Download Negative Controls of Cell Proliferation: Human

[CANCER RESEARCH 49, 3474-3481. July 1, 1989]
Negative Controls of Cell Proliferation: Human Prostate Cancer Cells and
Androgens1
C. Sonnenschein,2 N. Olea, M. E. Fasanen, and A. M. Soto
Department of Anatomy anil Cellular Biology, Tufts University Health Science Centers, Boston, Massachusetts 02111
ABSTRACT
LNCaP cells represent a useful tool to explore the mechanism of sex
hormone action on cell proliferation in an "in culture-in animal" model.
Results indicated that: (a) these cells were inhibited from proliferating
for extended periods (up to 30 days) when placed in charcoal-dextranstripped sera; they remained, however, viable because they proliferated
when sex hormones were added to this medium; (b) the inhibitory effect
of sera was reversed by the addition of Sa-dihydrotestosterone at 3 x
I0~"' M, 17/3-estradiol at 3 x 10~" M and higher concentrations, and
progesterone at 3 x Kl'" M and higher concentrations; (c) while the dose
response to androgens was biphasic (i.e., 5a-dihydrotestosterone at con
centrations higher than 3 x 10~'°M resulted in progressively lower cell
yields), estrogens and progestagens exhibited a monophasic pattern; (d)
these cells were exceedingly sensitive to the nutritional environment in
which they grew; (<•)
while these cells have androgen receptors (68 fmol/
mg protein; A,, = 2 x 10 ' M), estrogen and progestagen receptors could
not be detected by biochemical and ¡mmunocytochemicaltechniques; (/)
tumors grew at the site of inoculation in castrated nude mice carrying
170-estradiol and progesterone pellets and in intact male nude mice
implanted with placebo pellets, while tumors did not grow in castrated
nude mice implanted with a 5a-dihydrotestosterone pellet. Taken together
the data collected are compatible with the following conclusions: (a) the
proliferatiti- response in LNCaP cells seems not to be directly mediated
by their intracellular androgen receptors; (b) plasma-borne trypsin-sensitive inhibitors of the proliferation of these cells (androcolyone I) appear
to play a significant role in the proliferative event; (c) natural and
synthetic androgens, estrogens, and progestagens cancelled the inhibition
by duiri-oal-di-Mrail-stripped human sera; (¡I)only androgens were able
to trigger an inhibition of cell proliferation (shutoff effect) at concentra
tions higher than those that affected maximal cell yields (direct negative
hypothesis); and (<•)
a faulty shutoff response is probably a crucial event
for the tumorigenesis of these human prostate cells.
INTRODUCTION
Our knowledge of the mechanisms whereby cells proliferate
is sketchy and full of controversy. Several hypotheses have been
proposed and they vary significantly depending on the premises
on which they are based. For example, while hypotheses on
positively regulated mechanisms rely on the notion that cells
are waiting for a signal to trigger their entering the cell cycle
(1, 2), hypotheses on negatively controlled mechanisms are
based on the premise that cells are constitutively poised to
proliferate and will do so when all inhibitory signals are can
celled or no longer recognized (3). The most frequently explored
hypotheses have been those dealing with the positive options
(direct and indirect) (1, 2, 4-6). Alternatively, using a variety
of vertebrate cell types, we have proposed and presented evi
dence compatible with the notion that only negative controls
are prevalent on the regulation of cell proliferation (3). Still
others have proposed that a combination of growth factors and
growth inhibitors is responsible for an integrated regulation of
Received 9/13/88: revised 3/15/89; accepted 3/28/89.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
' This work was supported in part by Grants USPHS NIH CAI3410 and NSF
DCB8711746.
2To whom requests for reprints should be addressed.
the proliferation of cells in metazoans (7-10).
Sex hormone-sensitive mammalian cell lines represent useful
experimental models to study the control of cell proliferation
in animals because sex steroids can be easily manipulated to
regulate the proliferation rate of these cells. In addition, sex
hormone-sensitive tumors represent a sizable proportion of
tumors that increase the morbidity and mortality of human
populations (11). After testing several putative androgen-sensitive models for cell proliferation (12),' we focused on the
human prostate cancer cell line LNCaP described by Horoszewicz et al. (13). LNCaP cells provide a useful human prostate
"in culture-in animai" model to assess proximate and ultimate
elements of causality in this mechanism.
MATERIALS AND METHODS
Cell Line. LNCaP-FGC cells were graciously made available to us
by Dr. Julius Horoszewicz from the State University of New York, at
Buffalo. They were established from a metastatic supraclavicular lymph
node removed from a 50-year-old patient with a prostate adenocarcinoma (13). When we obtained them these cells were on their 60th
passage after their establishment.
Culture Conditions. Cells were maintained routinely in DIME4 pur
chased from GIBCO, Grand Island, NY, and supplemented with 5%
FBS from Sterile Systems (Lot 1115715, Logan, UT). Cells were
routinely grown in 75- and 250-cm2 plastic flasks (Corning Plastics,
Corning, NY). To test their proliferative rate and proliferation yields
cells were seeded in 12-well plates (Costar, Cambridge, MA) in 5%
FBS-supplemented DME for 48-72 h so that cells would attach to the
plastic surface. Then, seeding media were quickly changed to the
experimental media; steroid hormones were added according to the
experimental design (see below). Because LNCaP cells attach loosely
to the plastic surface every effort was taken not to disrupt them; for
quantitation purposes, the cell-lysing solution was added to the exper
imental media in each well so that nuclei from attached and floating
cells would be counted. For cell yield experiments, cultures were kept
untouched for 7-10 days after the experimental media were added; for
proliferation rate experiments, cells were harvested every 1-2 days and
experiments were stopped after the slopes of the curve reached a
plateau.
