- Email: [email protected]
Responses to estradiol in a human endometrial adenocarcinoma cell line (Ishikawa)
J. sferoid
Biochem.
Vol.
24, No.
1, pp.
85-89,
1986
0022-473
Printed in Great Britain. All rights &served
I,‘86$3.00
+ 0.00
Copyright Q 1986Pergamon Press Ltd
III. Steroids and Cancer RESPONSES TO ESTRADIOL IN A HUMAN ENDOMETRIAL ADENOCARCINOMA CELL LINE (ISHIKAWA) C. F. HOLINKA*$, H. HATA*, H. KURAMOTO~ and E. GURPIDE* *Department of Obstetrics, Gynecology and Reproductive Science, Mount Sinai School of Medicine, New York, NY 10029, U.S.A. and tDepartment of Obstetrics and Gynecology, Kitasato University, Kanagawa-Ken, Japan
Summary-A human endometrial adenocarcinoma cell line (Ishikawa) has been found to be estrogen responsive. The growth stimulatory effects of estradiol (IO-’ M) could be clearly demonstrated when cell cultures containing the hormone were compared with the maximal cell density achieved in control cultures. The approx. 3-fold increase in cell density observed 2-3 weeks after plating, with frequent medium changes, could by blocked by a tOO-fold molar excess of the antiestrogen trans-~monohydroxy~moxifen. When added to hormone-free cultures that had reached a plateau level of cell numbers on day 14 after plating, estradiol (low8 M) caused the resumption of proliferation: after 6 days in the presence of the hormone, the cultures contained nearly twice the cell numbers of controls. Effects of estradiol on Ishikawa cells were also evident from the several-fold increases in the levels of specific progesterone binders provoked by the hormone at lO-q--lO-6 M concentrations. Cells injected into nude mice formed tumors which contained estrogen and progesterone binders. The availability of a fast-growing (doubling time approx. 30 h) endometrial cancer cell line responsive to estradiol at near physiologic levels will facilitate biochemical studies of hormonal effects on the human
endometrium.
Studies on the effects and mechanisms of action of hormones on human target tissues have been facilitated by the availability of hormone responsive cell lines, like MCF-7 cells derived from a mammary tumor [l-3]. Unfo~unately, none of the established endometrial cell lines have been clearly responsive to ovarian hormones at near physiologic concentrations, even though some of them are endowed with specific estrogen binders. Previous studies at these laboratories have focused on two human cell lines, HEC- 1 and HEC-50, derived poorly moderately or specimens of from differentiated endometrial adenocarcinomas. These cells did not respond to either estradiol (E,) and, perhaps as a consequence, to progesterone (P), as determined by evaluating cell proliferation, levels of P receptors and activities of ornithine decarboxylase, peroxidase, alkaline phosphatase or estradiol 17/1 dehydrogenase. We now report responsiveness to estrogens of a human endometrial adenocarcinoma cell line (Ishikawa). Addition of E, to the medium had a marked growth promoting effect and enhanced specific binding of P. In addition, E2 also stimulated alkaline phosphatase and DNA polymerase 01activities in Ishikawa cells, as shall be reported elsewhere. EXPERIMENTAL
Cells Ishikawa cells were established as a permanent cell line by Nishida et ul.[4] from cells derived from a $To whom correspondence should be addressed. 85
specimen of well-differentiated endometrial adenocarcinoma. At our laboratory, the cells were cultured in minimum essential medium (Eagle) with Earle’s salts (MEM, Grand Island Biological Co., Grand Island, N.Y.) supplemented with 15% fetal bovine serum (FBS, Grand Island Biological Co.). In experiments aimed to determine hormonal effects, the serum was pretreated with charcoal to remove steroid hormones, as follows: 100 ml FBS mixed with 0.25 g activated charcoal (Sigma Chemical Co., St. Louis, MO.) and 0.025 g dextran (clinical grade, Sigma) was stirred at 56°C for 30 min and centrifuged to separate the dextran-coated charcoal pellet from the supernatant, which was again subjected to the same treatment at 37°C. The charcoal-stripped serum (SFBS) was filtered through a 0.20pm sterilization unit (Nalge Co., Rochester, N.Y.) and stored at -20°C. Growth kinetics Confluent cultures of Ishikawa cells in MEM-15% SFBS were exposed to a trypsin (O.O5%~E~TA (0.02%) solution (Flow Laboratories, McClean, Va) for lO--15min at 37°C and the collected cells were seeded into 6 cm culture dishes (Falcon Plastics, Los Angeles, Calif.) at a density of 0.25 x IO6cells/dish in MEM-15% SFBS medium confining E, (lo-* M, Steraloids, Wilton, N.H.), trans-4-monohydroxytamoxifen (OHTam; 10s6 M, ICI, Macclesfield, U.K.), a mixture of E, (10e8 M) and OHTam (10m6M), or only ethanol (0.2% v/v) to reproduce the concentration of the vehicle used for the addition of steroids to the other dishes. Medium was renewed on day 3 and 6 after plating and subsequently every
