1985;45:2608-2615. Published online June 1, 1985.Cancer Res Susan C. Aitken, Marc E. Lippman, Attan Kasid, et al. Mammary Tumor CellsGenes and Estrogen-stimulated Proliferation of MCF-7 Relationship between the Expression of Estrogen-regulated
Updated Version http://cancerres.aacrjournals.org/content/45/6/2608
Access the most recent version of this article at:
Relationship between the Expression of Estrogen-regulated Genes andEstrogen-stimulated Proliferation of MCF-7 Mammary Tumor Cells1
Susan C. Aitken, Marc E. Lippman, Attan Kasid, and Daniel R. Schoenberg2
Department of Pharmacology, Uniformed Services University of the Health Sciences, Bethesda, Maryland 20814-4799 [S. C. A., D. R. S.], and Medical Breast Cancer
Section, Medicine Branch, National Cancer Institute, Bethesda, Maryland 20205 [M. E. L, A. K.]
ABSTRACT
The growth of MCF-7 cells was arrested by 24 h of isoleucine
deprivation. Following replenishment of the medium, the incorporation of uridine and thymidine into trichloroacetic acid-precip-
itable material began to increase slowly and gradually rose tothe level of cycling cells. The addition of 5 x 10~9 M estradici to
growth-arrested cells dramatically shortened the time of onset
of macromolecular synthesis and increased the overall amountof precursor incorporation 2- to 4-fold over the level obtained by
arrested control cells. The increase in uridine incorporation preceded the increase in thymidine incorporation by 6 h. Inhibitionof protein synthesis with cycloheximide blocked the recovery ofmacromolecular synthesis in both control and estrogen-treatedcells. Actinomycin D was ineffective in blocking the estrogen-
stimulated recovery of macromolecular synthesis at concentrations known to inhibit pre-rRNA synthesis (10~8 M). At higher
concentrations, uridine and thymidine incorporation were inhibited in a dose-dependent manner. Inhibition of RNA polymeraseII activity with a-amanitin similarly blocked both the recovery ofthe cells from isoleucine starvation and the potentiation of thisby estradici. DihydrofoÃatereducÃaseand thymidine kinase activities are both stimulated by estradici in MCF-7 cells. In cyclingcells, estrogen stimulates a 2-fold increase in their messenger
RNAs (mRNAs) within 24 h. The level of dihydrofolate reducÃasemRNA is unaffected by isoleucine starvation, and estrogencaused no change in dihydrofolate reducÃasemRNA levels overa 24-h period following reversal of growth arrest. Similar resultswere observed for the 600-nucleotide pS2 mRNA lhal has beenidentified as an estrogen-induced RNA in MCF-7 cells. In con
trast, thymidine kinase mRNA was found to be increased byestrogen at 24 h, but not at 12 h, following reversal of growtharrest. This increase correlates with increases in thymidine, butnot uridine incorporation. These data indicate that the estrogen-
stimulated increase in thymidine incorporation following releasefrom growth arrest is dependent on new RNA synthesis. However, the hormone did not increase the levels of three estrogen-
regulated mRNAs coordinately with the increases observed inuridine incorporation.
INTRODUCTION
It is a tenet of human oncology that a proportion of mammarytumors respond to estrogen with increased growth. This hasprompted the application of estrogen receptor assays to clinical
1This work was supported in part by Protocol C07533 from the Uniformed
Services University of the Health Sciences.2To whom requests for reprints should be addressed, at the Department of
Pharmacology, Uniformed Services University of the Health Sciences, 4301 JonesBndge Road, Bethesda, MD 20814-4799.
Received 8/28/84; revised 11/21/84, 2/14/85; accepted 2/20/85.
isolates so as to assess the likelihood of success of endocrinetherapy in a given patient. It has also stimulated research into anumber of cell and tissue systems in an attempt to developsuitable models for hormone-responsive mammary tumors. Ofthose presently under investigation, the MCF-7 mammary tumorcell line has proven especially useful. MCF-7 and other mammary
tumor cell lines respond to estrogen with increases in growth(1-7), the induction of the progesterone receptor (8), the induction of intracellular (9) and secreted glycoproteins (3, 10-12),
and the transcription of new mRNAs (13,14). We are particularlyinterested in identifying the genes and gene products responsiblefor the estrogen-induced growth of these cells.
At the outset of this study, it was clear that a critical approachto identifying estrogen-regulated genes involved in cellular hy-
perplasia should use some form of cell cycle manipulation. In thismanner, it was hoped that estrogen might alter the manner inwhich an arrested cell population would enter S phase. A previous paper (15) described a method to partially synchronizeMCF-7 cells by maintenance in isoleucine-free medium for 30 h.
Reversal consisted of changing to complete medium, followingwhich thymidine incorporation began to increase after another30 h. We chose to use this method of growth arrest to determinethe effects of estradici on the onset of DNA synthesis and onthe expression of 3 estrogen-regulated mRNAs: DHFR3 (16);
pS2(14), andTK(17).
