Progestin stimulation of thymidine kinase in the human breast cancer cell line T74D

170

Bto~htmt~a et Btoph~ stca A~ta 1096 (1991) 170-174 • 1991 Elsewer Science Pubhshers B V 0925-4439/91/$03 50 A D O N I S 0925443991000647

BBADIS 61026

Progestin stimulation of thymidine kinase in the human breast cancer cell line T47D Michael R. Moore, Larry D Hathaway and James A Bircher Department of Bto~hemtstt3 Marshall Umterstt~ School of Medtcme Huntmgton 1411~(U S A )

(Recewed 16 May 1990) (Rewsed manuscript received 26 October 1990)

Key words Thymldme klnase, Progestm, Human breast cancer

Our laboratory has previously reported that progestlns stimulate growth of the human breast cancer cell line T47D In an attempt to probe further into thts stimulation, we are investigating progestin effects on thymtdlne klnase (EC 2 7.1.21), an enzyme known to be involved in growth regulation This report relates our finding that progestins stimulate thymidine kinase activity, at physiological progestin concentrations, in a dose-responsive manner Estradiol-1713 also stimulates, but testosterone, hydrocortisone and aldosterone do not. The antiprogestin RU486 inhibits progestin stimulation, but also stimulates on its own. Maximal by 24 h, the progestin stimulation then falls off with time Experiments with actinomycin D and cycloheximide suggest that the thymidine kinase stimulation depends on new RNA and protein synthesis. These data shed further light on progestln stimulation of the growth of human breast cancer. To our knowledge, this is the first report of progestin stimulation of thymidine kinase in human breast cancer cells

Introduction Several authors have recently reported progestln inhibition of growth of h u m a n breast cancer cells in long term tissue culture [1,2] These studies, however, were done in the presence of the p H indicator phenol red, which has recently been shown to contain estrogenic lmpunties [3,4] This leaves open the posslblhty that one reason for the progestins' growth inhibition was that they were acting as antlestrogens That is, it is possible that estrogenic impuntles in the phenol red were stimulating growth of the cells and the antlestrogenIc nature of the progestins acted to inhibit thas stimulation Still, the above reports may reflect the in VlVO situation to a large extent In m a n y breast cancer patients, there are circulating estrogens whose growth stlmulatory effects could be intublted by progestins However, there are also patients whose circulating estrogens have been at least parttally neutralized by the antIestrogen tamoxlfen Postmenopausal patients also have reduced estrogen levels In theory, this could leave circulating progestlns (either endogenous or exogenous) to act u p o n

Correspondence M R Moore, Department of Biochemistry, Marshall Umverslty School of Medicine, 1542 Spnng Valley Drive Huntington WV 25704 USA

the cells without or with reduced interaction with estrogens Our data have shown that in vitro progestlns alone stimulate the growth of h u m a n breast cancer cells rich in progesterone receptors [5 7] It is possible that this same situation, progestln stimulation of breast cancer growth, may occur m VlVO partially explaining why some estrogen receptor-positive, progesterone receptor-positive patients relapse during tamoxifen treatment In fact, R o b i n s o n and J o r d a n have shown that this does occur in the dimethylbenzanthracene-lnduced rat m a m m a r y carcinoma model [8] We have found that in the absence of phenol red progestms stimulate growth of wild type T 4 7 D cells from the A T C C in a dose-responsive manner in the physiological concentration range [7] Other authors have also recently reported progestln growth stimulation of breast cancer cells both in wvo and in vitro [8-13] Earlier in VlVO studies gave conflicting results, reporting both inhibition [14,15] and stimulation [16,17] Thymldlne klnase (EC 2 7 1 21) is an enzyme implicated in the regulation of growth, including that of h u m a n breast cancer cells [18], rat pituitary gland [19], developing rat cerebellum [20] and erythropoietic mouse spleen cells [21] L l p p m a n and his co-workers have documented tts increase in association with estrogen growth stimulation of the h u m a n breast cancer cell line M C F - 7 [18,22,23] In view of our finding of progestln