C. F. HOLINKA et al.
86
other day. Cells were harvested by trypsin-EDTA treatment and counted in a Neubauer-Levy hemacytometer under an inverted microscope. Hormone binding Cells from confluent cultures were seeded into IOcm dishes in a 1:2.5 split ratio in MEM-15% SFBS. After 1 day, the medium was changed to MEM-15% SFBS containing E, at various concentrations. The final levels of ethanol in each dish, including controls, was 0.1%. Cells were harvested 3 days after hormone addition by trypsin-EDTA treatment. Trypsin action was stopped by addition of SFBS (7.5%) and cells were pelleted and homogenized in the assay buffer (50 mM Tris-HCI, 15 mM sodium phosphate, 20% glycerol, 0.01% sodium azide and 1 mM dithiothreitol, pH 7.8). Specific hormone binding was measured by the hydroxylapatite (HAP) method [5] as follows. Aliquots of the homogenate (180 ~1) containing 1.5-3.5 mg/ml total protein. measured by the method of Lowry et a1.[6], and 0.25-0.58 mg/ml DNA, measured by the method of Burton[7], were transferred to 7 ml glass scintillation counting vials (Kimble. Toledo, Ohio) and mixed with 20 ~1 of an ethanol-assay buffer solution of either [‘HIP (20 nM) and cortisol (2 PM), to determine total binding, or [3H]P (20 nM), cortisol (2 PM) and P (2 PM), to measure nonspecific binding. The final ethanol concentration was 1%. At the end of a 3 h incubation period on ice, 0.25 ml of a 70% suspension of buffer-washed HAP (Biorad, Richmond, Calif.) was added to the assay mixture. After further incubation for 10min on ice, a HAP plus nuclear pellet was separated by centrifugation and washed with 2 ml of the assay buffer containing 0.2% Triton X-100 (Sigma) and then twice with 2 ml of assay buffer. The mixture was vortexed and centri-
fuged at 1800g for 5 min after each washing. The pellets were then suspended in Dimiscint (National Diagnostics, Somerville, N.J.) and radioactivity was measured with a liquid scintillation counter (LS 8000, Beckman Instruments, Palo Alto, Calif.). The same assay buffer and experimental procedures were used to determine specific EZ binding with the following changes: homogenates were incubated at 30°C in [3H]E2 (20 nM) solutions to determine total binding, and in [3H]E2 (20 nM) plus E? (2pM) solutions to measure nonspecific binding. Tumor formation in nude mice Approximately IO’ Ishikawa cells were S.C. injected into the anterior dorsolateral region of nude mice (BALB/c, nu/nu, females). Tumors were harvested 3 weeks after inoculation, portions were sectioned and stained with hematoxylin-eosin and the rest of the tissue was used to measure E, and P binding levels. RESULTS
Figure 1A presents a photomicrograph of Ishikawa cells in culture. The multilayer characteristic of these cultures is evident at the crater-shaped area and in several other regions. Figure 1B shows a section of a tumor removed 11 weeks after inoculation of Ishikawa cells into a nude mouse. Measurement of specific binding of [3H]E2 and [3H]P in three such tumors showed the presence of receptors for estrogens (160, 1000 and 1200fmol/mg DNA) and progestins (250, 240 and 200 fmol/mg DNA). Figure 2 shows growth curves corresponding to parallel cultures of Ishikawa cells in MEM-15% SFBS, either alone or in the presence of E, (1 O-’ M). OHTam ( 10m6 M) or a mixture of these compounds. Control cultures reached maximal cell numbers at 14
Fig. 1A. Phase-contrast photomicrograph of a high-density culture of Ishikawa cells. This micrograph illustrates the overgrowth of cells into several layers, as evident at the crater-shaped (x 110).
