MCF-7: A human breast cancer cell line with estrogen, androgen, progesterone, and glucocorticoid receptors

785 MCF-7: A HUMAN BREAST CANCER CELL LINE WITH ESTROGEN, ANDROGEN, PROGESTERONE, AND GLUCOCORTICOID RECEPTORS

K.B, Horwitz, M.E. Costlow, and W.L. McGuire

Department of Medicine University of Texas Health Science Center 7703 Floyd Curl Drive San Antonio, Texas 78284 .',eceivecl: g/24-/73 ABSTRACT We have identified receptors for glucocorticoids,progestins, and androgens in a human breast tumor cell line (MCF-7) known to have estrogen receptor. Sucrose density gradients show that MCF-7 cytosol contains approximately 100 fm/mg protein estradiol (E2-3H) receptor, more than 300 fm/mg protein progesterone receptor (measured with R5020-3H), about 40 fm/mg protein Sa-dihydrotestosterone (S~X-DHT-~H) receptor, and 800 fm/mg glucocorticoid receptor (measured with dexamethasone-3H). D' association constants obtained by Scatchard analyses were approximately 0.6 x 10-l"M (E2), 1 x 10egM (RSOZO), 2.8 x lo-1OM (Sa-DHT) and 8 x lo-SM (dexamethasone). No cross competition was found for estrogen receptor, but progestins competed for androgen and glucocorticoid binding. The androgen, but not the glucocorticoid,partially competed for R5020 binding to progesterone receptor. This first demonstration of 4 classes of steroid receptors in human breast cancer means that MCF-7 may be an excellent in vitro model for studying the mechanism of tumor response to endocr=etherapy as well as the complex relationships between binding and biological actions of these hormones.

INTRODUCTION Although 55 to 60% of estrogen receptor containing (ER+) human breast tumors respond to endocrine ablative or additive therapy, the growth of 40% of ER+ tumors is unaffected by endocrine manipulation (1)*

Because of this sizable predictive failure, we have speculated

that a more effective marker of an endocrine responsive tumor would be a product of hormone action, rather than the initial binding step. Progesterone receptor (PgR) may be such a marker (2) since in estrogen target tissues its synthesis is modulated by estrogen.

VoZwne 26, Nwnber 6

S

TREOXDU

December, 2975

S We have recently tumors

contain

endocrine

endocrine

we also

sought

is

the pleural.

have

While response

a stable of

whose disease substantial

responded

amounts of

should

also

have PgR if,

differentiated to

glucocorticoids

model

found

response

Culture

therapy

receptor

derived

long-term

for

cases,

studying

the

obtained

from

breast

(4) _

(5).

These

cells

Based on our

responsive

from such

culture,

all

four

yet

been

in rat

it

tumor

a tumor it has maintained

for

Conditions.

tumors

mammary tumors

known to line

(6).

for

cancer,

(7).

whether,

in addition

androgens influence

has potential

androgen

Receptors

in human breast

progesterone,

steroids cell

and progesterone,

described

to determine

the hormonal

endocrine

estrogen

in human breast

receptors

studying to

individual

metastatic

an ER+ hormone

for

we sought

in MCF-7, this for

that

MCF-7 is

are present

MCF-7 contained

present

to

correlation

originally

with

to hormone

after

have not

Consequently,

If

of

system

line

patient

estrogen

receptors

have been

ticoids.

the

respond

character.

In addition

they

model

ER+ breast

which will

establish

cell

a female

have PgR, and since

although

to

in vitro --

we would predict

receptors

fraction

and PgR in a series

epithelioid

should

its

that

55% of

PgR and hormone dependence.

effusion

hypothesis

approximately

attempting

an appropriate

of

MU-7

that

PgR, approximately

between

cancer

shown (3)

therapy.

relationship

IZPEOXDII

to ER,

and glucocor-

mammary function to be an excellent

interrelationships

involved

in tumor

therapy.