Steroids and Other Substances Tested. Estrone, 17/J-estradiol, and
estriol were from Calbiochem, Richmond, CA. DHT, 5/3-dihydrotestosterone, testosterone, 19-nortestosterone, progesterone, pregnenolone, androstenedione,
5n-androstane-3n,17/i-diol,
5«-androstane,
3/3,17/3-diol,hydrocortisone, ethynyl estradici, and DES were obtained
from SUTalo ids, Keene, NH. Moxestrol and R1881 were graciously
provided to us by Roussell-UCLAF, Romainville, France: R5020 was
purchased from Du Pont-New England Nuclear, Boston, MA. Org4333
(11/3-chloromethyl estradiol) was made available to us by Organon,
Oss, The Netherlands. Mibolerone (7a,17a-dimethyl-l 9-nortestoster
one) was a gift by Dr. A. Traish, Boston University School of Medicine.
Epidermal growth factor was purchased from Collaborative Research,
Lexington, MA (Lot 88-1341) and graciously supplied by Creative
Biomolecules, Inc., Holliston, MA (Lot 055-A); insulin (Lot 615-0753C. Sonnenschein and A. M. Soto, unpublished data.
4 The abbreviations used are: DME. Dulbecco's modification of Eagle's me
dium; FBS, fetal bovine serum; DHT. dihydrotestosterone; DES. diethylstilbestrol; IET. 100 ng/ml insulin. 100 ng/ml epidermal growth factor, and 2 ng/ml
transferrin; CD. charcoal-dextran-stripped;
RBA, relative binding affinity;
CDHuS. charcoal-dextran-stripped human serum.
3474
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSITIVE
from 5 x 10 '°to 5 x 10~6M. The assay was carried out as described
256) was from Eli Lilly Co., Indianapolis, IN; and human transform)
(Lot 100F-9002) was from Sigma. This combination, called IET, was
added to DME when cells were grown in "serumless" media.
Experimental Design. Because LNCaP cells are of human origin, we
tested their proliferation properties using homologous sera. Blood was
drawn from male and female healthy volunteers. Sera were obtained
from clotted blood, centrifuged to eliminate cells, heat inactivated for
30 min at 56°C,and stored frozen at —
20°C.When needed, these sera
were thawed and charcoal-dextran stripped according to a protocol
described elsewhere (14). CD sera were used either immediately after
stripping or were kept frozen at —20°C.
Proliferation yields of LNCaP cells in heat-inactivated human male
and female sera were measured to establish whether or not the sex
origin of the sera mattered. Next, similar experiments were run using
heat-inactivated alone and heat-inactivated plus charcoal-dextranstripped sera. In these experiments, the different sera were tested at
various concentrations between 0.5 and 50% to determine whether the
addition of various fixed sex hormone concentrations would elicit
comparable responses. Toward the end of their exponentially growing
period, cells were lysed with a detergent and nuclei were counted in
duplicate in a Coulter Counter Zf¡Model Apparatus. These data were
statistically analyzed and represented graphically using the LOTUS
123 and the Sigma plot packages in an IBM XT. A similar procedure
was followed for proliferation rate curves (3, 14).
Tumorigenesis in Nude Mice. Groups of 5 nude mice (nu/nu CD1
from Charles River Research Laboratories, Wilmington, MA) were
castrated and immediately implanted with pellets containing 15 mg of
DHT, 0.72 mg of 170-estradiol or 20 mg of progesterone. A group of
5 intact males and females were implanted with a placebo pellet and a
group of 5 castrated male mice were implanted with placebo pellets.
Pellets were purchased from Innovative Research of America, Toledo,
OH. One day after these procedures took place each nude mouse was
inoculated s.c. with approximately IO7cells suspended in 0.25 ml DME
in the interscapular area. Animals were observed twice weekly and
tumor development was recorded. Tumors were excised to measure
levels and localization of estrogen receptor. This experiment ended 75
days after cell inoculation. Sex hormone receptors were measured by
biochemical procedures described below; estrogen and progesterone
receptors were characterized also by a commercially available kit (Ab
bott Laboratories, Chicago, IL).
Binding Studies. Labeled steroids |ïI (testosterone (specific activity,
3.1 TBq/mmol), [3H]l7ß-estradio\ (specific activity, 3.8 TBq/mmol)
and [3H]R5020 (specific activity, 3.2 TBq/mmol) were purchased from
Du Pont-New England Nuclear. Intracellular sex steroid-binding pro
teins were investigated in whole cell extracts as well as in intact cells.
Cytosolic and nuclear extracts were obtained after centrifugation
(105,000 x g, 45 min) of a sonicated cellular suspension in 10 mM
Tris-500 mivi KC1-0.5 mM EDTA buffer, pH 7.4. Aliquots of this
extract were incubated for 18 h at 4°Cwith increasing concentrations
(2 x 10~'-2 x 10~8M) of labeled steroids, with and without a 200-fold
excess of the corresponding unlabeled steroid. Bound and free fractions
were separated by dextran-coated charcoal adsorption (14). Intact cells
grown in monolayer culture were assayed for specific steroid binding
(15). Briefly, the cells were incubated in 12-well plates during 50 min
at 37°Cin DME containing the labeled steroid with and without excess
of cold competitor. The range of concentrations used was the same as
that in the cytosolic-nuclear extract assay. Previously, it was determined
that specific uptake of labeled steroids in these cells reaches plateau
during the first 25 min of incubation.5 Intracellular radioactivity was
extracted with ethanol after ice-cold phosphate-buffered saline washing.
In both types of experiments, Scatchard analysis of saturation data was
used to quantify maximal binding capacity (/f,,,.,J and affinity parame
ters (dissociation constant, Ka) by the SCAFIT program (16).
Cytosol-nuclear extracts were used to study competition for androgen
receptors. Aliquots were incubated with 6 x 10~9 M [3H]testosterone
and variable concentrations of cold competitors (testosterone, DHT,
5«-androstane-3a, 17/3-diol, 5a-androstane-3/3,17/3-diol, mibolerone
R1881, 17/3-estradiol, estrone, estriol, ethynyl estradici, DES,
Org4333, moxestrol, progesterone, pregnenolone, and R5020) ranging
5 N. Olea, unpublished results.
TUMOR CELLS
above for the receptor assay. The RBA was calculated from the equation
RBA =
Testosterone (I50)
x 100
Test competitor (I50)
i.e., the ratio between the concentration of unlabeled testosterone and
that of the competitor that inhibits 50% (I50) of the [3H]testosterone
bound to androgen receptors.