171 sttmulatlon of growth, we were mterested in finding whether thymldlne klnase activity was hkewtse affected Ttus study presents our mitial findings m that regard Materials and Methods

Cell culture T47D wild type cells, obtained from the American Type Culture Collection, were grown routinely wtth phenol red but for experiments in the absence of phenol red in 5% CO 2 in atr at 3 7 ° C Growth medium was basal m e d i u m Eagle modffted-autoclavable (powdered) plus non-essential a m m o actds, 2 mM Lglutamlne, 10% fetal bovine serum (heat inactwated), 100 u n I t s / m l penicillin, 1 0 0 / z g / m l streptomycin (Grand Island Btologlcal) and 6 n g / m l insulin (Sigma) Cells were harvested by replacing the growth medium with Hanks' balanced salt solution without calcium and magnesium, but with 1 m M EDTA, incubating for 10 nun at 37 ° C and centrifuging For experiments cells were plated in the above medium except that the fetal bowne serum had been treated wtth dextran-coated charcoal and there was no phenol red After 5 days in this charcoal-stripped serum-containmg medium, during which there was one medium change, hormones were added m absolute ethanol, control flasks receiving absolute ethanol only Cells were grown for the times indicated in the figure legends, m e d m m changed as indicated Extraction and assay of thymldme kmase Cell pellets were homogenized in 0 6 ml of PG buffer (5 mM sodium phosphate, 1 mM dithiothreltol, 10% glycerol ( v / v ) p H 7 4, 4 ° C) using a motor-driven, steel-shafted, teflon-tipped pestle and a comcal ground-glass handheld homogenizer (Kontes) The mixture was kept m ice except during the actual homogemzatlon (20 up-anddown strokes with 30 s cooling in ice between each series of 20 strokes for a total of 60 strokes) Nuclei, etc were then centrifuged out usmg an SW60Tt rotor (Beckman) m a Beckman L5-75 ultracentrifuge at 46 000 x g for 1 h at 2 ° C and the supernatant was kept as the enzyme extract Assay of thymidme kmase activity was done using a modification of the method of Breitman [24] The assay mixture contained 21 ~M [3H]thymldme (0 8 C i / m m o l ) , 0 005 M ATP, 0 005 M MgC12, 0 25 M Trls-HCI (pH 7 8) and the enzyme extract in a final volume of 0 11 ml Incubation was performed at 37 ° C for various time mtervals up to 30 nun and the reaction stopped by plpettlng 25 /LI onto a Whatman DE81 filter disc The disc was then tmmediately put into 1 m M ammonium formate, washed three times wtth ammonium formate and once with ethanol After drying, the disc was counted m Scintlverse II (Fisher) scintillatton cocktail at 44% efficiency Protein concentrations were determined by the method of Bradford [25] D N A was measured by the method of Burton [26] 1 unit of

thymtdme kmase 1s the amount of enzyme necessary to convert 1 nmol of thyrmdlne to thyrmdme monophosphate per minute Chermcals R5020 (17,21-dimethyl-19-nor-4,9-pregnadtene-3,20-dlone) was purchased from New England Nuclear RU486 (17fl-hydroxy-ll-fl (4-dimethylanunophenyl-1)-17 a-(prop-l-ynil)estra-4,9-dlen-3-one) was a generous gaft from Dr R Deraedt of RousselU C L A F (Romalnvllle, France) ICN was the source of [3H]thymldlne, (74 C1/mmol), while non-radioactive thyrmdine, ATP, steroid hormones, lactic acid, N A D +, cyclohexlmlde and actlnomycln D were from Sigma All other chemicals used were secured from standard sources and were of ACS reagent grade or better purity Results

