Jan 12, 1984 - corticosteroid effects. RU-486 is shown to be a strong an- tagonist of the glucocorticosteroid-induced cytolytic response of the murine thymoma ...
The EMBO Journal vol.3 no.4 pp.751-755, 1984
DNA binding properties of glucocorticosteroid receptors bound the steroid antagonist RU-486
Suzanne Bourgeois*, Magnus Pfahl* and Etienne-Emile Baulieu1 Regulatory Biology Laboratory, The Salk Institute, San Diego, CA 92138, USA, and 'Lab-Hormones, INSERM, Faculte de MWdecine, 94-270 Bicetre, France *To whom reprint requests should be sent Communicated by G. Schutz
RU-486 is an anti-fertility steroid which also has anti-glucocorticosteroid effects. RU-486 is shown to be a strong antagonist of the glucocorticosteroid-induced cytolytic response of the murine thymoma lines W7TB and T1M,b, and of the induction of mouse mammary tumor virus (MMTV) mRNA in T1Mjb cells. The glucocorticosteroid receptor of W7 cells has high affinity for RU-486 (Kd = 3 x 10 - 9 M) but the complex formed has low nuclear transfer capacity. Binding of RU-486, as compared with the glucocorticosteroid agonist triamcinolone acetonide, to mouse receptor results in a decreased affinity for DNA in general and a reduced specific recognition of a site in the promoter region of MMTV proviral DNA. The RU-486 complex formed with rat liver receptor exhibits the same behavior; in addition, it is shown that only a fraction of these complexes are activated by temperature and these form highly salt-sensitive interactions with DNA. These results indicate that the binding of RU-486 to glucocorticosteroid receptors mimics pharmacologically the properties of a class of receptor variants (nt -) which are non-functional and have reduced nuclear transfer and altered DNA binding capacity. These results substantiate the importance of DNA binding in receptor function. Key words: DNA binding/glucocorticosteroid receptors/lymphocytolytic response/MMTV induction/steroid antagonist Introduction The effects of steroid hormones are mediated by intracellular receptors which, after binding of the hormone, undergo an 'activation' process and accumulate in the cell nucleus (reviewed in Baulieu et al., 1975). Recent evidence indicates that receptor-steroid complexes recognize specific DNA sequences (Payvar et al., 1981; Mulvihill et al., 1982; Compton et al., 1982; Govindan et al., 1982; Pfahl, 1982; Scheidereit et al., 1983; Pfahl et al., 1983). If interactions with DNA play a role in the regulatory functions of receptor-steroid complexes, binding of a steroid antagonist may alter their DNA binding properties. RU-486 is a new anti-fertility steroid which also displays anti-glucocorticosteroid activity (Herrmann et al., 1982; Philibert et al., 1981). Here we demonstrate the antiglucocorticosteroid effects of RU-486 in murine T-lymphoma cell lines, and characterize the DNA binding properties of the complexes formed by mouse and rat receptors with RU-486. The mouse thymoma lines W7TB and T1Mlb are glucocorticosteroid sensitive in that these cells lyse in the presence of hormone (Bourgeois and Newby, 1977; Mermod et al., 1981). ( IRL Press Limited, Oxford, England.
to
Both cell lines contain endogenous mouse mammary tumor virus (MMTV) proviral sequences, which are not expressed in the W7TB line while MMTV transcription is induced by glucocorticosteroids in the T1Mjb line (Mermod et al., 1983). Glucocorticosteroid-resistant variants have been isolated from both lines and have been shown to result from receptor defects (Mermod et al., 1983; Pfahl et al., 1978a). A collection of receptor-defective variants of the W7 line has been characterized and some have been shown to have a reduced nuclear transfer capacity (nt -), and altered sensitivity of the nuclear binding to ionic strength (Pfahl et al., 1978a). Nonfunctional nt - receptor-triamcinolone acetonide (TA) complexes have a reduced affinity for DNA in general as well as for specific sites in MMTV proviral DNA (Pfahl, 1982). Here we show that binding of RU-486 to glucocorticosteroid receptors mimics pharmacologically the properties of nt - receptor variants.