Partial Purification and Trypsin Treatment of the Plasma-borne In
hibitors. Human plasma was treated with 80 mM BaCI2, 15 mM sodium
citrate, and 1 mM benzamidine for l h at 4°C.After centrifugation at
2000 rpm for 20 min, the supernatant depleted of coagulation factors
was dialyzed against 25 mM Tris-HCl, pH 7.4. Plasma protease inhib
itors (a,-antitrypsin, a2-macroglobulin, etc.) were removed from 10 ml
of BaCl2-treated plasma by chromatography through a 25-ml Cibacronblue agarose column (Pierce Chemical Co., Rockford, IL); the retained
fraction was eluted with 2 M NaCl-25 mM Tris, pH 7.4. Aliquots of
this fraction were treated as follows: (a) chromatography through a 5ml benzamidine-Sepharose column (Pharmacia-LKB, Piscataway, NJ)
as a control; and (¿>)
incubation with 0.5 mg/ml of trypsin for l h at
37°Cfollowed by chromatography in benzamidine-Sepharose to remove
trypsin from the androcolyone preparation. Both benzamidine-Sepha
rose eluates were concentrated by ultrafiltration, dialyzed against 100
mM NaCl-25 mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonic
acid,
pH 7.4, and rendered sterile by filtration before adding to IET-supplemented DME. Proliferation yield experiments to test the effect of these
plasma fraction-supplemented media were done as described above and
elsewhere (14).
RESULTS
Proliferative Properties of LNCaP Cells
Effects of Heat-inactivated Human Male and Female Sera.
Human female serum was more inhibitory than male serum
when heat inactivated (Fig. I A). DHT increased cell yields in
10% female serum; in contrast, DHT added to male serumsupplemented medium did not significantly increase cell yields.
However, 3 x 10~7 M DHT consistently inhibited cell prolifer
ation (Fig. IA); this was also seen when 10% female serum was
used.
a 2
UJ
IO
1
1CT12
IO"10
10"e
DHT CONCENTRATION
1CT
(M)
10-12
ESTRADIOL
,0-10
,0-e
CONCENTRATION
,0-6
(M)
Fig. 1. Cell proliferation yields obtained when established human prostate
tumor cells (LNCaP) were grown in media supplemented with 10% human male
(O) and female (•)sera that were heat-inactivated (A and C) or heat-inactivated
and charcoal-dextran-stripped sera (B and D). These sera were supplemented with
either 3 x KT12-3 x IO'" M DHT (A and A) or 3 x l(T"-3 x irr7 M 17/3estradiol (C and D). Values are mean ±SD (bars) of 4 experimental measure
ments; when SD was less than or equal in magnitude to the size of the plotted
data point bars were not plotted. Cells were grown undisturbed in experimental
media for 7 days.
3475
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSITIVE
TUMOR CELLS
Effect of DHT on Cell Proliferation Yields. Charcoal-dextranstripped male and female human sera showed a comparable
inhibitory effect on the proliferation yield of LNCaP cells (Fig.
\B). Addition of DHT to either female or male CD serumsupplemented media generated a biphasic proliferative response
(Fig. \B); the peak was reached at 3 x 10~'" M. Concentrations
above 3 x 10"'" M DHT resulted in progressively lower cell
yields. As a result of this type of experiment, heretofore, we
used sera from either sex when heat-inactivated CD-stripped
human sera was required.
Effect of Other Sex Hormones on Cell Proliferation. Fig. 1, C
and D, shows the effect of 17/3-estradiol at concentrations
between 3 x 10~" and 3 x 10~7 M when supplemented to (a)
intact heat-inactivated and (b) heat-inactivated and charcoaldextran-stripped 10% human male and female sera. Cell yields
were approximately 2-fold higher in plates treated with 3 x
NT* M 17/3-estradiol than with 3 x 1(T10M DHT.
Testosterone showed a proliferative pattern similar to that
obtained with DHT (not shown). 3«-Androstanediol was 103fold and 3/3-androstanediol was 10-fold less potent than DHT
(Fig. 2A). However, 3/3-androstanediol increased the cell yield
to values approximately 2-fold higher than those obtained with
DHT; this effect was comparable to that seen with 17/3-estradiol. In addition, 30-androstanediol concentrations above those
required for maximal cell yield resulted in a shutoff effect
similar to that observed with DHT (Fig. 2A). 19-Nortestosterone was 10-fold more potent than DHT (not shown). On the
other hand, 5/3-DHT and androstenedione were 100-fold less
potent than DHT (not shown). Estrone and estriol did not
affect the proliferation yields of LNCaP cells. 17a-Estradiol
was 100 times less effective than 17/3-estradiol (Fig. 2B). Pro
gesterone (Fig. 2C) and pregnenolone (not shown) elicited a
monophasic proliferative pattern. Cholesterol and hydrocortisone did not induce LNCaP cell proliferation over control
values (not shown).
Effect of Synthetic Androgens, Estrogens, and Progestagens
on Cell Proliferation. The synthetic androgens R1881 (Fig. 2A)
and mibolerone (not shown) were 100 times more potent than
DHT. Both androgens triggered the shutoff response as DHT
did.
Among synthetic estrogens neither DES nor moxestrol was
active; Org4333 increased the proliferation yield of these cells
at concentrations of 3 x 10~9 M and higher (Fig. 2B). The
proliferative pattern of Org4333 was similar to that shown by
17/3-estradiol, i.e., no shutoff effect. Ethynyl estradici behaved
similarly to 17/3-estradiol (Fig. 2B).