MATERIALS AND METHODS
Cell Culture. MCF-7 cells (18) were maintained at 37°Cin 95% air-5% CO2 in monolayer culture in Richter's IMEM (NIH Media Unit) supple
mented with 5% fetal bovine serum, gentamicin (40 mg/liter), and gluta-
mine (0.6 g/liter). Prior to each experiment, cells were grown for 2passages in IMEM (containing isoleucine) supplemented with 2.5% charcoal-treated calf serum and 10~7 M insulin (IMEM + isoleucine). For each
experiment, cells were replicately plated at a density of 2 x 105/well in
6-well culture dishes. After 48 h, the medium was changed to IMEM +isoleucine for the controls or to IMEM lacking isoleucine (IMEM -
¡soleucine)to obtain arrest growth (19). The metabolic block was reversed 24 h later by replacing the medium with IMEM + isoleucinecontaining the indicated hormone, drugs, or vehicle control. To determinethe rate of precursor incorporation into nucleic acids, the cells were pulselabeled for 2 h prior to harvest with 1 to 5 ¿<Ciof radioactive thymidineor uridine. Cells were harvested with 1 ml of 0.04% EDTA in phosphate-
buffered saline and collected by a brief centrifugaron. Cell pellets werestored at -20°C prior to analysis.
Analytical Procedures. Cell pellets were suspended in 1 ml of H-,O
and were briefly sonicated. Samples were removed for determination ofprotein by the method of Lowry ef al. (20) and for fluorometric determination of DNA (21). Precursor incorporation into RNA and DNA was
3The abbreviations used are: DHFR, dihydrofolate reducÃase; IMEM, improvedminimal essential medium; TK, thymidine kinase; SSC, 0.15 M NaCI-0.015 M sodium
determined by liquid scintillation counting of 10% TCA-precipitable ma
terial on glass fiber filters.RNA Dot Blot Analysis. Cells were harvested in Hanks' balanced salt
solution containing 0.04% EDTA and cycloheximide (50 /ig/ml). The cellpellet was collected by brief centrifugation in a microfuge, following whichit was washed with cold Hanks' salts containing cycloheximide to remove
EDTA. After another brief centrifugation, the cells were suspended in 90M! of cold Hanks' salts containing cycloheximide and 10 mw vanadyl
ribonucleoside complex. Five n\ each of 10% (w/v) sodium deoxycholateand Brij 35 were added followed 2 min later by 10 //I of proteinase K (2mg/ml). The tubes were placed in a 37°Cwater bath for 30 min and thenstored at -70°C until the time of assay.
The dot hybridization assay was essentially that of Bresser et al. (22).An equal volume of supersaturated Nal was added to each thawedsample, and appropriate dilutions were prepared in saturated Nal. Eachsample was filtered onto BA-85 nitrocellulose (Schleicher and Schuell)that was presoaked for at least 24 h in 6x SSC (1x SSC = 0.15 M NaCI-
0.015 M sodium citrate). The soaking was necessary for efficient binding.The filter washing, prehybridization, and hybridizations were as described(22). After the hybridization, filters were washed 3 times for 1 h at 68°Cin 2x SSC-0.1% sodium dodecyl sulfate and once for 1 h at 68°Cin 1x
SSC-0.1% sodium dodecyl sulfate. Autoradiography was performed at-70°C with Kodak X-OMat AR film and Dupont Cronex lightning-plus-
intensifying screens.The DHFR clone pJC201 consisted of a human intronless pseudogene
cloned into pBR322 (23). This was a gift from Dr. Arthur Nienhuis of theNIH. The hybridization probe consisted of the 4-kilobase insert excised
from the plasmid by digestion with EcoRI and radiolabeled by nicktranslation. Clone pS2 (14) was a gift from Dr. Pierre Chambón, Institutde Chimie Biologique, Strasbourg, Strasbourg, France. The human TKprobe (24) was a gift from Dr. Harvey Bradshaw, Louisiana StateUniversity Medical Center, New Orleans, LA.
Isotopes, Enzymes, and Reagents. All restriction enzymes andproteinase K were obtained from Bethesda Research Laboratories.Ultrapure Nal was purchased from Alfa-Thiokol. Cycloheximide, actino-
mycin D, common salts, and detergents were purchased from Sigma.«-Amanitin was purchased from Boehringer-Mannheim, and 17/3-estra-diol was purchased from Steraloids, Inc., Pawling, NY. [mef/)y/-3H]Thy-midine (40 Ci/mmol), [L/-'4C]uridine (248 Ci/mmol), and [5-3H]uridine (28Ci/mmol) were purchased from New England Nuclear. [a-MP]dATP (2000
to 3000 Ci/mmol) was purchased from Amersham.
RESULTS
Kinetics of DMA and RNA Synthesis. In the initial experiments, we repeated the study of Jakesz er a/. (15) on therecovery of DNA synthesis following a period of isoleucine deprivation in order to determine what effects estradici might haveon this process. The metabolic block was produced by maintaining cells for 24 h in IMEM - isoleucine. The experiment was
initiated (time zero) by replacing the medium with IMEM +isoleucine ±5 x 10~9 M estradici. The cells were then harvestedat the indicated times following a 2-h pulse with [3H]thymidine or[14C]uridine. The data in Chart 1 confirm the previous observation
that the rate of thymidine incorporation into TCA-precipitable
material increases slowly after the reversal of the metabolicblockade, reaching the levels obtained by cycling cells between12 and 24 h. Changes in the rate of [14C]uridine incorporation
into RNA parallel this for the most part and preceded changesin [3H]thymidine incorporation by 4 to 6 h.