photoregion
Effects of estradiol on Ishikawa cells
Fig. t B. Photomicrograph
of a tumor formed in a nude mouse after S.C.injection of 10 million lshikawa cells into the dorsotateraf area (x 340).
days. At that time, cultures
with E, (lWs tained twice as many cells as controls, numbers still increasing to over three times trol levels on day 18. Hydroxytamoxifen
M) conwith ceil the con-
(10m6M) had no effect on growth but counteracted markedly the actions af E?. Figure 3 indicates that E, caused the resumption of proliferation when added to high-density control cultures on day 14. at a time when the cuhures had 18 -
14
reached a plateau. After 6 days of hormone exposure these cells contained twice the cell numbers found in parallel control dishes. As apparent from both Figs 2 and 3, E2 affected primarily cell density and only to a minor extent growth rates. The doubling time in hormone-exposed and control cultures was similar (approx. 30 h). Figure 4 illustrates the effect of E, on the Ievets of specific binding of P when present in the medium for 3 days at concentrations ranging from 1W9 to 10-6M. Significant increases in binding relative to
Control
0
16 -
87
A E, . Ez+OH-TRM x OH-TAM
12 c a 2
lo-
p: fs g
a-
f 6-
4-
2-
0
3
6
8
10 12 14 16
18
Days in culture
Fig, 2. ~r~wt~-prorn~tin~ etkt of &radio1 in Isbikawa c&s and its reversal by OHTam. Each point represents the mean of three dishes. The SD was below IO% of the mean in ail cases.
Days
in culture
Fig. 3. Resumption of cell prolife~tion after addition of estradioi to hj~-density cultures at a plateau level. Each point represents the mean of three dishes. The SD was below IO% of the mean in all cases.
88
C.
0
lo-!+ 10-0
10-7
F. HOLINKA et al.
10-6
F23
Fig. 4. Enhancement of specific P binding by estradiol. The hormone was added in the indicated concentrations to confluent cultures and P binding was measured 3 days thereafter. Values represent the mean + SEM of three dishes, assayed in triphcate. (*P < 0.02 relative to control; Student’s f-test.)
controls (1.6- to 3-fold) were noted, with maximal increases at 10e8 M E2. Since the DNA/protein ratio
in the cells was not affected by E,, P binding patterns were not different when related to protein or DNA content. DISCUSSION
Responsiveness of Ishikawa cells to E, has been demonstrated in this study by evaluating growthpromoting effects and enhancement of specific P binding levels. The responses to E, indicate that the cells possess functional estrogen receptors; we found 33Ofmol/mg DNA specifically bound E, in cultures at the plateau phase 18 days after plating and 3 days after the last change of medium. However, we do not attempt to characterize Ishikawa cells by their recep tor content since previous experience with endometrial cells in monolayer made us aware of marked changes in these levels with time, which were found to reflect varying ratios of intracellular concentration of cGMP and cAMP[8]. In the present study, we observed a targe dispersion of specific P binding levels among samples of confluent cells from different cultures. However, parallel cultures of cells with identical chronology showed close agreement in P binding, as evident from triplicate values for each time point in the experiment described in Fig. 4. The present study clearly indicates that the E2 effects on growth of Ishikawa cells are mainly evident by comparing the maximal cell density achieved in control cultures with the density of parallel cultures in the presence of the hormone. Under these experimental culture conditions, using media containing 15% FBS, no obviously significant E2 effects on the doubling time during the exponential phase of proliferation were observed. The resumption of proliferation obtained by adding E to control cultures