Of synthetic progestagens, R5020 was approximately 2 or
ders of magnitude less potent than progesterone (Fig. 2C);
higher concentrations of either progesterone or R5020 did not
result in lower cell yields.
Proliferation Rate of LNCaP Cells. The population-doubling
time of LNCaP cells was 40 h when 3 x 10-"' M DHT was
added to 10% charcoal-dextran-stripped
human sera-supple
mented medium (Fig. 3). The inhibitory effect of CD human
sera was not instantaneous; at the end of 3-4 days the popula
tion increase was halted and remained stable for up to 30 days
without media changes. That these cells were then still viable
is suggested by increased cell proliferation rates when 3 x IO'"'
M DHT was added (not shown). The proliferation rate of
LNCaP cells growing in the presence of DHT was significantly
faster than that of the control (minus DHT). 170-Estradiol
showed a slightly shorter population-doubling time. It appears
as if 17/3-estradiol allowed for an additional population dou
bling over what DHT influenced (Fig. 3).
-10
1CT1U
10~8
STEROID CONCENTRATION
10r«
(M)
Fig. 2. Dose-response cunes in medium supplemented with 10% CDHuS and
with different hormones. A, DHT (O). 5«-androstane-3n,17^-diol (•),5ir-androstane-.V.17rf-diol (A). R1881 (A). B, DHT (O), 17«-estradiol(A). 17f(-estradiol
(•),cthynyl estradiol (A). Org4333 (D). C. DHT (O). progesterone (•).R5020
(A). Value* are mean ±SD (bars) of 4 experimental measurements. Cells were
kept in experimental media without change for 7 days. This is 1 of 20 represent
ative experiments.
Serum Concentration Effects on Proliferation Yields. Absolute
yields varied according to the serum and sex hormone concen
trations used (Fig. 4). The proliferative effect of DHT appeared
to be influenced by serum concentration in the media; lower
DHT concentrations (3 x 10~" M) were necessary to cancel the
inhibitory effect of low CD human serum-supplemented media
(1-5%). Higher DHT concentrations (3 x 10"'°M) were needed
to cancel the inhibitory effect of high CD human serum-supple
mented media (10-50%). In addition, lower serum concentra
tions generated lower yields when compared with higher serum
concentrations in experiments seeded and harvested simulta
neously. A comparable pattern was obtained with 17/3-estradiol
(Fig. 4B).
Effect of Putative Growth Factors and Growth Inhibitors.
Testing of proliferation yields under putative serumless condi
tions revealed information worth reporting. As described above
(see "Experimental Design"), cells were kept for 48 h in 5%
FBS so that a sufficient number of cells would become attached
3476
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSITIVE
TUMOR CELLS
the proliferative potency of some of the compounds tested.
[3H]17/3-Estradiol binding was displaced by androgens and pro
gestagens. DES and moxestrol, which bind to estrogen receptor
present in estrogen-sensitive cells, did not compete with [3H]17/3-estradiol binding. These observations, as well as the ab
sence of immunoreaction with monoclonal antibodies against
human estrogen receptor (not shown), indicate that no estrogen
receptor is present in these cells; this confirms data of others
(17). R5020 binding was also displaced by androgens and
estrogens with RBAs comparable to those observed when com
petition between ['Hjtestosterone and progestagens was meas
ured. These observations indicate that steroid-binding activity
seems to represent androgen receptor.
While a good correlation between relative binding affinities
and proliferative potency was established for many compounds
tested, R1881 and mibolerone were more potent to induce cell
proliferation than predicted by their RBA. In addition, both
progesterone and pregnenolone were more potent inducers of
cell proliferation than predicted by their RBAs; pregnenolone
displayed the most extreme lack of correlation between its RBA
(<0.01%) and its proliferative potency (1%) when compared
with DHT values.
LU
ce
m
m
5
D
Z
LU
ü
10 5_
Tumorigenesis in Nude Mice
Tumors developed in castrated mice implanted with 17/3estradiol and progesterone and in intact male animals (Table
3). Tumors did not develop in castrated mice implanted with
placebo pellets, in mice implanted with DHT pellets, and in
female nude mice. Plasma hormone levels were not determined
in these mice.
4x10
3
5
time (days)
Fig. 3. Proliferation curve of the LNCaP cells. Similar inocula were grown in
10% CDHuS with 3 X IIT10 M DHT (•),with 3 X KT" M 17/3-estradiol (D), or
without hormone (O). Media were not changed for 7 days when cells were
harvested. Each point represents three wells. Values are mean ±SD (bars).
to the plastic surface. Cells seeded and kept for 48 h in media
supplemented with increasing concentrations of FBS (from 1
to 40%) showed a significantly different cell yield when chal
lenged to proliferate in IET-supplemented DME; a direct pos
itive correlation could be established between the serum con
centration in the seeding media and the cell yield achieved in
DME plus IET (Fig. 5). When cells were seeded in 1% FBS the
difference between cell yields of controls and sex hormonesupplemented cultures were less significant than those recorded
when cells were seeded in 5% or higher FBS concentrations.
Effect of Trypsin Treatment on Partially Purified Plasmaborne Inhibitors. Table 1 shows that the partially purified frac
tion inhibited the proliferation of LNCaP cells when compared
to IET medium. In addition, DHT added to the purified fraction
significantly increased cell yield, whereas IET was ineffectual.
Trypsinization of the plasma fraction resulted in a significant
cell yield increase in the absence of androgen; moreover, DHT
did not increase the cell yield when added to the trypsin-treated
purified fraction. Remarkably, adding 10% CDHuS to the
trypsin-treated fraction restored both inhibition and hormone
responsiveness.
3-2-jIti¡S**
T'
Ei
1-^^to
^Iss
^^Blr*i/^f|in/
f\
fey//vl fsvs^sssssN
_!
r1ifi*
o's«"(flm3Z
3-immtLUa
sssssS
NSSsNs
2-1-n-fr
_Ls\s\s\utl.