The inclusion of 5 x 10~9Mestradiol caused a dramatic changein the rates of macromolecular labeling. The rate of [14C]uridine
incorporation began to rise before 6 h, and by 15 h, this increased
C3J
gzm5T)
OU>X
36 48
TIME (hours)
Chart 1. Time course of nudeotide incorporation following 24 h of isoleucinestarvation. Cultures were grown as described in "Materials and Methods." Twenty-
four h prior to the start of the experiment, the medium was replaced with IMEM +isoleucine ±5 x 10~* M estradiol (E2)(cycling control cultures) or IMEM - isoleucine
(arrested). At time zero, the cycling cultures were harvested following a 2-h pulsewith [3H]thymidine (1 /¿Ci/dish)or [uC]uridine (0.6 >iCi/dish), and the medium wasreplaced on the arrested cultures with IMEM + isoleucine ±5 x 10"' M estradiol.
Cells were harvested at the indicated times, and the relative rates of nucleotideincorporation were determined by TCA precipitation of total cellular homogenates.Samples were removed for DNA and protein analysis (Table 1). Points indicate theamounts of undine (right axis) and thymidine (left axis) incorporation in cyclingcontrol cultures (+ILE) and in cultures maintained for 24 h in IMEM - isoleucine(-ILE). The graph represents the time course for recovery of undine (A, A) and
thymidine (O, •)incorporation following replenishment with IMEM + isoleucine.The closed symbols are the control cultures, and the open symbols are for culturescontaining estradiol. Points, mean; oars, SE (n = 3).
to a level 4 times that of cycling control cultures. Similar changesoccurred in the rate of [3H]thymidine incorporation into DNA.
This latter response succeeded the alterations in the rate of[14C]uridine incorporation by 6 h.
In these first experiments, the data are presented on a perdish basis because the actual amounts of DNA and proteinbegan to change after 24 h of estrogen stimulation. The datapresented in Table 1 detail the changes that occurred in DNAand protein content during the 48-h time course shown in Chart
1. No change occurred in the DNA content of control cultures;however, the amount of DNA began to increase after 24 h ofestrogen treatment and continued to increase to an amount 1.5times that of the control culture by 48 h. Similar changes occurredin the total amount of protein per dish, where there was littlechange in the control cultures, and estradiol caused a 1.5-foldincrease within 48 h. It should be noted that the estradiol-
stimulated changes in the rates of undine and thymidine incorporation observed in Chart 1 preceded gross changes in theamount of DNA and protein.
It was apparent from the data shown in Chart 1 that the majorchanges in macromolecular labeling occurred within the first 24h, so the time course was repeated with more frequent samplingintervals (Chart 2). Within 24 h following the change to IMEM +
Table 1EHect of estradiol on DNA and protein content ot arrested cultures of MCF-7 cells
Cells were plated at 2 x 10*/dish in IMEM, and after 48 h, the medium waschanged to IMEM - ¡soleucine.After an additional 24 h, the medium was changedto IMEM + isoleucme ±5 x 10 9 M estradiol. Cultures were harvested at the
indicated times, and protein and DNA assays were performed as described in'Matehals and Methods." Statistical analyses were performed using Duncan's
" Mean ±SE for triplicate determinations.b Significant differences between time zero and succeeding time points at P <
0.05.c Significantdifferences between control and estrogen treatment at P < 0.05.
250
200
o
X150
I100
p
*0<i
-ILE
-ILE
l—
Control-ILE
10
8 o
2D
I
4 genX
12
TIME (hours)
18 24
Chart 2. Time course of nucleotide incorporation following 24 h of isoleucmestarvation. The experiment in Chart 1 was repeated over a shorter interval (24 h)with more frequent sampling times. ["HJThymidine (2.5 /*C¡)and [14C]uridine (0.25
//CO were added to each dish. Closed symbols represent the rates of undine (A)and thymidine (•)incorporation in control cultures. Open symbols represent therate of undine (A) and thymidine (O) incorporation in cultures supplemented with 5x 10"' M estradiol (E2).Points, mean; oars, SE (n = 3).
isoleucine no change occurred ¡nthe rate of [14C]uridine incor
poration in the control cultures. In those cultures that received 5x 1CT9M estradiol, the rate of [14C]uridine incorporation began
to rise after 8 h and continued to rise to a maximum at 20 h. Inthis experiment, the maximal rate of uridine incorporation was50% greater than that observed in cycling control cultures. Onceagain, changes in the rate of [3H]thymidine incorporation followed
the changes in the rate of uridine incorporation by 6 to 8 h and,in this case, reached the same level as that observed in cyclingcontrol cultures. Therefore, the data in Charts 1 and 2 and Table1 indicate that estradiol caused substantial changes in the timeat which cells arrested by isoleucine deprivation enter S phase,
in the relative rates of uridine and thymidine incorporation obtained within 48 h following reversal of arrest, and in the amountsof DNA and protein per culture.