The data of Table Ia demonstrate dose-responsiveness in stimulation of thymldine klnase by the synthetic progestm R5020 We used R5020 because of its stability in this system as compared to the naturally ocurrlng hormone progesterone, which is very rapidly metabohzed [1] Slmdar stimulatton occurred whether analyzed per mg D N A or per mg protein, as was the case in all other experiments tn this report Increases m thymldlne kmase activity occurred throughout the phystologlcal TABLE I

Concentration-dependence and hormonal specificity of progestm stimulation of thymldme kmase (a) Concentration dependence *

(b) Hormonal speclfloty **

- log[R5020] (M)

103 × units /mg protem

hormone

103 × umts /mg protem

Control 11 10 9 8 7 6

4 71 + 2 65 720+019 8 90 + 1 18 11 01 _+3 07 13 17 + 1 62 18 3 6 + 6 16 12 60_+4 13

control R5020 estradlol-17B progesterone hydrocortlsone testosterone aldosterone

9 14_+ 0 28 2 2 5 9 + 1 18 22 2 4 + 1 42 12 76 + 0 79 9 16 + 0 26 9 1 6 + 1 24 8 70_+0 44

After plating 7 5 105 cells per flask and pretreatment for 5 days as described in Materials and Methods duphcate T-75 flasks were treated with varying concentrations of hormone or with vehicle (0 1% ethanol) for 24 h Cells were then harvested and assayed as described m Materials and Methods Data refer to m e a n + S D for duphcate flasks and are representatwe of three separate experiments After plating and pretreatment as above, cells were treated with 10 -8 M hormone or ethanol for 24 h Final ethanol concentration was 0 1% in all flasks Results refer to mean_+S D for trlphcate flasks and are representative of three separate experiments Control. hydrocortlsone, testosterone and aldosterone are statistically the same as one another They are all statistically different from R5020, estradlol and progesterone R5020 and estradlol are the same but are both different from progesterone All differences are at P < 0 01 by the Student-Newman-Keuls multiple comparison procedure

172 500400 .

300

/

0 (o

"i, \\

/

200

\\

/

i00

0 0

2

4 DAY

Fig 1 Dependence of progestln stimulation of thynudme kmase on ume Cells were plated and pre-treated as m Table I, then treated w~th 10 -8 M R5020 or ethanol for varying periods of time Final ethanol concentration was 0 170 m all flasks There were tnphcate hormonetreated and control flasks at every t~me point Results refer to mean_+ S D ( u m t s / m g protein) for tnphcate flasks and are representative of three separate experiments All values were statistically different from their respective controls and from one another by the Student-Newman-Keuls multiple comparison procedure ( P < 0 01)

progestm concentration range, becormng opUmal at around 10 -7 M and appearing to fall off from ttus peak as the mlcromolar range was reached Growth, as measured by #g DNA per flask, parallelled thymldlne kmase actwlty (data not shown) Time dependency stu&es indicate that the phenomenon becomes maximal wltlun 24 h after addmon of progestm (Fig 1) Sttmulauon of thynudme klnase then falls off with time Hormonal speoficlty stu&es reveal, as shown m Table Ib, that the phenomenon has speclfloty for progestms and estrogens Progesterone ttself is less effective than the synthetic progestm R5020, probably because of the rapid half-hfe of progesterone as compared to the

stable R5020 [1] We have previously reported this same comparattve trend as regards growth sttmulatory effects of these two compounds [7] Estradlol and R5020 sumulate thymldme klnase to about the same extent after 1 day of treatment Th~s ~s also what we observed m our first report on progestm stimulation of growth It ts only after about 3 days of treatment that estra&oltreated cells begin to prohferate much faster than progestm-treated cells [5] The data of Table IIa show that progestin sumulatton of thymldme kmase is inhibited by the anuprogestm RU486 They also m&cate that RU486 alone stimulates thym~dme kmase Ttus ts consistent with our earher report [27] that RU486 alone can sttmulate growth Table IIb dlustrates the effects of the transcrtptton inhibitor actmomycm D which prevents progestm sumulat~on of thyrmdme kmase, suggestmg that the effect is dependent on new RNA synthests Three separate experiments showed that the protein synthests mhlbltor cyclohex~mlde hkewise prevents stimulation, suggesting a dependence on protem synthesas (data not shown) Discussion