with cell numbers stabilized at a plateau level indicates that the hormone is capable of overcoming density-de~ndent growth arrest. Whether Ez acts directly in promoting cell division or indirectly by enhancing autocrine growth factors or blocking the production of such growth factors remains to be examined. The availability of estrogen responsive (Ishikawa) and estrogen unresponsive (HEC- 1, HEC-50) endometrial aden~arcinoma cells, all containing specific estrogen binders [9], offers the possibility of studying factors affecting responsiveness to hormones. Furthermore, production of hormonally responsive tumors in the nude mouse makes possible the investigation of strategies for the regulation of human endometrial cancer growth in experimental animals. It is reassuring that results obtained with endometrial cancer cells in culture can be reproduced in fresh endometrial tissue, such as the regulation of estrogen binding by cyclic nucleotides [8]. Another interesting characteristic of ishikawa cells is their lack of responsiveness to P, as determined by evaluating the effect of P on estradiol 178 dehydrogenase activity or as an antagonist to the E, effects on cell numbers and alkaline phosphatase activity (unpublished). Since basal and E,-induced levels of P receptors are clearly present in Ishikawa cells, these cells may be useful in the search for explanations of unresponsiveness to progestins in endometrial cancers that possess P receptors. Acknowledgements-This
work was supported by Grants
CA 15648, awarded by the National Cancer Institute, and HD 07197, awarded by the National Institutes of Child Health and Human Development. REFERENCES 1. Berg C. D., Nawata H., Bronzert D. A. and Lippman
2.
3.
4.
5.
6.
M. E.: Altered estrogen and anti-estrogen responsiveness in clonal variants of human breast cancer cells. In Progress in Cancer Research and Therap) (Edited by F. Bresciani, R. J. B. King, M. E. Lippman, M. Namer and J.-P. Raynaud). Raven Press, New York, Vol. 31 (1984) pp. 161-170. Edwards D. P., Adams D. J. and McGuire W. L.: Estrogen regulation of growth and specific protein synthesis in human breast cancer cells in tissue culture. In Hormones and Cancer (Edited by W. W. Leavitt). Plenum Press, New York, Vol. 138 (1982) pp. 133~-149. Rochefort H., Chalbos D., Capony F., Garcia M., Veith F. and Westley B.: Effect of estrogen in breast cancer cells in culture: released proteins and control of cell proliferation. In Hormones and Cancer (Edited by E. Gurpide, R. Calandra, C. Levy and R. J. Soto). Liss, New York, Vol. 142 (1984) pp. 37-51. Nishida M., Kasahara K., Kaneko M. and Iwasaki H.: Establishment of a new human endometrial adenocarcinoma cell line, Ishikawa cells, containing estrogen and progesterone receptors. Acta obstet. gynaec. japonica 37 (1985) 1103-l 11I. (In Japanese.) Hoffman P. G., Jones L. A., Kuhn R. W. and Siiteri P. K.: Progesterone receptors: saturation analysis by a solid phase hydroxylapatite adsorption technique. Cancer 46 (1980) 2801-2804. Lowry 0. H., Rosenbrough N. J., Farr A. L. and
Effects of &radio1 Randall R. J.: Protein measurement with the Folin phenol reagent. J. biol. Chem. 193 (1951) 2655275. 7. Burton K.: A study of the conditions and mechanisms of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid. Biochem. J. 62 (1956) 315-323. 8. Fleming H.. Blumenthal R. and Gurpide E.: Rapid
on Ishikawa
cells
89
changes in specific estrogen binding elicited by cGMP and CAMP in cytosol from human endometrial cells. Proc. natn. Acad. Sci. U.S.A. 80 (1983) 2486-2490. 9. Fleming H., Blumenthal R. and Gurpide E.: Characteristics of cyclic nucleotide dependent regulation of cytoplasmic E2 binders in cultured endometrial and breast cells. J. sreroid Biochem. 20 (1984) 5-9.