Effects of Inhibitors of RNA and Protein Synthesis. Thepreceding data were suggestive of a mechanism by which estrogen stimulated the synthesis of RNAs and/or proteins that mightalter the processes by which cells recovered from the metabolicblock and proceeded into S phase. This possibility was furtherinvestigated by the use of specific inhibitors of protein and RNAsynthesis. In the first such experiment, the estradiol-inducedchanges in thymidine and uridine incorporation were studied inthe presence of the protein synthesis inhibitor cycloheximide.Cells were arrested by maintaining them for 24 h in IMEM -
isoleucine, and they were reversed by changing the medium toIMEM + isoleucine ±5 x 10~9 M estradiol ±varying concentrations of cycloheximide between 2 x 10~8 and 10~5 M. The rates
of uridine and thymidine incorporation were assayed at 16 and24 h, because the data in Chart 2 indicated that significantdifferences with respect to incorporation into RNA and DNAwere likely to be detected between estrogen and control groups.Moreover, 16 h (RNA) and 24 h (DNA) represented fairly earlytime points in the response of cells to estrogen. Cycloheximideat all concentrations greater than 10~7 M inhibited precursor
incorporation into DNA and RNA in a dose-dependent manner(data not shown). In order to simplify the presentation of results,only the data for 10"6 M cycloheximide are shown here. The data
in Chart 3, A and B (16 and 24 h after reversal, respectively),indicate that cycloheximide blocked increases in thymidine anduridine incorporation in both control and estradiol-treated cultures (Chart 3, A and B, respectively). The rates of macromolec-
ular labeling observed in the presence of cycloheximide wereidentical to the basal rates obtained with cells arrested for 24 hin IMEM - ¡soleucineand assayed at time zero. Therefore, it is
evident that recovery from isoleucine deprivation and the estrogen-stimulated enhancement of that recovery both require pro
tein synthesis.Actinomycin D is a well-characterized inhibitor of RNA synthe
sis. At 10~8 M, it can block the synthesis of pre-rRNA (25, 26),and at 10'7 M, it effectively inhibits all classes of RNA synthesis.
We therefore used this drug as an agent to determine if, ingeneral, RNA synthesis is required for the estrogen-stimulatedincreases in thymidine incorporation. MCF-7 cells were arrestedby maintaining cultures for 24 h in IMEM- isoleucine and at time
zero, the medium was replaced with IMEM + isoleucine ±5 x10~9 M estradiol and increasing concentrations of actinomycin D(10~8 to 10~6 M). Once again, cells were pulse labeled for 2 h
prior to assay with [3H]thymidine and [14C]uridine, and the assays
were performed at 16 and 24 h so as to coincide with the majorincreases in macromolecular labeling observed in Chart 2. Thedata shown in Chart 4, A and 8 (16 and 24 h, respectively),indicate that 10"8 M actinomycin D has no effect on the incor
poration of thymidine or uridine in control cultures, and it alsohas no effect on the stimulation of nucleotide incorporationproduced by estradiol. Increasing the concentration of actinomycin D to 10 7 M caused a pronounced inhibition in nucleotide
incorporation in both the control and hormone-treated culturesin a concentration-dependent manner. Concentrations of 10~6 M
actinomycin D proved to be cytotoxic. These data indicate thatthe recovery of cells from isoleucine deprivation and the estradici-
stimulated increases in thymidine incorporation are dependent
Chart 3. Effect of Cycloheximide on nucleotide incorporation. Cells were grown to a density of 8 x TCP/well as described in "Materials and Methods," and 24 h prior
to the beginning of the experiment, the medium was replaced with IMEM - isoleucine. At time zero, the medium was changed to IMEM + isoleucine containing vehiclecontrol, 10"* M Cycloheximide, 5 x 10"* M estradici, or estradici + Cycloheximide. The base-line incorporation at time zero (-ILE) is indicated on each axis for uridine and
thymidine incorporation. Cells were harvested at 16 h (A) and 24 h (B), and the TCA-precipitable radioactivity was assessed. Columns, mean; bars, SE (n = 3).
on RNA synthesis.In the final series of inhibitor experiments, we examined the
effect of a-amanitin on the recovery from isoleucine starvationand on the estrogen-stimulated increases in thymidine and uridine incorporation. At a concentration of 10~7 M, a-amanitin
inhibits the activity of RNA polymerase II by 50% (mRNA, heterogeneous nuclear RNA) (27, 28). Concentrations of inhibitorexceeding 10~5 M block RNA polymerase III (tRNA, 5S RNA),
while RNA polymerase I (rRNA) activity is unaffected by a-
amanitin. In this experiment, cells were arrested for 24 h in IMEM- isoleucine and reversed at time zero in the presence of 10~7M a-amanitin ±5 x 10~9 M estradici. Cells were then harvested
at 16 and 24 h to determine the rate of macromolecular labeling.The data in Chart 5A (16 h) indicate that a-amanitin inhibited
the recovery in [3H]thymidine labeling of control cultures some
what but reduced the rate of thymidine incorporation in theestradiol-treated cultures to the level found in the nonestrogen-
ized controls. These data suggest that an RNA polymerase IIproduct is required for the stimulation of thymidine incorporationabove control levels. Examination of ["CJuridine incorporation
indicates a similar result. The 50% decrease in the rate of uridineincorporation is not surprising, as a-amanitin at this concentra
tion should only affect mRNA synthesis, and the dosage used is50% inhibitory of RNA polymerase II. Similar results were obtained at 24 h (Chart 5B) with the exception that the rates ofmacromolecular labeling were greater in the control and estrogen-treated cultures. Although indirect, we believe that these
data and those obtained with actinomycin D and Cycloheximide
indicate that the estradiol-stimulated changes in the overall rateof [3H]thymidine incorporation and the accelerated initiation of
DNA synthesis caused by the hormone are the result of specificestrogen-stimulated mRNA synthesis.
