• Abstract and Figures
• Public Full-text

[CANCER RESEARCH 45,167-173, January 1985] Growth Inhibition and Increase of Insulin Receptors in Antiestrogen-resistant T47DCOHuman Breast Cancer Cells by Progestins: Implications for Endocrine Therapies1 Kathryn B. Horwitz2 and Gary R. Freidenberg3 Departments oíMedicine and Biochemistry, Biophysics, and Genetics, University of Colorado Health Sciences Center, Denver, Colorado 80262 ABSTRACT and the duration of remissions ranges from several months to 1 year or more (1, 25, 31, 33, 39, 43, 45). These results are There is renewed interest in the use of progestins to treat comparable to those obtained with antiestrogens, the most advanced breast cancer because results with these agents are common drug used for endocrine therapy (16, 39, 44). It is not comparable to those obtained with antiestrogens. However, it is known whether progestins act directly on the regressing tumors not known whether progestins inhibit the growth of breast tumor or indirectly through other hormones, (b) In support of the cells directly and independently of estradici. To study this, we hypothesis of Horwitz et al. (20), presence of PR4 may be the have used T47Dco human breast cancer cells. The progesterone single best marker for predicting both the hormone dependence receptors in these cells do not require estrogen induction, and of tumors and the disease-free survival of patients (11, 35). This this permits study of pure progestin effects without interference has led to speculation that PR-rich tumors would be especially by estradiol. We report here that, in the absence of estradici, sensitive to progestin treatment (26). (c) In small series of studies, physiological concentrations of progestins directly inhibit prolif positive responses to progestins have been obtained in tumors eration of these cells. At the same time, progestins increase the that have failed to respond to or have become resistant to levels of the receptors for insulin, a common cell mitogen. Ten antiestrogens or other endocrine therapies (14) or in tumors days of treatment with 1 or 10 nw of the synthetic progestin lacking ER (25, 39, 48). This suggests that some of the mecha R5020 suppresses cell growth approximately 50 to 60%. This is nisms responsible for the antitumor effects of progestins and consistent with the concentrations that either partially (approxi antiestrogens must differ. mately 10%) or more extensively (>60%) translocate cytoplasmic There is little experimental evidence underlying any of these progesterone receptors. Even a brief 1-hr pulse of R5020 has assumptions. It has been difficult to study the direct biological long-term growth-inhibitory effects. Progesterone is also antipro- actions of progestins separate from those of estrogens, because liferative, but its effects are attenuated because, unlike R5020, estrogen priming is required to maintain elevated, and presum it is rapidly metabolized in the medium. Other synthetic proges ably functional, levels of PR (19). Thus, the contaminating pres tins also inhibit cell growth, but unrelated steroids (estradiol, ence of estradiol in the system under study has precluded androgens, glucocorticoids, 1,25-dihydroxyvitamin D3) are inef accurate assessment of true progestin effects in vivo. Further fective. While growth is suppressed by R5020, insulin receptors more, for reasons that are not entirely clear, progesterone action increase rapidly and then fall to a new, elevated steady state as has also been difficult to study in vitro, where conditions can the cells slowly begin to proliferate. Only progestins have this usually be more carefully controlled. As a result, there are several effect on insulin receptors. We conclude that the hormonal established cell lines that are used for studies of glucocorticoid regulation of breast tumor cell growth is complex and includes and estrogen action in vitro and that contain estrogen-inducible progestins among the regulating factors. Furthermore, since PR but in which direct responses to progesterone have not been T47Dco cells are antiestrogen-resistant and estrogen receptor- demonstrable. negative, the antiproliferative effects of progestins must be me T47D is a relatively new human breast cancer cell line that has diated through mechanisms that differ from the cytotoxic effects PR and in which progesterone responses have recently been of antiestrogens. We propose that, clinically, antiestrogens and described (9, 10, 27, 49). In the "clone 11" subline, which has progestins may have complementary uses in breast cancer ER and responds to estrogens, progestins have antiestrogenic treatment, and we outline two therapeutic strategies. actions. They suppress growth-stimulatory effects of estrogens but have no growth effects alone (9,10). A variant subline in our INTRODUCTION laboratory, T47Dco, contains unusually high levels of PR that are independent of estrogen controls. The cells lack ER and are For several reasons, there is renewed interest in the use of antiestrogen resistant; their PR, although structurally and bio synthetic progestins to treat men and women