Biochem.
Vol.
24, No.
1, pp.
85-89,
1986
0022-473
Printed in Great Britain. All rights &served
I,‘86$3.00
+ 0.00
Copyright Q 1986Pergamon Press Ltd
III. Steroids and Cancer RESPONSES TO ESTRADIOL IN A HUMAN ENDOMETRIAL ADENOCARCINOMA CELL LINE (ISHIKAWA) C. F. HOLINKA*$, H. HATA*, H. KURAMOTO~ and E. GURPIDE* *Department of Obstetrics, Gynecology and Reproductive Science, Mount Sinai School of Medicine, New York, NY 10029, U.S.A. and tDepartment of Obstetrics and Gynecology, Kitasato University, Kanagawa-Ken, Japan
Summary-A human endometrial adenocarcinoma cell line (Ishikawa) has been found to be estrogen responsive. The growth stimulatory effects of estradiol (IO-’ M) could be clearly demonstrated when cell cultures containing the hormone were compared with the maximal cell density achieved in control cultures. The approx. 3-fold increase in cell density observed 2-3 weeks after plating, with frequent medium changes, could by blocked by a tOO-fold molar excess of the antiestrogen trans-~monohydroxy~moxifen. When added to hormone-free cultures that had reached a plateau level of cell numbers on day 14 after plating, estradiol (low8 M) caused the resumption of proliferation: after 6 days in the presence of the hormone, the cultures contained nearly twice the cell numbers of controls. Effects of estradiol on Ishikawa cells were also evident from the several-fold increases in the levels of specific progesterone binders provoked by the hormone at lO-q--lO-6 M concentrations. Cells injected into nude mice formed tumors which contained estrogen and progesterone binders. The availability of a fast-growing (doubling time approx. 30 h) endometrial cancer cell line responsive to estradiol at near physiologic levels will facilitate biochemical studies of hormonal effects on the human
endometrium.
Studies on the effects and mechanisms of action of hormones on human target tissues have been facilitated by the availability of hormone responsive cell lines, like MCF-7 cells derived from a mammary tumor [l-3]. Unfo~unately, none of the established endometrial cell lines have been clearly responsive to ovarian hormones at near physiologic concentrations, even though some of them are endowed with specific estrogen binders. Previous studies at these laboratories have focused on two human cell lines, HEC- 1 and HEC-50, derived poorly moderately or specimens of from differentiated endometrial adenocarcinomas. These cells did not respond to either estradiol (E,) and, perhaps as a consequence, to progesterone (P), as determined by evaluating cell proliferation, levels of P receptors and activities of ornithine decarboxylase, peroxidase, alkaline phosphatase or estradiol 17/1 dehydrogenase. We now report responsiveness to estrogens of a human endometrial adenocarcinoma cell line (Ishikawa). Addition of E, to the medium had a marked growth promoting effect and enhanced specific binding of P. In addition, E2 also stimulated alkaline phosphatase and DNA polymerase 01activities in Ishikawa cells, as shall be reported elsewhere. EXPERIMENTAL
Cells Ishikawa cells were established as a permanent cell line by Nishida et ul.[4] from cells derived from a $To whom correspondence should be addressed. 85
specimen of well-differentiated endometrial adenocarcinoma. At our laboratory, the cells were cultured in minimum essential medium (Eagle) with Earle’s salts (MEM, Grand Island Biological Co., Grand Island, N.Y.) supplemented with 15% fetal bovine serum (FBS, Grand Island Biological Co.). In experiments aimed to determine hormonal effects, the serum was pretreated with charcoal to remove steroid hormones, as follows: 100 ml FBS mixed with 0.25 g activated charcoal (Sigma Chemical Co., St. Louis, MO.) and 0.025 g dextran (clinical grade, Sigma) was stirred at 56°C for 30 min and centrifuged to separate the dextran-coated charcoal pellet from the supernatant, which was again subjected to the same treatment at 37°C. The charcoal-stripped serum (SFBS) was filtered through a 0.20pm sterilization unit (Nalge Co., Rochester, N.Y.) and stored at -20°C. Growth kinetics Confluent cultures of Ishikawa cells in MEM-15% SFBS were exposed to a trypsin (O.O5%~E~TA (0.02%) solution (Flow Laboratories, McClean, Va) for lO--15min at 37°C and the collected cells were seeded into 6 cm culture dishes (Falcon Plastics, Los Angeles, Calif.) at a density of 0.25 x IO6cells/dish in MEM-15% SFBS medium confining E, (lo-* M, Steraloids, Wilton, N.H.), trans-4-monohydroxytamoxifen (OHTam; 10s6 M, ICI, Macclesfield, U.K.), a mixture of E, (10e8 M) and OHTam (10m6M), or only ethanol (0.2% v/v) to reproduce the concentration of the vehicle used for the addition of steroids to the other dishes. Medium was renewed on day 3 and 6 after plating and subsequently every