MCF-7 cells,

a gift

(Michigan Cancer foundation) were received

are

of

Dr. Herbert Soule

at passage

102 and have

been serially cultured in our laboratory for approximately one year. Cells were routinely subcultured (1:5) every 3 weeks using trypsin (0.125%)-EDTA (1 mM) in Ca++Mg++ free Hank's balanced salt solution (HBSS). For the experiments described here, cells at passage number 117 to 124 were grown in Falcon plastic flasks (75 cm2) or in glass roller bottles (692 cm2). At confluency, flasks contained approximately 2 x lo7 cells, and bottles 1.6 x lo* cells. The cells were grown in 5% CO2 in air at 37'C in medium consisting of Earles-based minimal essential medium (MEM, Gibco), supplemented with non-essential amino acids (Gibco), 2mM L-glutamine, (Gibco), 0.006 ng/ml insulin (Sigma), 10% calf serum (Gibco) and 50 ug/ml gentamicin (Schering). Five days before use, growth medium was exchanged for medium containing 10% calf serum stripped of endogenous hormones by a 30 minute, 4'C incubation in a dextran-coated charcoal (DCC) pellet (0.25% Norit A, 0.0025% Dextran in 0.01 M Tris-HCl, pH 8.0, 1.5 ml/ml serum). Hydrocortisone (14 ng/ml), estradiol (27 ng/ml), insulin (6 ng/ml), ovine prolactin (1 ug/ml) and 0.1% bovine serum albumin were also added. 18 hours before the cells were assayed, this medium was replaced by MEM containing 10% stripped calf serum and insulin. Cytosol Preparation. Cells were washed with Ca++-Mg++ free HBSS, then scraped from flasks with perforated cellophane (Microbiological Associates) and pooled, or detached from glass bottles with a 15 minute, 37OC incubation in Ca++-Mg++ free HBSS containing 1 mM EDTA. The cells were then chilled to 4"C, washed twice in HBSS without EDTA, once in homogenization buffer, and resuspended in 2 packed cell volumes of buffer. The homogenization buffer (PG buffer), prepared immediately before use, was 5 mM sodium phosphate, pH 7.4 at 4OC, containing 1 mM thioglycerol and 10% glycerol (8). Cells were homogenized in a Dounce homogenizer (Kontes) using the B pestle, until they were more than 90% disrupted and clean nuclei were seen by phase contrast microscopy (30-50 strokes). Cytosol was harvested from the cellular debris by centrifugation for 50 minutes at 40,000 rpm (105,000 xg), 4'C, with a Beckman 75 Ti rotor. Cytosol protein concentration was estimated by measurement of its OD at 2601280 nm (9) and measured by the method of Lowry -et al (10). Sucrose Gradients. R5020 (17,21-dimethyl-19-nor-4,9-pregnadiene-3, 20-dione-6,7-SH, 51.4 Ci/m mol, Roussel-UCLAF). dexamethasone (9-fluoro-~l~,l~a,21-trihydrox~-16a-methyl-1,~~pregnadiene-3, 20-dione-1,2,4-3H; 10 Ci/m mol, Schwartz/Mann), estradiol-178-2,4,6, 7-3H (114 Ci/m mol, New England Nuclear) or 5a-dihydrotestosterone1,2,4,5,6,7- 3H (175 Ci/m mol, Amersham-Searle) were added in 2 ul ethanol to 250 ~1 cytosol and incubated for 4 hours at 4'C. Parallel samples were preincubated for 15 minutes with 100 fold excess unlabeled hormone (progesterone,R5020, diethylstilbestrol [4,4'-dihydroxy-a, 8-diethylstilbene],dexamethasone, or 5a-dihydrotestosterone)added in 1 ul ethanol. Pellets were prepared from a 1 ml suspension of DCC by a 10 minute centrifugation at 3200 rpm (2000 g). The charged cytosol was transferred onto the pellet, mixed, and incubated 15 min