Effect of Estradiol on the Levels of DHFR, TK, and pS2mRNA. DHFR is a key enzyme of DNA synthesis, the activity ofwhich is stimulated by estradici in MCF-7 cells (16). We therefore
chose to study the effects of estradici on the levels of this mRNAto determine: (a) whether growth arrest produced by isoleucinestarvation decreased the levels of DHFR mRNA; (b) whetherestradici increased DHFR mRNA in randomly cycling cells; and(c) whether the level of DHFR mRNA reflected the observedchanges in the overall rate of uridine incorporation.
To implement these studies, triplicate cultures of MCF-7 cellswere seeded into 6-well dishes and grown to a density of 8 x105/well. At this point, the medium was replaced with fresh IMEM
+ isoleucine ±estradici on the control cultures. The reason forthis was to eliminate possible potentiation of the hormone effectsby growth factors known to be secreted by these cells. Thecontrol cells were harvested 24 h later for determination of cellnumber and DHFR mRNA. The remaining cultures were transferred to IMEM - isoleucine for 24 h, after which the mediumwas changed to IMEM + isoleucine ±5 x 10~9 M estradiol. Cells
were harvested at the times indicated in Fig. 1C for DHFR mRNAdetermination, and cell number determinations were performedat time zero and 24 h for both the control and growth-arrested
cells. No change in cell number was observed in either thecontrol or arrested cultures over the period of time chosen for
Chart 4. Dose-response curve for actinomydn D inhibition of nucleotide incorporation. Cells were arrested by 24 h of isoleucine starvation. At time zero, the mediumwas replaced with IMEM + isoteucine (ILE) ±5 x 10"* M estradici (Ei) containing concentrations of actinomycin D ranging from 10"" to 10"* M. Cells were harvested at
16 h (A) and 24 h (8) for determination of the rate of nucleotide incorporation in a 2-h pulse. Cultures that received estradici are shown in the open symbols, and control
cultures are shown in the closed symoo/s. The data for undine incorporation (O, •)and thymidine incorporation p, •)represent the mean of triplicate determinations.
this experiment.DHFR mRNA was determined by the "quick blot" procedure
that involves the selective immobilization of mRNA to nitrocellulose in the presence of detergents and saturating concentrationof Nal. The experiment presented in Fig. "M shows that no
change occurred in the levels of DHFR mRNA in cells that wererandomly cycling or arrested by 24 h of isoleucine deprivation.These data confirm the results of Collins ef al. (29) in which theyobserved that the levels of total cellular DHFR mRNA remainunchanged in isoleucine starved mouse fibroblasts. They alsoobserved that the metabolic block resulted in delayed processingof nuclear DHFR transcripts. This phenomenon was not examined in the present series of experiments.
We next sought to determine if the estradiol-induced increase
in DHFR activity in cycling cells reported by Cowan ef a/. (16) isthe result of an increase in DHFR mRNA. The data in Fig. 10show that, after 24 h, estradici induced a 2-fold increase in the
level of DHFR mRNA. Quantitative analysis was obtained bydensitometric scanning of the autoradiograms and liquid scintillation counting of the dots.
Then, we sought to determine if estradici induced increasesin the levels of DHFR mRNA that might coincide with the hormone-induced changes in uridine and thymidine incorporation ingrowth-arrested cells. The time-course analysis for DHFR mRNA
levels in Fig. 1C shows that no change occurred in the level ofthis mRNA over the 24-h period following the change from IMEM- isoleucine to IMEM + isoleucine, regardless of the presenceof the hormone. Therefore, the early estrogen-induced changes
in uridine and thymidine incorporation observed at 8 and 14 h,
respectively (Chart 2), are uncoupled from changes in the levelsof DHFR mRNA. We next examined the same extracts forchanges in the levels of the 600 nucleotide estrogen-induced
mRNA that corresponds to clone pS2 isolated from a cDNAlibrary of MCF-7 cells by Masiakowski ef al. (14). Like DHFRmRNA, estradici induced a 2-fold increase in pS2 mRNA over24 h in cycling cells. However, no estradiol-induced change wasobserved in the level of pS2 mRNA over the 24-h period following
reversal of metabolic blockade (data not shown).Finally, we examined changes in TK mRNA, reasoning that
increased thymidine incorporation might be coupled to increasedTK mRNA. It has been shown that TK activity is stimulated byestradici in MCF-7 cells (17), and recent data from one of ourlaboratories have shown that estradici causes a 2-fold increasein TK mRNA in cycling cells.4 In the experiment shown in Fig. 2,
cells were arrested by isoleucine starvation and reversed in thepresence of estradiol. No change in TK mRNA was observableby 12 h, when uridine incorporation was increasing (Chart 2).However, at 24 h, a 2-fold elevation of TK mRNA was observed
that correlates with the maximum rate of thymidine incorporationcaused by estradiol.
DISCUSSION
This paper defines a new approach to studying the estrogen-stimulated proliferation of MCF-7 mammary tumor cells thatutilizes a nonhormonal method of growth arrest, followed by
4A. Kasid and M. E. Uppman, manuscript in preparation.
