., and Smith, G. H. (1987) Cancer Res. 47,4686-4693.
P., and Mahn,