///tf/
Intracellular Sex Steroid-binding Proteins
LNCaP cells have been reported to contain androgen recep
tors (13). We confirmed this. The KA of the receptor-DHT
complex was 3 x 10~9 M; the concentration of binding sites was
68 fmol/mg of protein in the 100,000 x g extract (Fig. 6). The
whole cell assay indicated about 40,000 binding sites/cell.
Testosterone binding was displaced by other androgens, some
estrogens, and progestagens. Table 2 compares the RBA and
_
riJ| r1fEs^s
ÕA[
SÃŽ
t/
///
/
1%
5%
10%
20%
30%
40%
50%
Fig. 4. Serum concentration effects on proliferation when LNCaP cells were
grown in media supplemented with 1-50% CDHuS P, A and B); plus DHT (A:
m. 3 x 10~" M; P, 3 x 10"'°M; ®,3 x 10"' M) or plus 17i)-estradiol (B: •.3 x
10"' M; M, 3 x I0~8 M). Cells were harvested at the end of 7 days of culture
without media change. Values are mean ±SD (bars) of 4 experimental measure
ments.
3477
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSmVE
25201510Ltl
5
5-
o
C
o1%
ce.
£ 25-
5%
10%
20%
40%
B
Z
20-
105-
1%
5%
10%
20%
40%
Fig. 5. Serum concentration effects on LNCaP cell proliferation. Cells were
seeded in 1-40% FBS and 48 h later were changed to IET-supplemented DME
alone (O, A and A) or plus DHT (A: •,3 x 10~" M;^, 3 x \0~">M;M, 3 X 10~"
M) or plus 17/i-estradiol (B: •,3 X 10"* M; Ü,3 X 10"' M). Cells were harvested
after 7 days of culture without media change. Values are mean ±SD (bars) of 4
experimental measurements.
Table 1 Trypsin treatment of partially purified androcolyone I
Results are expressed as cell number/well (mean ±SD) after 5 days of exposure
to experimental media.
TreatmentAndrocolyone
frac
ratio151,050±
+DHT
TUMOR CELLS
that androgens directly stimulate cell proliferation; we will
expand below on why we consider this latter option unlikely.
On the other hand, androgen-mediated inhibition of LNCaP
cell proliferation appears to be a consequence of the androgenspecific induction of a gene product the function of which is
shutting off the entry of these cells in the cell cycle. These
conclusions are not significantly different from those drawn
more than a decade ago by Bruchovsky et al. (18). These
interpretations will be confirmed when the following evidence
will become available: (a) the purification of the plasma-borne
specific inhibitor of the proliferation of androgen-sensitive cells
(androcolyone I); (b) the characterization of specific receptors
for these inhibitors; and (c) the characterization and purification
of the intracellular androgen-induced proliferation inhibitor
(androcolyone II).
While the task of generating evidence to verify these inter
pretations is completed using the LNCaP model, it is worth
while attempting the reconstruction of probable regulatory
mechanisms of the proliferation of these human androgensensitive cells.
Tumor Nature of LNCaP Cells. LNCaP cells were established
from a supraclavicular lymph node metastasis of a malignant
prostatic adenocarcinoma in a patient who died within 2 years
of the apparent onset of the disease (13). Figs. 1 to 5 indicate,
however, that these cells can be effectively arrested by sex
steroid-depleted sera. Both male and female human sera were
effective and to the same extent as far as we could measure it.
This inhibition, however, was cancelled by physiological con
centrations of natural androgens (Figs. 1 and 2A). That this is
not just a fortuitous, idiosyncratic "in culture" event is indicated
by the lack of tumor formation by these cells in castrated nude
mice (Table 3). Comparable data have been generated when
another human prostate carcinoma was challenged to prolifer
ate in nude mice (19). These data are, therefore, compatible
with the notion that human and mouse sera contain an inhibitor
of the proliferation of human prostate cells. Bovine, horse, pig,
±5,184
18,835
tion"
Trypsin-treated an
145.540 ±10,440 153,720 ±5,399
0.92.35
drocolyone*
10% CDHuS-DHT98,93084.260 ±11,527+DHT/-DHT
198,1 20 ±25,8801.5
Trypsin-treated an- 101,290 ±10,603 224,610 ±10,503
2.22
drocolyone +
10% CDHuS
IET
143,020 ±6,747
148,430 ±31,954
1.04
°Androcolyone fraction was obtained by Cibacron blue and benzamidine-
Table 2 Comparison between RBAs to the androgen receptor and proliferation
potencies of diverse steroidal and nonsteroidal compounds on LNCaP cells
HormoneDHT5a-Androstane-3a,
proliferative
potency*1000.11010,00010,0001—
7/3-diol5«-Androstane-3/3,l
1
7/3-diolR1881Mibolerone1
Sepharose chromatography and added to IET to a final concentration of 3 mg/
ml.
* Androcolyone fraction was incubated with 0.5 mg/ml trypsin for l h at 37"C;
trypsin was removed by benzamidine-Sepharose chromatography; 3 mg/ml were
added to IET. Cells were harvested after 5 days of culture without changing
media.
7/3-EstradiolEstroneEstriolEthynyl
'1__1—10011
estradiciR2858Org4333DESProgesteronePregnenoloneR5020RBA°1001.214200150.90.1<0.011.
DISCUSSION
The human prostate tumor LNCaP cell line has proved to be
a reliable tool to explore competing hypotheses on the mecha
nism of androgen regulation of cell proliferation. Results sug
gest a complex interaction among the protagonists of such
regulation. Thus far, our data point to the participation of
distinctive pathways favoring the proliferation and the inhibi
tion of this proliferation occurring in a successive fashion. On
the one hand, data are compatible with the notion that prolif
eration occurs when some natural and synthetic sex hormones
cancel the inhibition generated by human serum depleted of sex
hormones by charcoal-dextran stripping. While we favor this
interpretation we cannot yet completely rule out the alternative
" Calculated as indicated in "Materials and Methods."