C. F. HOLINKA et al.
86
other day. Cells were harvested by trypsin-EDTA treatment and counted in a Neubauer-Levy hemacytometer under an inverted microscope. Hormone binding Cells from confluent cultures were seeded into IOcm dishes in a 1:2.5 split ratio in MEM-15% SFBS. After 1 day, the medium was changed to MEM-15% SFBS containing E, at various concentrations. The final levels of ethanol in each dish, including controls, was 0.1%. Cells were harvested 3 days after hormone addition by trypsin-EDTA treatment. Trypsin action was stopped by addition of SFBS (7.5%) and cells were pelleted and homogenized in the assay buffer (50 mM Tris-HCI, 15 mM sodium phosphate, 20% glycerol, 0.01% sodium azide and 1 mM dithiothreitol, pH 7.8). Specific hormone binding was measured by the hydroxylapatite (HAP) method [5] as follows. Aliquots of the homogenate (180 ~1) containing 1.5-3.5 mg/ml total protein. measured by the method of Lowry et a1.[6], and 0.25-0.58 mg/ml DNA, measured by the method of Burton[7], were transferred to 7 ml glass scintillation counting vials (Kimble. Toledo, Ohio) and mixed with 20 ~1 of an ethanol-assay buffer solution of either [‘HIP (20 nM) and cortisol (2 PM), to determine total binding, or [3H]P (20 nM), cortisol (2 PM) and P (2 PM), to measure nonspecific binding. The final ethanol concentration was 1%. At the end of a 3 h incubation period on ice, 0.25 ml of a 70% suspension of buffer-washed HAP (Biorad, Richmond, Calif.) was added to the assay mixture. After further incubation for 10min on ice, a HAP plus nuclear pellet was separated by centrifugation and washed with 2 ml of the assay buffer containing 0.2% Triton X-100 (Sigma) and then twice with 2 ml of assay buffer. The mixture was vortexed and centri-
fuged at 1800g for 5 min after each washing. The pellets were then suspended in Dimiscint (National Diagnostics, Somerville, N.J.) and radioactivity was measured with a liquid scintillation counter (LS 8000, Beckman Instruments, Palo Alto, Calif.). The same assay buffer and experimental procedures were used to determine specific EZ binding with the following changes: homogenates were incubated at 30°C in [3H]E2 (20 nM) solutions to determine total binding, and in [3H]E2 (20 nM) plus E? (2pM) solutions to measure nonspecific binding. Tumor formation in nude mice Approximately IO’ Ishikawa cells were S.C. injected into the anterior dorsolateral region of nude mice (BALB/c, nu/nu, females). Tumors were harvested 3 weeks after inoculation, portions were sectioned and stained with hematoxylin-eosin and the rest of the tissue was used to measure E, and P binding levels. RESULTS
Figure 1A presents a photomicrograph of Ishikawa cells in culture. The multilayer characteristic of these cultures is evident at the crater-shaped area and in several other regions. Figure 1B shows a section of a tumor removed 11 weeks after inoculation of Ishikawa cells into a nude mouse. Measurement of specific binding of [3H]E2 and [3H]P in three such tumors showed the presence of receptors for estrogens (160, 1000 and 1200fmol/mg DNA) and progestins (250, 240 and 200 fmol/mg DNA). Figure 2 shows growth curves corresponding to parallel cultures of Ishikawa cells in MEM-15% SFBS, either alone or in the presence of E, (1 O-’ M). OHTam ( 10m6 M) or a mixture of these compounds. Control cultures reached maximal cell numbers at 14
Fig. 1A. Phase-contrast photomicrograph of a high-density culture of Ishikawa cells. This micrograph illustrates the overgrowth of cells into several layers, as evident at the crater-shaped (x 110).