at 4OC to adsorb unbound radioactivity, After recentrifugation for 10 minutes at 3200 rpm, a 200 ~1 aliquot of the supernatant was layered onto a 5 to 20 percent sucrose gradient prepared in the homogenization buffer. Bovine serum albumin-14C (ll), 1500 cpm/lO ul buffer was added to each gradient as an internal marker. Gradients were centrifuged in a Beckman SW56 rotor at 50,000 rpm (246,000 g,,) for 17 hours. Four drop fractions were collected and counted in 5 ml modified Bray's solution (12). Scatchard Analysis. Saturation kinetics of estradiol and dexamethasone binding were determined using the dextran-coated charcoal method (13). 200 ul cytosol prepared as above was incubated with increasing quantities of 3H-hormone added in 50 ul PG buffer. Parallel incubations contained excess nonradioactive hormone. After 16 hours (estradiol) or 4 hours (dexamethasone)of incubation at 4OC, 0.5 ml of the charcoal suspension was added, and the mixture was vigorously shaken for 30 minutes in an Eberbach shaker (estradiol) or mixed and incubated 15 minutes (dexamethasone). The charcoal was sedimented with a 10 minute, 3200 rpm centrifugation, and an aliquot of the supernatant was counted in 5 ml Bray's solution. The data was analyzed according to Scatchard (14), after subtraction of nonspecific binding calculated from the preparations competed with nonradioactive hormone. Kinetics of R5020 and So-dihydrostestosteronebinding were assayed using protamine precipitated receptor (15). Acid washed 12 x 75 mm glass tubes were incubated at 30' for 10 minutes with 1 ml PG buffer containing 0.1% bovine serum albumin. The tubes were then rinsed with 1 ml ice-cold PG buffer, and subsequent procedures were carried out at 0' unless otherwise stated. Protamine sulfate (LJSP injection, without phenol preservative, Eli Lilly Co.) was diluted to 1 mg/ml in PG buffer and 250 ul was added to each tube. After addition of 200 ul cytosol, the mixtures were vortexed and the precipitates were sedimented by centrifugation at 2000 g for 10 minutes. The precipitated receptors were incubated with 'H-hormone of increasing concentrations added in 500 1.11 PG buffer. Parallel incubations contained lOO-fold excess unlabeled hormone. Tubes were incubated at 4'C for 4 hours (progesterone)or 16 hours (androgen), then washed three times with 1 ml cold PG buffer. Bound radioactivity was extracted twice with 1 ml 100% ethanol at room temperature. The extracts were dried in scintillation vials and counted in 5 ml fluor (4.0 gm PPO, 0.05 gm POPOP, 1 liter toluene) in a Beckman LS233 counter. The data was analyzed as above.

RESULTS We have used the synthetic steroids R5020 and dexamethasone to measure progesterone and glucocorticoid receptors because they do not bind to corticosteroid binding globulin (CBG) (16, 17), a

component of the serum used in the cells' growth medium. The fact that specific receptors sedimented primarily at 8s while CBG sediments only at 45 also aids in distinguishing the two types of binding. Although estradiol and dihydrotestosteronecan also bind to a high affinity serum protein (sex steroid binding globulin [SBG]) which sediments at 4S, no such binding appeared in MCF-7

cytosols. We

therefore used the natural sex hormones to measure high affinity cytoplasmic receptors. After a year of serial trypsin passages in our laboratory the MCF-7 cell line retained estrogen receptor (Fig. 1). The number of sites as determined from sucrose gradients was more than 100 fm/mg cytosol protein. This value is slightly higher than the 63 fm/mg originally described (4). The dissociation constant (Kd) was 0.57 x 10-1°M (Fig. 2) , 50 times lower than that reported by Brooks et al (5). This --

lower Kd more nearly approximates the average value

obtained in human breast tumors (1). We have found that MCF-7 has, in addition, receptors for the other steroid hormones known to control mammary gland function (Fig. 1). The cells contained more than 300 fm/mg protein progesterone receptor, measured with R5020-3H (or progesterone-3H, data not shown), about 40 fm/mg androgen receptor, and over 800 fm/mg glucocorticoid receptor. Scatchard plots of the specific binding are shown in figure 2.

The Kd for R5020 was approximately 1.1 x lO_gM,

for DHT was

2.8 x lo-I'M and for dexamethasone was 8.1 x 10s9M. To demonstrate that 4 distinct classes of receptors were being measured, each hormone was used as a competitor for the others (Table 1).

s

790

WDmOXDI

c\

R-5020

4

0

1

IO

I

20

I

IO

TOP

20

TOP

DEX

I

TOP FRACTION NUMBER

IO

1

20 FRACTION NUMBER

1

TOP

Figure 1. Estrogen, progesterone, androgen and glucocorticoid receptors in MCF-7. Sucrose density gradients of cell cytosols (3.45 to 7.50 mg/ml protein) were incubated with estradiol-176-3H (4 x lo-DM), R5020-3H (2 x lO_SM), 5a-dihydrotestosterone-3H (DHT, 2 x 10m8M) or dexamethasone-3H (5 x lo-SM), in the presence (0) or absence (0) of 100 fold unlabeled The arrow indicates diethylstilbestrol, R5020, DHT or dexamethasone. 4.6s sedimentation of bovine serum albumin -14C.