*The relative proliferative potency was the ratio (xlOO) between the concen
tration of testosterone and agonist necessary to elicit maximal cell yield for a 104
cell inoculum grown in CDHuS for 8 days.
c—, no proliferative potency at the range of concentrations tested.
Table 3 Tumorigenesis when iff1 LNCaP cells were inoculated in nude mice
Intact female + placebo
Intact male + placebo
Castrate male + placebo
Castrate male + DHT pellet
Castrate male + 17/3-estradiol pellet
Castrate male -I-progesterone pellet
0/5
4/5
0/5
0/5
3/5
3/5
3478
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSITIVE
dog, and chicken sera also inhibit the proliferation
LNCaP cells.6
of these
Proliferative and Inhibitory Effects of Androgens. Androgen
administration
generated a biphasic proliferative pattern
whereby higher concentrations than those required for maximal
cell yield resulted in a dose-dependent inhibitory' effect (Figs. 1
and 2A). This phenomenon resembles the time-dependent
shutoff effect of androgens on the proliferation of prostate
epithelium in castrated rats (18).
Proliferative Effect of Estrogens and Progestagens. Addition
of 17/3-estradiol, 17a-estradiol, ethynyl estradici, and Org4333
to 10% charcoal-dextran-stripped
human serum-supplemented
medium resulted in increased cell yields. Both estrogens and
progestagens showed a monophasic dose-response curve. Un
like androgens, estrogens and progestagens failed to trigger the
inhibitory shutoff response. Estrogens and progestagens also
induced proliferation of LNCaP cells inoculated into castrated
male nude mice (Table 3). Thus, good correlation between in
culture and in animal studies has been established in this model.
Correlation between Intracellular Receptor Binding and Pro
liferative Potency of Sex Steroids and Synthetic Agonists.
LNCaP cells contain an intracellular androgen-binding moiety
which binds DHT (Ka 3 x 10~9 M) and is present at a concen
tration of 68 fmol/mg protein (Fig. 6). These cells lack estrogen
receptors because (a) no immunoreactive estrogen receptor was
detected; (¿>)
the Kd for [3H]17/3-estradiol was lower than ex
pected for human estrogen receptor (20); and (c) [3H]17/3estradiol binding was not displaced by DES and moxestrol
while it was displaced by androgens and progestagens. That
pH]17|0-estradiol binding by LNCaP cytosol represents androgen receptor is suggested by the fact that DES does not displace
[3H]DHT from the rat ventral prostate androgen receptor (21),
while 17/3-estradiol is known to reduce ['HJDHT binding to
androgen receptors; this evidence was collected by sucrose
gradient analysis (22) and by measuring 17ß-estradioleffects
on the nuclear retention of [3H]DHT in prostate slices (23).
When comparing the proliferative potency of diverse hor
mones with their relative binding affinities a good correlation
was found for 3/3-androstanediol, DHT, testosterone, and 6 of
the 7 estrogens shown in Table 2. On the other hand, proges
terone, pregnenolone, mibolerone, and R1881 were more po
tent to induce cell proliferation than predicted by their relative
binding affinities for the androgen receptor. For example, the
proliferative potency of pregnenolone was 1% of the DHT
potency, while its RBA was less than 0.01% when compared to
DHT. In addition, 3a-androstanediol, estrone, and R5020 dis
played lower proliferative potency than predicted by their re
ceptor-binding affinities. Lack of correlation between affinity
for the androgen receptor and activity was reported for both
synthetic agonists and antagonists using the rat ventral prostate
model (21). These data are interpreted as evidence for (a)
metabolic conversion of the hormone into a more potent me
tabolite when hormonal action exceeds binding activity or (b)
fast metabolic clearance when hormonal activity is lower than
anticipated by the receptor binding data. The proliferative effect
of estrogens and progestagens could not be explained by meta
bolic conversion to androgens since (a) estrogens and proges
tagens displayed a monophasic proliferative response while all
androgens show a biphasic response and (b) addition of estro
gens and progestagens consistently resulted in a 2-fold higher
cell yield than that generated by the androgens tested. Because
progestagens showed a proliferative effect comparable in po
6Unpublished data.
TUMOR CELLS
LU
O
rr
LU
« 0.2CO
I
0.1 Ul
LU
ce
o
D
Å“
0
25
50
75
BOUND [3H] -TESTOSTERONE (fmol/mg protein)
Fig. 6. Scatchard plot of the saturation data obtained when LNCaP cell cytosol
was incubated during 18 h at 4°Cwith [3H]testosterone with and without cold
competitor. The maximal binding capacity and the value of the dissociation
constant were 68.73 fmol/mg of cytosol protein and 3.2 x 10"' M, respectively.
Correlation coefficient of point fit 0.977.
tency to DHT while displaying cell yields consistently higher
than those evoked by androgens, a critical evaluation on the
role of the androgen receptors is in order.
The direct positive hypothesis for androgen action on cell
proliferation proposes that hormone action is directly mediated
by steroid binding to the intracellular receptor. Hormone recep
tor complexes would interact with regulatory sequences of genes
under androgen control and elicit their expression. However,
there is no compelling evidence supporting this pathway for the
control of cell proliferation. In fact, if one assumes that cell
proliferation was directly mediated through receptor binding a
strong correlation between receptor binding affinity and prolif
erative potency should be apparent. Data on LNCaP cells do
not provide evidence for this mode of action because: (a) while
estrogens and progestagens induce cell proliferation, these cells
do not have estrogen or progesterone receptors; (b) if the action
of estrogen and progesterone were mediated through interaction
with the androgen receptor, a good correlation between binding
affinity and proliferative potency should have been apparent
(this is not the case at least for estrone and the progestagens
tested); and (c) if the proliferative action of estrogens and
progestagens were mediated through the androgen receptor, a
similar response pattern should have been obtained with all
these sex steroids. Instead, all androgens displayed both a
proliferative and an inhibitory response (shutoff) whereas estro
gens and progestagens only displayed the proliferative response
and consistently induced higher cell yields than androgens
(Table 2).