Thyrmdme kmase converts thymldme to thym~dylate, and ts an enzyme of one of the salvage pathways of nucleoUde synthesis Found m essentially all growing cells [28], tts levels are sttmulated durmg condmons of increased growth rate The enzyme has been lmphcated m regulatton of growth of human breast cancer cells [18], rat pttmtary gland [19], developmg rat cerebellum [20] and erythropolettc mouse spleen cells [21] Bronzert et al [29] found that levels of estradtol whach maximally sttmulate [3H]thyn~dlne tncorporatlon mto DNA also

TABLE II

Effects of the annprogestm R U486 and of actmomycm D (b) Effects of actmomycm D * *

(a) Effects of RU 486 * treatment

103 × u m t s / m g protein

treatment

103 x u n l t s / m g protein

Control R5020 RU486 R5020 + RU486

6 22 12 9

control R5020 act D act D + R5020

8 19 + 0 47 22 06 + 1 64 9 2 7 + 0 30 9 0 6 + 0 73

80 +_1 27 17 + 1 70 13___1 96 2 3 + 0 52

* Cells were plated and pre-treated as m Table I Treatment was then carried out for 24 h with 10 -9 M R5020, 1 0 - l 0 M RU486, both of the above or ethanol Ethanol concentrauon was 0 2% m all flasks Data refer to mean + S D for t nphc a t e flasks and are representatwe of three separate experiments R5020 is statistically &fferent from all others at the P < 0 01 level,RU486 Is &fferent from [R5020+ RU486] at P < 0 05 and from all others at P < 0 01 by the Student-Newman-Keuls multiple comparison procedure ** Cells were plated and pre-treated as m Table I 1 h before addttlon of hormone, all flasks were changed to fresh medium A c t m o m y c m D dissolved m water was added to appropriate flasks at 1 ~ g / m l 1 hour later 10 -8 M R5020 m ethanol or ethanol alone was added to appropriate flasks Final ethanol concentrauon in all flasks was 0 17o After 24 more h of treatment, cells were harvested, extracted and assayed for thyrmdme klnase We have shown the above treatment to intubxt R N A synthesis more than 9570 and to have no affect on cell vmbdlty [33] Data refer to mean + S D for tnphcate flasks and are representauve of three separate experaments R is stausucally different from all others by the Student-Newman-Keuls multiple comparison procedure