Chart 5. Effects of a-amanitin on nucleotide incorporation. Cells were prepared as described in Chart 3. At time zero, the medium was replaced with IMEM +isoleucine containing vehicle control, 10~7 M a-amanitin, 5 x 10"* M estradiol, or both estradici and a-amanitin. The base-line incorporation (-ILE) at time zero for a 2-h
labeling is indicated on each axis. Cells were harvested at 16 h (A) and 24 h (B), and the TCA-precipitabte radioactivity was assessed. Columns, mean; oars, SE (n = 3).
reversal in the presence of estrogen. In this, we have confirmeda previous observation (15) on the use of isoleucine starvationto effect growth arrest and have demonstrated that estrogencan substantially shorten the time required for arrested cells toresume DNA synthesis. The experiment shown in Chart 1 demonstrates that estradici doubled the rate of incorporation ofundine and thymidine into TCA-precipitable material, and thehighest estrogen-stimulated levels exceeded the rate of labeling
obtained with randomly cycling cells. To define better the natureof estrogen-stimulated changes in macromolecular labeling, thesame experiment was repeated over a 24-h time course with
shorter sampling intervals (Chart 2). In this experiment, estradicistimulated undine incorporation to increase by 8 h, and thiscontinued throughout the 24-h interval. The increases in uridine
incorporation were paralleled 6 h later by similar increases inthymidine incorporation. Thus, estrogen induced changes in theamount of precursor incorporation into nucleic acids, but moredramatically, it altered the kinetics at which this response occurred.
To determine whether the hormonally stimulated increases inthymidine incorporation were dependent on the synthesis of newproteins or RNAs, we turned to the use of the inhibitors cyclc-heximide, actinomycin D, and a-amanitin. Both cycloheximideand actinomycin D have been shown to inhibit estradiol-stimu-lated changes in plasminogen activator activity in MCF-7 andZR-75-1 cells at concentrations known to effectively block pro
tein and total RNA synthesis (10,11 ). The data in Chart 3 indicatethat a concentration of cycloheximide sufficient to inhibit protein
synthesis effectively blocked the recovery of thymidine and uridine incorporation following isoleucine deprivation, and it alsoeliminated the estrogen-stimulated changes in the kinetics of
recovery. This is not surprising, because the metabolic blockproduced by isoleucine deprivation is likely to involve an inhibitionin protein synthesis, and cycloheximide treatment will effectivelyblock recovery through its intrinsic action on protein synthesis.The data in Chart 4 indicate that 10~8 M actinomycin D [which is
reported to specifically block pre-rRNA synthesis (26)] had no
effect on the recovery from the metabolic block or on thealteration in the kinetics of recovery produced by estradici. Whenthe concentration of actinomycin D was increased, the rates ofthymidine and uridine incorporation were decreased in a dose-
dependent manner, suggesting a direct effect of this inhibitor onthe recovery from isoleucine starvation.
To ascertain whether an RNA polymerase ll-transcribed prod
uct is involved in the estrogenic stimulation of thymidine anduridine incorporation, the same experiments were performedwith a-amanitin at a concentration (10~7) known to produce a
50% inhibition in the activity of this enzyme. The data in Chart 5indicate that inhibition of RNA polymerase II with this agentreduced the rate of thymidine incorporation in control cultures toessentially that of the starting level at the time of reversal. Forthe most part, a-amanitin also blocked the stimulation in thymi
dine incorporation caused by estradiol, reducing this to the levelof the uninhibited control values at 16 h and slightly greater thanthe uninhibited control values at 24 h. The rates of uridineincorporation in both the control and estradiol-treated cultures
at both times studied were inhibited 50% by o-amanitin, which
was the expected result.All of the above studies indicated that estradici might act to
stimulate DMA synthesis by inducing (or enhancing) the synthesisof specific mRNAs. A critical feature of this phenomenon is theobservation of Jakesz ef al. (15) that estrogen receptor activityis reduced to 10% of control levels by 24 h of isoleucine starvation and only begins to reappear 14 h after reversal with completemedium. The reappearance of the estrogen receptor coincides(although somewhat later) with the estradiol-induced changes in
uridine incorporation. We therefore sought to determine if theseearly increases in uridine incorporation were merely the result ofa generalized increase in the expression of all mRNAs broughtabout by the combined recovery of the cells by isoleucine replenishment and the binding of estradiol present in the mediumto the rising levels of receptor.
The first mRNA we chose to examine was that for DHFR.Cowen ef al. (16) have demonstrated that estradiol causes a 1.5-to 2-fold increase in the levels of DHFR enzymatic activity incycling, methotrexate-resistant MCF-7 cells. Furthermore, Col
lins ef al. (29) have reported that synchrony of mouse fibroblastsachieved by isoleucine starvation results in delayed processingof DHFR heterogeneous nuclear RNA with no change in the totalamount of cellular DHFR mRNA. Since DHFR is a key step innucleotide biosynthesis, we reasoned that estradiol might preferentially stimulate an increase in the level of its mRNA followingreversal of growth arrest. The data presented in Fig. 1 indicate:(a) like mouse fibroblasts, there is no change in the total cellularamounts of DHFR mRNA in cycling or 24-h isoleucine-starvedcells; (b) estradiol induces a 2-fold increase in the amount of
DHFR mRNA within 24 h when given to cycling cells; and (c)estradiol does not increase DHFR mRNA levels within 24 h afterreversal of growth arrest with complete (+ isoleucine) medium.A similar experiment was performed with clone pS2. This cloneis a complementary DNA of a 600 nucleotide mRNA that isinduced by estradiol following pretreatment of MCF-7 cells with
nafoxidine. The data we obtained were identical to those obtained with DHFR; estrogen caused a 2-fold stimulation of mRNA
levels in cycling cells, and no change in mRNA levels wasobserved in the 24-h period following reversal of growth arrest
(data not shown). We did, however, find that TK mRNA wasincreased by 24 h, but not 12 h, of estrogen treatment (Fig. 2).The stimulation observed at 24 h correlates with the maximalrate of thymidine incorporation, but the absence of any increaseat 12 h suggests that changes in TK, like DHFR and pS2 mRNA,are not primary events in the growth response to estradiol.