photoregion
Effects of estradiol on Ishikawa cells
Fig. t B. Photomicrograph
of a tumor formed in a nude mouse after S.C.injection of 10 million lshikawa cells into the dorsotateraf area (x 340).
days. At that time, cultures
with E, (lWs tained twice as many cells as controls, numbers still increasing to over three times trol levels on day 18. Hydroxytamoxifen
M) conwith ceil the con-
(10m6M) had no effect on growth but counteracted markedly the actions af E?. Figure 3 indicates that E, caused the resumption of proliferation when added to high-density control cultures on day 14. at a time when the cuhures had 18 -
14
reached a plateau. After 6 days of hormone exposure these cells contained twice the cell numbers found in parallel control dishes. As apparent from both Figs 2 and 3, E2 affected primarily cell density and only to a minor extent growth rates. The doubling time in hormone-exposed and control cultures was similar (approx. 30 h). Figure 4 illustrates the effect of E, on the Ievets of specific binding of P when present in the medium for 3 days at concentrations ranging from 1W9 to 10-6M. Significant increases in binding relative to
Control
0
16 -
87
A E, . Ez+OH-TRM x OH-TAM
12 c a 2
lo-
p: fs g
a-
f 6-
4-
2-
0
3
6
8
10 12 14 16
18
Days in culture
Fig, 2. ~r~wt~-prorn~tin~ etkt of &radio1 in Isbikawa c&s and its reversal by OHTam. Each point represents the mean of three dishes. The SD was below IO% of the mean in ail cases.
Days
in culture
Fig. 3. Resumption of cell prolife~tion after addition of estradioi to hj~-density cultures at a plateau level. Each point represents the mean of three dishes. The SD was below IO% of the mean in all cases.
88
C.
0
lo-!+ 10-0
10-7
F. HOLINKA et al.
10-6
F23
Fig. 4. Enhancement of specific P binding by estradiol. The hormone was added in the indicated concentrations to confluent cultures and P binding was measured 3 days thereafter. Values represent the mean + SEM of three dishes, assayed in triphcate. (*P < 0.02 relative to control; Student’s f-test.)
controls (1.6- to 3-fold) were noted, with maximal increases at 10e8 M E2. Since the DNA/protein ratio
in the cells was not affected by E,, P binding patterns were not different when related to protein or DNA content. DISCUSSION
Responsiveness of Ishikawa cells to E, has been demonstrated in this study by evaluating growthpromoting effects and enhancement of specific P binding levels. The responses to E, indicate that the cells possess functional estrogen receptors; we found 33Ofmol/mg DNA specifically bound E, in cultures at the plateau phase 18 days after plating and 3 days after the last change of medium. However, we do not attempt to characterize Ishikawa cells by their recep tor content since previous experience with endometrial cells in monolayer made us aware of marked changes in these levels with time, which were found to reflect varying ratios of intracellular concentration of cGMP and cAMP[8]. In the present study, we observed a targe dispersion of specific P binding levels among samples of confluent cells from different cultures. However, parallel cultures of cells with identical chronology showed close agreement in P binding, as evident from triplicate values for each time point in the experiment described in Fig. 4. The present study clearly indicates that the E2 effects on growth of Ishikawa cells are mainly evident by comparing the maximal cell density achieved in control cultures with the density of parallel cultures in the presence of the hormone. Under these experimental culture conditions, using media containing 15% FBS, no obviously significant E2 effects on the doubling time during the exponential phase of proliferation were observed. The resumption of proliferation obtained by adding E to control cultures