173 stimulate thymadme kmase activity 1 5-4-fold Usmg MCF-7 cells whose growth was arrested and cell cycle synchromzed by lsoleucme starvation, Altken et al [22] found that thymldme klnase m R N A levels were mcreased by 24 h after rescue of these cells m m e d m m containing estrogen There were no changes m the levels of m R N A for two other growth-related gene products, dlhydrofolate reductase and pS2 m R N A More recently, Kasld et al [23] reported that the level of thymldlne kmase m R N A maximally increased 2-3-fold after 24 h of estradlol stimulation and decreased to below control levels m the presence of antlestrogens Tins parallelled changes m [3H]thymldme incorporation into D N A Using endogenous nuclear run-off transcription, they also demonstrated that m R N A levels were regulated by transcnpuonal control The data suggested that the thymldlne kmase gene may be one whose actlwty ~s dtrectly influenced by the activated estrogen receptor complex, and an attractive system for mvesugatlon of hormonal regulation of growth Bronzert et al [29] saw about a 2-fold increase m thynudme klnase specific activity (per mg protein) at o p a m a l estrad~ol concentrations, which were from 10-9 through 10 -7 M, after 24 h of treatment with estradlol17/3 Their experiments were done m MCF-7 cells m the presence of phenol red, before it was known that phenol red contained estrogenic lmpurmes [3,4] It Is possible that tf there had been no phenol red m the m e d m m there would have been a greater estrogen stimulation of thymldme kmase We, on the other hand, treated T47D cells, a hne with high levels of progesterone receptor, with the progestm R5020 Tins treatment resulted m a 2-4-fold stimulation of thyrmdlne kmase as measured per mg protem Our experiments were done m the absence of phenol red These are conditions which we have prewously shown result in growth stimulation by progestlns [5-7] Our dose response studies show that the effect occurs in the physiological range of progestm concentration These findings are s~rmlar to our earher report regarding progestm growth stimulation of T47D cells Llppman and has co-workers showed a similar dose-response effect of estrogen on thynudme kmase m MCF-7 cells [291 Our t~me dependency stttdles indicate that thyrmdme kmase stimulation peaks at about 24 h and then falls off w~th time Tins effect of a peak m specific actwlty followed by a dechne seems to parallel growth stlmulahon That ~s, the extent of growth stxmulaUon by progestms m these experiments decreased w~th time just as thyrmdme kmase stimulation d~d (data not shown) Reports on estrogen stimulation of thymldme kmase by Llppman and has co-workers [29] indicate that enzyme stimulation ~s maximal at 24 h, but they do not comment on what happens after tins

We have previously shown that progestms and estrogens stimulate growth of T47D cells [5-7] Thymldme kanase sUmulat~on seems to be speclflC for these two classes of steroid hormones m these cells, as neither testosterone, hydrocortlsone nor aldosterone sUmulated the enzyme These specificity data, along with mhabmon of progestm stimulation by the antlprogestm RU486, suggest that the progestm stimulation is occurring through the progesterone receptor We have earher shown that RU486 seems to exinblt not only antlprogestm effects on growth m these T47D cells, but also progestm-hke effects when used alone, sttmulatmg growth [27] These new data with thynudme kmase confirm and extend those findings in that RU486 both inhabits thyrmdme kanase stimulation by progestms and stimulates the enzyme when admamstered alone We have also earher shown that RU486, cons~stent with ~ts partml progestm agomst actw~ty, acts m an anuestrogenlc capacity as regards growth Prehrmnary experiments suggest that ~t also has ant~estrogenlc actwlty on thyrmdme kmase (data not shown) Kasld et al [23] recently reported that the human thyrnldme kmase g e n e m MCF-7 cells is transcriptionally regulated by estrogen and may be directly influenced by the activated estrogen receptor complex The results of our experiments w~th actmomycln D and cyclohexamlde suggest that progestm stimulation of thyrmdme kmase may also occur at the level of m R N A However, studies using a thyimdme kmase c D N A wdl be necessary to clarify the level and mechamsm of progestln regulation of thymldlne kmase These studies are now in progress m our laboratory The data m thas report are consistent with our earher reports of progestm stimulation of growth of human breast cancer They add a new enzyme to those thus far found to be stimulated by progestlns They are also consistent w~th the possibility that thymadme kmase stlmulanon may be revolved m the progestln growth sUmulatlon They lend further support to our previous argument [5-7,27) for a chmcal regimen using combined antlestrogen and antlprogestm therapy, as potentrolly better than anuestrogen alone Others have also recently supported this concept [30,31]

Acknowledgments We thank Dr R Deraedt of Roussel Uclaf for the generous gift of RU486 and Mrs Dolores Brumfield for preparation of the manuscript Tins study was supported by grants from the N I H (RR05870) and Huntington Chmcal Foundation