These data indicate that the stimulation of thymidine anduridine incorporation caused by estradiol in these cells is uncoupled from hormone-induced increases in the expression of several estrogen-regulated genes. The results observed with DHFR
and TK are particularly striking, as we expected these mRNAsto increase in a manner parallel to the increases seen in uridineincorporation in estradiol-treated cells. The data suggest that
estradiol enhances (or induces) the expression of a specific setof genes that are intimately involved in DNA synthesis. It ispossible that the hormone merely increases the expression ofthe same genes that are involved in cell replication in randomlycycling, nonestrogenized MCF-7 cells. However, we believe the
data obtained with DHFR and TK mRNAs suggest that a differentsubset of genes might play a role in estrogen-induced DNA
synthesis. With this working hypothesis, we are presently preparing cDNA libraries from poly(A) RNA of MCF-7 cells to identifygenes and gene products directly involved in estrogen-stimulated
DNA synthesis.
ACKNOWLEDGMENTS
The authors wish to thank Dr. Anna TäteRiegel for her critical review of thismanuscript.
REFERENCES
1. Aitken, S. C., and üppman,M. E. Hormonal regulation of net DNA synthesisin MCF-7 human breast cancer cells in tissue culture. Cancer Res., 42:1727-
1735, 1982.2. Chalbos, D., Vignon, F., and Rochefort, H. Effects of estrogen on cell growth
and protein synthesis in the T47D human breast cancer cell line. In: G. Sato,A. Pardee, and D. Sirbasku (eds.), Growth of Cells in Hormonally DefinedMedia, pp. 845-849. Cold Spring Harbor, NY: CSH Press, 1982.
3. Katzenellenbogen, B. S., Norman, M. J., Eckert, R. L., Peltz, S. W., andMangel, W. F. Bioactivities, estrogen receptor interactions, and plasminogenactivator-inducing activities of tamoxifen and hydroxytamoxifen isomers inMCF-7 human breast cancer cells. Cancer Res., 44:112-119,1984.
4. üppman,M. E., Bolán,G., and Huff, K. The effects of estrogen and antiestro-gens in hormone-responsive human breast cancer in long-term tissue culture.Cancer Res., 36: 4602-4609,1976.
5. Natoli, C., Sica, G., Natoli, V., Serra A., and lacobellÃ,S. Two new estrogen-supersensitive variants of the MCF-7 human breast cancer cell line. BreastCancer Res. Treat., 3: 23-32,1983.
6. Page, M. J., Field, J. K., Everett, N. P., and Green, C. D. Serum regulation ofthe estrogen responsiveness of the human breast cancer cell line MCF-7.Cancer Res., 43: 1244-1250,1983.
7. Weichselbaum, R. R., Hellman, S., Piro, A. J., Nove, J. J., and Little, J. B.Proliferation kinetics of a human breast cancer cell line in vitro followingtreatment with 17tf-estradiol and 1-/a-o-arabinofuranosylcytosine. Cancer Res..38:2339-2342,1978.
8. Horwitz, K. B., and McGuire, W. L. Estrogen control of progesterone receptorin human breast cancer. J. Biol. Chem., 253: 2223-2228,1978.
9. Mairesse, N., Dcvleeschouwer, N., Ledere. G., and Galand. P. Estrogen-induced protein in the human breast cancer cell line MCF-7. Biochem. Biophys.Res. Commun., 97:1251-1257,1980.
10. Butler, W. B., Kirkland, W. L., and Jorgensen, T. L. Induction of plasminogenactivator by estrogen in a human breast cancer cell line (MCF-7). Biochem.Biophys. Res. Commun.. 90: 1328-1334,1979.
11. Huff, K., and LJppman, M. E. Hormonal control of plasminogen activatorsecretion in ZR-75-1 human breast cancer cells in culture. Endocrinology, 714:1702-1710,1984.
12. Westtey, B., and Rochefort, H. A secreted glycoprotein induced by estrogenin human breast cancer cell lines. Cell, 20: 353-362,1980.
13. Adams, D. J., Edwards, D. P., and McGuire, W. L. Estrogen regulation ofspecific messenger RNA's in human breast cancer cells. Biochem. Biophys.
Res. Commun., 97; 1354-1361,1980.14. Masiakowski, P., Breathnach, R., Btoch, J., Gannon, F., and Chambón, P.
Cloning of cDNA sequences of hormone-regulated genes from the MCF-7human breast cancer cell line. Nucleic Acids Res., 70: 7895-7903, 1982.