with cell numbers stabilized at a plateau level indicates that the hormone is capable of overcoming density-de~ndent growth arrest. Whether Ez acts directly in promoting cell division or indirectly by enhancing autocrine growth factors or blocking the production of such growth factors remains to be examined. The availability of estrogen responsive (Ishikawa) and estrogen unresponsive (HEC- 1, HEC-50) endometrial aden~arcinoma cells, all containing specific estrogen binders [9], offers the possibility of studying factors affecting responsiveness to hormones. Furthermore, production of hormonally responsive tumors in the nude mouse makes possible the investigation of strategies for the regulation of human endometrial cancer growth in experimental animals. It is reassuring that results obtained with endometrial cancer cells in culture can be reproduced in fresh endometrial tissue, such as the regulation of estrogen binding by cyclic nucleotides [8]. Another interesting characteristic of ishikawa cells is their lack of responsiveness to P, as determined by evaluating the effect of P on estradiol 178 dehydrogenase activity or as an antagonist to the E, effects on cell numbers and alkaline phosphatase activity (unpublished). Since basal and E,-induced levels of P receptors are clearly present in Ishikawa cells, these cells may be useful in the search for explanations of unresponsiveness to progestins in endometrial cancers that possess P receptors. Acknowledgements-This
work was supported by Grants
CA 15648, awarded by the National Cancer Institute, and HD 07197, awarded by the National Institutes of Child Health and Human Development. REFERENCES 1. Berg C. D., Nawata H., Bronzert D. A. and Lippman
2.
3.
4.
5.
6.
M. E.: Altered estrogen and anti-estrogen responsiveness in clonal variants of human breast cancer cells. In Progress in Cancer Research and Therap) (Edited by F. Bresciani, R. J. B. King, M. E. Lippman, M. Namer and J.-P. Raynaud). Raven Press, New York, Vol. 31 (1984) pp. 161-170. Edwards D. P., Adams D. J. and McGuire W. L.: Estrogen regulation of growth and specific protein synthesis in human breast cancer cells in tissue culture. In Hormones and Cancer (Edited by W. W. Leavitt). Plenum Press, New York, Vol. 138 (1982) pp. 133~-149. Rochefort H., Chalbos D., Capony F., Garcia M., Veith F. and Westley B.: Effect of estrogen in breast cancer cells in culture: released proteins and control of cell proliferation. In Hormones and Cancer (Edited by E. Gurpide, R. Calandra, C. Levy and R. J. Soto). Liss, New York, Vol. 142 (1984) pp. 37-51. Nishida M., Kasahara K., Kaneko M. and Iwasaki H.: Establishment of a new human endometrial adenocarcinoma cell line, Ishikawa cells, containing estrogen and progesterone receptors. Acta obstet. gynaec. japonica 37 (1985) 1103-l 11I. (In Japanese.) Hoffman P. G., Jones L. A., Kuhn R. W. and Siiteri P. K.: Progesterone receptors: saturation analysis by a solid phase hydroxylapatite adsorption technique. Cancer 46 (1980) 2801-2804. Lowry 0. H., Rosenbrough N. J., Farr A. L. and
Effects of &radio1 Randall R. J.: Protein measurement with the Folin phenol reagent. J. biol. Chem. 193 (1951) 2655275. 7. Burton K.: A study of the conditions and mechanisms of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid. Biochem. J. 62 (1956) 315-323. 8. Fleming H.. Blumenthal R. and Gurpide E.: Rapid
on Ishikawa
cells
89
changes in specific estrogen binding elicited by cGMP and CAMP in cytosol from human endometrial cells. Proc. natn. Acad. Sci. U.S.A. 80 (1983) 2486-2490. 9. Fleming H., Blumenthal R. and Gurpide E.: Characteristics of cyclic nucleotide dependent regulation of cytoplasmic E2 binders in cultured endometrial and breast cells. J. sreroid Biochem. 20 (1984) 5-9.