who have ad chemically normal, are neither induced by estradiol nor sup vanced breast cancer, (a) Clinical results using pharmacological pressed by antiestrogens (17, 21, 22, 37). These cells can progestin concentrations have been very encouraging: the drugs therefore be used in estrogen-free conditions, to assess the are well tolerated, remissions occur in 35 to 45% of patients, direct biological actions of progestins, distinct from their anties trogenic ones. We have now studied the role of progestins on 1This work was supported by Grant CA-26869 from the NIH. 1Recipient of Research Career Development Award CA-00694 from the NIH. cell growth and on the receptors for insulin, a common cell To whom requests for reprints should be addressed. mitogen. The studies show that progestins alone inhibit cell 3Presentaddress: Departmentof Medicine,Universityof Californiaat San Diego, 4The abbreviations used are: PR, progesterone receptors; ER, estrogen recep La Jolla, CA 92093. ReceivedApril 2, 1984; accepted September 27, 1984. tors. CANCER RESEARCH VOL. 45 JANUARY 1985 167 Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. GROWTH REGULATION BY PROGESTINS growth and increase insulin receptor levels at concentrations RESULTS that bind to and translocate PR. These direct actions of proges tins, distinct from their antiestrogenic ones, have important im Progestins and Cell Growth. Chart 1 shows that the doubling plications for the design of endocrine therapies against breast time of log-phase T47D«,cells in medium containing 5% char cancer, and 2 treatment strategies using progestins are de coal-stripped fetal calf serum is approximately 54 hr. Addition of scribed. 1 nw R5020, a synthetic progestin, slowed the doubling time to approximately 72 hr; at 10 nw R5020, the doubling time was 5 days. As a result, after 10 days, progestin-treated flasks had MATERIALS AND METHODS only 49% (1 nM R5020) and 39% (10 nw R5020) of the cell Hormones. [3H]R5020 (17,21-dimethyl-19-nor-4,9-pregnadiene-3,20- number present in control flasks. Though cell growth is slowed dione) and unlabeted R5020 were obtained from New England Nuclear by the presence of R5020, the cells are not killed, but continue (Boston, MA). Other steroids were from Sigma Chemical Co. (St. Louis, to proliferate sluggishly through many passage generations. MO), except 1,25-dihydroxyvitamin D3, which was a gift from David Normal growth rate is gradually restored in such cells once the Feldman (Stanford University School of Medicine). 125l-Ai4porcine insulin hormone is removed (data not shown). (340 to 360 /iCi///g) and unlabeled insulin were supplied by Eli Lily Table 1 compares the growth-inhibitory properties of R5020 (Indianapolis, IN). with those of progesterone and shows, at the same concentra Cell Culture and Growth Studies. The routine culture of the T47D subline (T47DÅ“) of human breast cancer cells has been described (21). tions, the ability of the 2 hormones to translocate PR. The data Briefly, cells were grown in buffered RPMI Medium 1640 containing 5% show that, like R5020, progesterone also suppresses cell growth heat-inactivated fetal calf serum, nonessential amino acids, glutamine, but that it does so less efficiently. There are 2 reasons for this: penicillin, streptomycin, and porcine insulin (6 ng/ml). Stock cells were first, a higher concentration of progesterone than of R5020 is plated at a density of 4 x 10* celts (a 1:4 split) in 175-sq cm plastic flasks required to translocate the same number of cytoplasmic PR in and grown at 37°in a humidified atmosphere of 5% CO2 and air. For intact cells (Table 1); and, second, progesterone is rapidly (fw, growth studies, cells were harvested at confluence, replated in 25-sq cm approximately 2 hr) metabolized in the medium (22). In practice, flasks at the densities described in the figures for 18 to 48 hr, and then therefore, the cells are exposed to progesterone for only a few switched to modified insulin-free medium containing 5% heat-inactivated fetal calf serum stripped of endogenous hormones by 30 min of incuba tion at 45° with a dextran-coated charcoal pellet [0.25% Norit A, 0.0025% dextran in 0.01 M Tris-HCI (pH 8.0) at 4°;1 ml charcoal per 2 350 ml serum]. Steroids were prepared in ethanol at 1000-fold final concen tration and added to the medium. At harvest, the cells were suspended and well dispersed and, except where noted, triplicate aliquots from 300 Control triplicate flasks were taken for DNA and protein analyses. The determi nation of DNA was performed according to the method of Burton (7), and protein content was measured by the method of Lowry ef al. (32). Progesterone Receptors. The in situ receptor binding affinity of 250 progestins was assessed by their acute ability to translocate PR. Briefly, cells were incubated with progestins for 5 min at 37° as described previously (37). They were then harvested, cooled, and homogenized, and cytosols were prepared. These were incubated at 4°for 4 hr with 200 20 nM [3H]R5020 only or together with a 100-fokJ excess of unlabeled R5020 to measure nonspecific binding.

Load more

  8

Copy Link