References 1 Horwltz, K B and Fnedenberg, G R (1985) Cancer Res 45, 167-173

174 2 Sutherland, R L, Hall, R E, Pang, G Y N Musgrove, E A and Clarke, C L (1988) Cancer Res 48, 5084-5091 3 Berthols, Y , Katzenellenbogen, J A and Katzenellenbogen B S (1986) Proc Natl Acad So USA 83 2496-2500 4 Bmdal, R D Carlson, K E , Katzenellenbogen, B S and Katzenellenbogen, J A (1988) J Steroid Biochem 31,287-293 5 Hlssom J R and Moore M R (1987)Biochem Blophys Res C o m m u n 145, 706-711 6 Moore, M R , Hagley, R D and Hlssom J R (1988) m Progress in Chmcal and Biological Research (Hanklns W D and Puett, D , eds ), Vol 262, pp 161-179, Lss, New York 7 Hlssom J R , Bowden, R T and Moore M R (1989) Endocrinology 125, 418-423 8 Robinson, S P and Jordan, V C (1987) Cancer Res 47, 5386-5390 9 Manta, A , Badger B Wright, C , Ahmed, S R and Dehmers, L M (1987) Cancer Res 47, 3066-3071 10 Manta, A , Wright C Badger, B, Demers, L and Bartholomew, M (1988) Cancer Res 48, 3058-3061 11 Braunsberg, H Coldham, N G , Leake R E and Wong, W (1987) Eur J Cancer C h n Oncol 23, 563 571 12 Longman, S M and Buehnng, G C (1987) Cancer 59, 281-287 13 Kass, R , Pandaens, R J Heuson, J C and Danguy H H (1986) J Natl Cancer Inst 77, 173-176 14 Kledzlk, G S Bradley, C J and Meltes, J (1974) Cancer Res 34, 2953-2956 15 Welsch, C W , Clemens J A and Meltes, J (1968) J Natl Cancer lnst 41,465-471 16 Huggms, C (1965) Cancer Res 25 1163 1167 17 Hugglns, C and Yang, N C (1962) Sctence 137, 257-262 18 Altken, S C and Llppman M E (1985) Cancer Res 45 1611-1620

19 Valotalre, Y LeGuellec, R Kercet, H , Guellaen, G and Duval J (1975)Mol Cell Endot.r 3, 117-127 20 Welchsel, M E (1974) Brain Res 78, 455 465 21 Roodman G D Hutton J J and Bollum, F J (1976)Blochlm B~ophvs Acta 425 478-491 22 Altken S C Llppman, M E, Kasld, A and Schoenberg D R (1985) Cancer Res 45, 2607-2615 23 Kasld, A Davldson, M E Gelman, E P and Llppman M E (1986) J Blol Chem 261 5562 5567 24 Breltman, T R (1963) Btochlm Blophys Acta 67 153 155 25 Bradford, M M (1976)Anal Blochem 72 248-254 26 Burton, D A (1956)Biochem J 62 315-323 27 Bowden, R T , Htssom, J R and Moore, M R (1989) Endocrinology 124 2642-2644 28 Bresmck E (1978)Methods Enzymol 51 360-365 29 Bronzert, D A , Monaco M E , Pmkus, L Altken, S and Llppman, M E (1981) Cancer Res 4 1 , 6 0 4 - 6 1 0 30 Maudelonde, T Romleu G Ulmann, A Pujol H Gremer, J Khalaf, S Ca',ahe, G and Rochefort, H (1987) m Hormonal Mampulatlon of Cancer Pept~des Grov, th Factors and Ne'0, (Antt)Steroldal Agents (Khjn, J G M , Pandaens R and Foekens J A e d s ) pp 55-59 Raven Press New York 31 Khln, J G M De Jong, F H Bakker G H Lamberts, S W J Rodenburg C J and Alexle,,a-Flgusch, J A (1989) ( a n c e r Res 49 2951-2956 32 Horwltz, K B Mockus M B Pike, A B, Fennessey P V and Sheridan R L (1983)J Blol Chem 258 7603 7610 33 Hagle~ R D Hlssom, J R and Moore M R (1987) Btochlm Blophys Acta 930, 167-172