The data discussed above are incompatible with, but do not
rule out, the direct and indirect positive hypotheses. To the
contrary, these data are compatible with the direct and indirect
negative hypotheses; according to the indirect negative hypoth
esis androgens, estrogens, and progestagens induce cell prolif
eration by neutralizing a serum-borne specific inhibitor of androgen-sensitive cell proliferation (androcolyone I). On the
other hand, the inhibitory effect of androgens seems to be
mediated through the androgen receptor since (a) there is a
good correlation between the potency of these hormones to
induce the shutoff effect and their relative binding affinities to
the androgen receptor and (b) estrogen and progestagens are
unable to evoke the shutoff response.
Proliferative Response of LNCaP Cells to Human Sera. Our
previous work on estrogen-sensitive cell proliferation revealed
3479
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSITIVE
that (a) charcoal-dextran-stripped
serum inhibited the prolif
eration of estrogen-sensitive cells in a dose-dependent manner
and (b) estrogen sensitivity was not apparent in serumless
medium where cells proliferated maximally regardless of the
presence of estrogens (3, 14, 20). Similar results were expected
with LNCaP cells. The proliferation yields of LNCaP cells in
1-50% CD human serum were comparable while the yields
obtained with DHT or 17/3-estradiol increased with serum
concentration reaching a maximum at 20% CDHuS (Fig. 4).
The hormone concentration needed to obtain maximal cell
yields also increased as the serum concentration increased (Fig.
4). These results suggest that serum plays a prominent nutritive
role which overlaps either with the inhibitory role of the androcolyone I it carries (indirect negative hypothesis) or with the
stimulatory role of sex hormones (direct positive hypothesis).
Fig. 5 shows that the cell yield of LNCaP cells in serumless
media (DME plus IET) was also dependent on the fetal bovine
serum concentration in the seeding medium, i.e., prior to the
cells being exposed to serumless media. The more serum in the
seeding medium, the higher was the cell yield in both the control
and sex hormone-treated wells; however, the proliferative yield
generated by DHT and 17/3-estradiol-treated cells was consid
erably smaller in comparison to that of controls when cells were
seeded in 1% fetal bovine serum. These data are compatible
with the notion that LNCaP cells are exceedingly sensitive to
nutritional availability and that sex hormones do not compen
sate for lack of nutrients. Also, these data suggest that heatinactivated fetal bovine serum contains a surplus of active
androcolyone I that is cancelled by DHT and 17/3-estradiol.
These data by itself, however, do not allow us to discriminate
between sex hormones inducing cell proliferation directly or
indirectly (by cancelling the inhibitory effect of the serum-borne
androcolyone I).
Table 1 further substantiates the presence of a plasma/serumborne androcolyone I; trypsin treatment of a partially purified
androcolyone I preparation resulted in the loss of both inhibi
tory and hormone effects. That trypsinization releases a growth
factor is unlikely because adding 10% CDHuS to this prepara
tion restored inhibition and hormone responsiveness.
Role of Putative Growth Factors on the Proliferation Yield of
LNCaP Cells. The indirect positive hypothesis proposes that
sex steroids induce the synthesis of growth factors that in turn
trigger the entry of their target cell into the cell cycle (1, 2, 4,
19,24). We tested the effect of insulin, epidermal growth factor,
and transferrin when added to phenol red-free DME plus 1%
and 10% CDHuS. The results shown in Fig. 7 indicate that the
cell yield of LNCaP cells was similar regardless of whether or
not these growth factors were added to the serum-containing
media. DHT added to serumless, growth factor-supplemented
media resulted in a marginal increase of the cell yield; however,
the effect of DHT was proportionally higher as the CD serum
concentration in the media increased. The growth factors tested
appear to play no significant role in these sex steroid-specific
proliferative events.
Results described in this paper provide interesting clues on
the hormonal control of cell proliferation. They can be sum
marized as follows, (a) The proliferative response obtained with
androgens, estrogens, and progestagens challenges the notion
that this event is mediated through the androgen receptor
pathway and suggests an alternative hypothesis for hormone
action on cell proliferation, i.e., the involvement of plasmaborne inhibitors of androgen-sensitive cell proliferation (andro
colyone I), (b) The potent inhibitory effect of androgens on cell
proliferation seems to be mediated through the androgen recep-
TUMOR CELLS
3 -
2 -
o
i H
"^
ce
LU
m
C
D
111
U
B
3 -
2 -
1 -
10%
IET
1%
Fig. 7. Cell proliferation yields obtained when LNCaP cells were grown in
IET or 1 and 10% CDHuS-supplemented DME medium. A and B: D, control
wells (no hormone added); A: DHT supplemented at 3 x 10"'°M (•),3 x 10"'
M (^), and 3 x 10~s M (^), respectively. B: 17/3-estradiol supplemented at 3 x
10~' M(•),3 x IO'8 Mp), and 3 x 10~7M (M), respectively. Cells were harvested
after 7 days of culture in media kept undisturbed. Values are mean -I-SE (bars)
of 4 experimental measurements.
tor pathway (direct negative hypothesis); because the prolifer
ative response seems to be triggered by all three sex steroid
groups and the inhibitory response is only triggered by andro
gens, these results suggest that the ultimate proliferative control
in normal prostate cells is the shutoff mechanism triggered by
high androgen plasma levels, (c) The development of an androgen-dependent tumor is probably due to a defective shutoff
mechanism; this will allow proliferation to occur in the presence
of androgen concentrations otherwise sufficient to trigger the
shutoff mechanism in normal cells. In addition, these results
provide leads to gain a mechanistic understanding of the role
of sex steroid on the control of cell proliferation by searching
for: (a) a serum-borne inhibitor of androgen-sensitive cell pro
liferation (androcolyone I); (b) androcolyone I receptors on
androgen sensitive cells; (c) sex hormone-androcolyone I inter
actions; (d) the gene product(s) responsible for the androgentriggered shutoff effect (androcolyone II); and (e) target genes
for the action of androgens on the induction of androcolyone
II (colyogenes) (3).