15. Jakesz, R., Smith, C. A., Aitken, S., Huff, K., Schuette, W., Shackney, S., andLJppman, M. Influence of cell proliferation and cell cycle phase on expressionof estrogen receptor in MCF-7 breast cancer cells. Cancer Res., 44: 619-625,
1984.16. Cowan, K. H., Goldsmith, M. E., Levine, R. M., Aitken, S. C., Douglass, E.,
Clendenin, E., Nienhuis, A. E., and LJppman, M. E. Dihydrofolate reducÃasegene amplification and possible rearrangement in estrogen-responsive methotrexate-resistant human breast cancer cells. J. Biol. Chem., 257: 15079-15086,1982.
17. Bronzen, D. A., Monaco, M. E., Pinkus, L., Aitken, S., and üppman,M. E.Purification and properties of estrogen-responsive thymidine kinase from human breast cancer. Cancer Res., 41: 604-610,1981.
18. Soule, H. D., Vasquez. J., Long, A., Albert, S., and Brennan, M. A human linefrom a pleural effusion derived from a breast cancer. J. Nati. Cancer Inst., 57:1409-1415,1973.
19. Tobey, R. A., Crissman, H. A., and Kremer, P. M. A method for comparingeffects of different synchronizing protocols on mammalian cell cycle traverse.J. Cell Biol., 54: 638-645, 1972.
20. Lowry, 0. H., Rosebrough, N. J., Fair, A. L., and Randall, R. J. Proteinmeasurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275,
1951.21. Brunk, C. K., Jones, K. C., and James, T. W. Assay for nanogram quantities
of DNA in cellular homogenates. Anal. Chem., 92: 497-500,1979.
22. Bresser, J., Doering, J., and Gillespie, D. Quick blot: selective mRNA or DNA
immobilizationfrom whole cells. DMA(NY),2: 243-254,1983.23. Chen, M., Shimado, T., Moulton, A. D., Harrison, M., and Nienhuis, A. W.
Intronless human dihydrofolate reducÃasegenes are derived from processedRNA molecules. Proc. Nati. Acad.Sci. USA, 79. 7435-7439,1982.
24. Bradshaw, H. D. Molecular cloning and cell cycle-specific regulation of afunctionalhuman thymidinekinasegene. Proc. Nati. Acad. Sci. USA,80:5588-
5591,1983.25. Ernst, S. G., and Oleinik, N. L. Actinomycin D in tetrahymena. Exp. Cell Res.,
110: 363-373, 1977.26. Perry, R. P., and Keltey, D. E. Inhibition of RNA synthesis by actinomycin D:
characteristic dose-response of different RNA species. J. Cell. Physiol., 76:127-140,1970.
27. Lindell, T. J., Weinberg, F., Morris, P. W., Roeder, R. G., and Rutter, W. J.Specific inhibitionof nuclear RNA polymerase II by a-amanitin.Science(Wash.DC), 770:447-450,1970.
28. Weinmann,R., and Roeder, R. G. Roleof DNA-dependentRNA polymeraseIIIin the transcription of the tRNA and 5S RNA genes. Proc. Nati. Acad. Sci.USA, 71: 1790-1794,1974.
29. Collins, M. L., Wu, J. R., Santiago, C. L., Hendrickson, S. L., and Johnson, L.F. Delayedprocessingof dihydrofolate reducÃaseheterogeneousnuclearRNAin amino acid-starvedmouse fibroblasls. Mol. Cell. Biol., 3:1792-1802,1983.
A. B.+ile
a eb ec •
d
-ile +E
b
c
d
C.
-E
Time After Reversal (hr)
O24681012162024
+EFig. 1. Dot blot analysis of DHFR mRNA. In A, cells were grown to a density of
8 x ICC/well, the medium was changed to IMEM + isoleucine (+/te) or IMEM -isoleucine (-/te), and RNA was extracted 24 h later for analysis by Nal immobilization as described in "Materials and Methods." Dots were cut out and counted.The equivalent amounts of cellular extract applied to each dot are 1 x 10s(a),2.5x 104(6), 6 x 103(c), and 1.6 x 103(d) cells, cpm were: - isoteucine(Lanec, 205;Lane d, 52); + isoleucine (Lane c, 213; Lane d, 51). In B, cells were grown to adensity of 8 x lO'/well, and the medium was changed to IMEM + isoleucine±5x 10~* M estradici (E). After 24 h, the cells were harvested, and the RNA was
processedby Nal immobilization.The sameamountsof cellularextract were appliedas above, cpm were: - estradici (Lane c, 128; Lane d, 51); + estradici (Lane c,252; Lane d, 140). In C, cells were grown to a density of 8 x I05/well, and themedium was changed to IMEM - isoleucine for 24 h. At time zero, the mediumwas replacedwith IMEM + isoleucine±5 x 10"' Mestradici. RNA was extractedat each time point, and the equivalent of 6 x 103 cells was immobilized ontonitrocellulose by the Nal procedure, cpm were: - estradici [205 (time zero), 268(12 h), 220 (24 h)]; + estradici [241 (time zero), 271 (12 h), 273 (24 h)].
Time After Reversal (hr)
*
Fig.2. Effect of estradici (E) on the expression of TK mRNA. MCF-7 cells werearrested for 24 h in IMEM - isoleucineand reversed in the presenceof 5 x 10"*
M estradici. Cells were harvested at 12 and 24 h, and total RNA was prepared.Seven /ig (a), 3.5 «ig(b), and 1.8 *ig (c) of each RNA preparation were bound tonitrocellulose and hybridized to purified insert DNA from a cloned exon of thehumanTK gene that was radiolabeledby nick translation.