I

0

,

t

0.02 Cl04 0.06 B R-5020 Wl)

B ESTRADIOL (nM)

0.12

0.08

0.04

-I-

I

1

I

0.02 0.04 0.06 B DHT (nM)

O

k m0

I

0.6 B DEX (nM)

I

I.2

Figure 2. Scatchard plots of estrogen, R5020, dihydrotestosterone (DHT) and dexamethasone binding to MCF-7, Cytosol (1.83 mg/ml protein) binding of 5 x lo-l1M to 8 x 10"9M estradiol-17$-5H; cytosol binding of 8 x lo-1OM to 2.5 x lo-8M dexamethasone-3H assayed by the dextran coated charcoal method. C tosol (1.22 mg/ml protein) binding of 4 x lo-llM to 10m8M R5020-?!I H; cytosol (2.62 mg/ml protein) binding of 5 x 10-llM to 8 x 10m9M dihydrotestosterone-5Hassayed by the protamine sulfate method. Nonspecific binding has been subtracted as indicated in Methods.

TABLE I Percent inhibition of estradiol-17&3H, R5020-3H, YKxDHT-~H and dexamethasone-3Hbinding in MCF-7 cytosols by unlabeled competitors

Competitor

Diethylstilbestrol R5020 Progesterone Dihydrotestosterone Dexamethasone

E2-3H 8S 4S

94 0 4 10 6

R5020-3H 8s 4s

-

14 94 90 55 5

5 86 79 35 7

SCXDHT-~H 8s 4S

5 82 78 93 0

-

Dex-3H 8S 4S

8 91 87 0 93

24 73 68 0 77

Cell cy$osols (3.45 to 7.50 mg/ml protein) were incubated with 4 x lo- M estradiol-17B-3H, 2 x 10V8M R5020-3H, 2 x lo-'M 5clDHT-3Hor 5 x lo-'M dexamethasone-3H. Parallel samples were preincubated with 100 fold excess unlabeled hormones as shown. Competition was determined from percent depression of the area under the peaks in sucrose density gradients. (-) indicates absence of 4S peak.

Estrogen Receptor. Only diethylstilbestrol (an estrogen which does not bind SBG) was able to compete for binding of estradiol-3H. The glucocorticoid, progestins, and androgen were ineffectual. Progesterone Receptor. Progesterone and R5020 completely suppressed R5020-3H binding. No effective competition was seen with the estrogen or glucocorticoid. Dihydrotestosterone,however, depressed the 8S binding of R5020 by more than 50%. Androgen Receptor. Unlabeled dihydrotestosterone,R5020, and progesterone inhibited binding of So dihydrotestosterone-3H. The estrogen and glucocorticoid were poor competitors. Glucocorticoid Receptor. Unlabeled dexamethasone, R5020, and progesterone inhibited binding of dexamethasone-3H. The estrogen and androgen were poor competitors.

S

793

TlBBlOIDDISCUSSION

We have androgens All

MCF-7 cells

receptor-bound

sucrose

90% of

binding

of

the binding

sedimented

plots

100-fold

a major of

the

criterion

were

binding,

confirming

the presence

of

were

agreement

these

steroids

in other

Our findings

of

and glucocorticoids known to have

provides

a mechanism

who have

shown that stimulate

repeatedly

demonstrated

two classes

Our data

makes the

competed

for

dexamethasone

of

for

findings

steroids

latter

glucocorticoid failed

(17,22),

to

of

operate

possibility binding, suppress

reported

agents bind

androgen

with

the

binding

androgen

et al --

That

has been

a single

reverse of

(Zl),

in the

has even been proposed

unlikely.

are

receptor

androgens

glucocorticoids

through

for

androgens

B-glucuronidase.

and it

high

of Mowszowicz

can synergize

binding

of

progestins,

and many progestins

an androgen-inducible

can compete

nonspecific

Progestational

for

increasing

19).

between

the

progestins

of

binder

(18,

specificity.

with

those

were

indicating

cytosol

with

progestins

to explain

progestins

these

of

of

hormone,

receptor

subtraction

surprising.