ACKNOWLEDGMENTS
The authors would like to acknowledge the technical support of
RenéeSilvia and the secretarial help of Virginia Olson and Mary
Currier.
REFERENCES
1. Erikson, R. L. Regulation of cell proliferation by oncogene products and
growth factors. Somatic Cell Mol. Genet., 13: 459-461, 1987.
2. Heldin, C. H., Betsholtz, C., Claesson-Welsh, L., and Westermark, B. Sub
version of regulatory pathways in malignant transformation. Biochim. Biophys. Acta, 907: 219-244, 1987.
3. Soto, A. M., and Sonnenschein, C. Cell proliferation of estrogen sensitive
cells: the case for negative control. Endocr. Rev., 8: 44-52, 1987.
4. Dickson, R. B., Huff. K. K., Spencer, E. M., and Lippman, M. E. Induction
3480
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
HUMAN ANDROGEN-SENSITIVE
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
of epidermal growth factor-related polypeptides by 17(j-cstradiol in MCF-7
human breast cancer cells. Endocrinology, US: 128-142, 1985.
Matuo, Y., Nishi, N., and Wada, F. Growth factors in the prostate. Arch.
Androl., 19: 193-210, 1987.
Schuurmans, A. L., Bolt, J., and Muller, E. Androgens stimulate both growth
rate and EOF receptor activity of the human prostate tumor cell LNCaP.
Prostate, 12: 55-63, 1988.
Keski-Oja, J., and Moses, H. L. Growth inhibitory polypeptides in the
regulation of cell proliferation. Med. Biol., 65:13-20, 1987.
Epifanova, O. I-, and Polumovsky, V. A. Cell cycle control in higher eukaryotic cells: resting state or a prolonged G, period? J. Theor. Biol., 120:467477, 1986.
Bullough. W. S. Chalones and cancer. In: B. Drewinko and R. M. Humphrey
(eds.). Growth Kinetics and Biochemical Regulation of Normal and Malig
nant Cells, pp 77-89. Baltimore: Williams & Wilkins. 1977.
Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H., Raybuck, M.,
Weinberg, R. A., and Harlow, E. Association between an oncogene and an
antioncogene: the adenovirus EIA proteins bind to the retinoblastoma gene
product. Nature (Lond.), 334: 124-129. 1988.
Isaacs. J. T., Schultze. H.. and Coffey. D. S. Development of androgen
resistance in prostatic cancer. In: G. P. Murphy et al. (eds.). Prostate Cancer,
Part A, pp. 21-31, New York: Alan R. Liss, Inc., 1987.
Sonnenschein, C., and Soto, A. M. Abstract No. 663. Control of the prolif
eration of mouse androgen target cells in culture. 67th Annual Meeting of
the Endocrine Society, June 19-21, 1985, Baltimore, MD.
Horoszewicz, J. S., Leong, S. S., Kawinski, E., Kerr, J. P., Rosenthal, H.,
Chu, T. M., Mirand. E. A., and Murphy. G. P. LNCaP model of human
prostatic carcinoma. Cancer Res., 43: 1809-1818, 1983.
Soto, A. M., and Sonnenschein, C. The role of estrogens on the proliferation
of human breast tumor cells (MCF-7). J. Steroid Biochem., 23:87-94, 1985.
TUMOR CELLS
15. Olea-Serrano. N., Devleeschouwer. N., Leclercq, G., and Heuson, J. C. Assay
for estrogen and progesterone receptors of breast cancer cell lines in monolayer culture. Eur. J. Cancer Clin. Oncol.. 21:965-973. 1985.
16. Munson, P. J., and Rodbard, D. Ligand: a versatile computerized approach
for characterization of ligand-binding systems. Anal. Biochem., 707: 220239, 1980.
17. Berns, E. M. J. J.. de Boer, W., and Mulder, E. Androgen dependent growth
regulation of and release of specific protein(s) by the androgen receptor
containing human prostate tumor cell line LNCaP. Prostate. 9: 247-259,
1986.
18. Bruchovsky, N., Lesser, B., van Doom, E.. and Craven, S. Hormonal effects
of cell proliferation in rat prostate. Vitam. Horm.. 33: 61-102, 1975.
19. van Steenbrugge, G. J., van Dongen, J. J., Reuvers, P. J., deJong, F. H., and
Schroeder, D. H. Transplantable human prostatic carcinoma (PC-82) in
athymic nude mice. I. Hormone dependence and the concentration of androgens and tumor tissue. Prostate, //.' 195-210, 1987.
20. Soto, A. M., Murai, J. T., Siiteri, P. K., and Sonnenschein, C. Control of
cell proliferation: evidence for negative control on estrogen-sensitive T47D
human breast cancer cells. Cancer Res., 46: 2271-2275. 1986.
21. Shain, S. A., and Boesel, R. W. Saturation analysis of the binding of
androgens, antiandrogens and estrogens by the cytoplasmic high affinity
androgen receptor of the rat ventral prostate. J. Steroid Biochem.. 6: 43-50,
1975.
22. Baulieu. E-E., and Jung, I. A prostatic cytosol receptor. Biochem. Biophys.
Res. Commun., 38: 599-606. 1970.
23. Fang, S., and Liao, S. Steroid- and tissue-specific retention of a /3-hydroxy5«-androstan-3-one-protein complex by the cell nuclei of ventral prostate. J.
Biol. Chem., 246: 16-24, 1971.
24. James, R., and Bradshaw, R. A. Polypeptide growth factors. Annu. Rev.
Biochem., 53: 259-292, 1984.
3481
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.
Negative Controls of Cell Proliferation: Human Prostate Cancer
Cells and Androgens
C. Sonnenschein, N. Olea, M. E. Pasanen, et al.
Cancer Res 1989;49:3474-3481.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/49/13/3474
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1989 American Association for Cancer Research.