Competition

of

and cells

effects,

8s region

cases

of

a principal

cross-competition

androgenic

(20) .

to

tissues

are not

receptor

kidney

in good

after

receptor.

unlabeled

in all

concentrations

The Kd’s

progestins,

45 components

excess

hormone

affinity.

in the

interaction

linear

for

to an estrogen

was abolished

was saturable,

The Scatchard

receptors

R5020 and dexamethasone,

In the presence

seen.

contain

in addition

hormones

with

gradients;

more than that

that

and glucocorticoids

four

also

found

receptor

Although was not

R5020.

that (22).

progestins true,

since

794

S

TPlEOXDI

We had predicted that progesterone receptors would be present in this cell line, since it was derived from an ER+, endocrine responsive tumor. That it also contained receptors for androgens and glucocorticoidswas fortuitous. To our knowledge, the finding of four classes of steroid receptors in one tissue is unique.

It means

that the MCF-7 line should prove to be extraordinarily useful in studying not only mechanisms of tumor endocrine response, but also the complex interrelationshipsbetween the binding and biological actions of these four steroids.

ACKNOWLEDGMENTS We thank Doctor J.-P. Raynaud for providing R5020. This work has been aided by the USPHS CA-11378, CB-23862, the American Cancer Society BC-23D, and the Robert A. Welch Foundation.

REFERENCES 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

McGuire, W.L., Pearson, O.H. and Segaloff, A. In ESTROGEN RECEPTOR IN HDMAN BREAST CANCER (W.L. McGuire, P.P. Carbone and E.P. Vollmer, eds.). Raven Press, New York, p. 17 (1975). Horwitz, K.B. and McGuire, W.L. STEROIDS 25, 497 (1975). Horwitz, K.B., McGuire, W.L., Pearson, 0-H. and Segaloff, A. SCIENCE 189, 726 (1975). Soule, H.D., Vazquez, J., Long, A., Albert, S. and Brennan, M. J. NATL. CANCER INST. 51, 1409 (1973). Brooks, S.C., Locke, E.R. and Soule, H.D. J. BIOL. CHEM. 248, 6251 (1973). Poortman, J., Prenen, J.A.C., Schwarz, F., Thijssen, J.H.H. J. CLIN. ENDO. AND METAB. 40, 373 (1975). Gardner, D.G. and Wittliff, J.L. BIOCHIM. ET BIOPHYS. ACTA 320, 617 (1973). Faber, L.E., Sandmann, M.L. and Stavely, H.E. FED. PROC. 22, 229 (1973). Layne, E. In METHODS IN ENZYMOLOGY, Vol. III (S.P. Colowick and N.O. Kasan, eds.) pp. 447-454 Academic Press, N.Y. (1957). Lowry, O.H., Rosebrough, N-J., Farr, A.L. and Randall, R.J. J. BIOL. CHEM. 193, 265 (1951). Rice, R.H. and Means, G.E. J. BIOL. CHEM. 246, 831 (1971).

12. Bray, G.A. ANAL. BIOCHEM. 1, 279 (1960), 13. McGuire, W.L. and De La Garza, M, J. CLIN. ENDO. AND MBTAB. 37, 986 (1973). 14. Scatchard, G. ANN. N.Y. ACAD. SCI. 51, 660 (1949). 15. Chamness, G.C., Huff, K. and McGuire, W.L. STEROIDS 25, 627 (1975). 16. Philibert, D. and Raynaud, J.-P. ENDOCRINOLOGY 94, 627 (1974). 17. Rousseau, G.G., Baxter, J.D. and Tomkins, G.M. J. MOL. BIOL. 67, 99 (1972). King, R.J.B. and Mainwaring, W.I.P. In STEROID-CELL INTERACTIONS. 18. University Park Press, Baltimore (197q. 19. Shyamala, G. BIOCHEMISTRY 12, 3085 (1973). 20. Bullock, L.P., Barthe, P.L., Mowszowicz, I., Orth, D.N., and Bardin, C.W. ENDOCRINOLOGY 97, 189 (1975). 21. Mowszowicz, I., Bieber, D.E., Chung, K.W. Bullock, L.P. and Bardin, W.L. ENDOCRINOLOGY 95, 1589 (1974). 22. Wittliff, J.L. SEMINARS IN ONCOLOGY 1, 109 (1974).