Results Anti-glucocorticosteroid effect of RU-486 on the cytolytic response of thymoma cell lines Both the murine T-lymphoid cell lines W7TB and T1Mlb undergo cytolysis in the presence of glucocorticosteroids (Bourgeois and Newby, 1977; Mermod et al., 1981). As illustrated in Figure IA, in the presence of 10-8 M dexamethasone a culture of W7TB cells reaches only 1 - 2o of the turbidity observed in a control without hormone. This represents dead cells and cell debris. The known anti-glucocorticosteroids cortexolone and progesterone (Rousseau et al., 1972) act as partial agonists in this response, i.e., they cause cell killing in the range of concentrations of 5 x 10-8 M to 10-5 M. RU-486 has no glucocorticosteroid activity on W7TB cells up to a concentration of 10-6 M and only minimal effect at 5 x 10-6 M, a concentration at which the culture reached 7007o of the control without hormone. This could reflect a weak glucocorticosteroid effect, or the presence of an agonist impurity present at ) non-specific DNA.
0) 0
E'a
E _E
C
z
z
0
0
0
0 0 H
0~
LLJ
Lu
0 LLJ 11
LuI
RECEPTOR ADDED
(o)N
RECEPTOR ADDED (pmoles)
S
(M)
IONIC STRENGTH (M)
Fig. 4. Characterization of the DNA binding of partially purified receptor from rat liver. For these experiments receptor extract from rat liver cytosol was passed over phosphocellulose prior to incubation with [3H]RU-486 or [3H]TA, to remove other DNA-binding proteins. The preparation contained 4.0 pmol of receptor-TA complex of 7.6 pmol of receptor-RU-486 complex per ml, total protein concentration was 11 mg/ml. (A) Increasing amounts of R-TA or R-RU-486 were incubated with a constant amount of DNA-cellulose (5 mg/assay tube). Other experimental details are as in Figure 3. (B) 40 ,Ld of R-RU-486 or R-TA labeled extracts were incubated with various amounts DNA-cellulose. Corresponding experimental points in A and B are circled (equal amounts of receptor and DNA-cellulose/assay tube). (C) The standard ionic strength of the binding buffer is 0.07 M. To obtain lower molarities the buffer was diluted with 107o glycerol in H20; to obtain higher molarities, desired amounts of a 2 M NaCI solution were added.
labeled receptor-steroid complex to the DNA-cellulose is observed. We have used this DNA-cellulose competition assay with both R-TA and R-RU-486. Two competing DNAs were used, both obtained from cloned MMTV proviral DNA (Hynes et al., 1981). First, a purified 550-bp fragment containing a high affinity glucocorticosteroid receptor binding
site in the promoter region of MMTV DNA (Pfahl, 1982), and second a pBR322 DNA fragment of similar size which was shown not to contain such a specific binding site (Pfahl, 1982). Results for the R-TA complex are shown in Figure 3B. The R-TA complex has a >3-fold increased affinity for the DNA fragment containing the specific binding site, compared 753
S. Bourgeois, M. Pfahl and E.-E. Baulieu
with the non-specific DNA fragment. The competition experiments were carried out under conditions in which the competing DNA fragments were present at molar excess over the receptor complexes in the reaction mixture. Under these conditions each base pair can start a binding site; if this site is small ( - 25 bp) relative to the 550 bp length of the fragment, then this fragment contains 500 potential binding sites, taking into account some end effects. The observed 3-fold increased affinity, then, represents binding in the presence of an 500-fold excess of non-specific binding sites. The apparent 3-fold increase in affinity could therefore actually correspond to an 1500- to 2000-fold higher affinity for that specific site (for further discussion of this calculation see Pfahl, 1982). The R-RU-486 complex, when tested in this system (Figure 3C), showed only an 1.7-fold increase in affinity for the 550-bp fragment over the non-specific fragment. DNA binding properties of the receptor-RU-486 complex from rat liver Since the rat liver glucocorticoid receptor has been extensively purified (Wrange et al., 1979; Govindan and Manz, 1980), characterized (Carlstedt-Duke et al., 1982) and used in DNA binding studies (Payvar et al., 1981; Govindan et al., 1982; Pfahl, 1982; Scheidereit et al., 1983), we examined the effect of RU-486 on the DNA binding properties of this receptor as well. The results are virtually identical with the results obtained with the W7 mouse lymphoid cell receptor (data not shown). To obtain further information on the DNA binding properties of the R-RU-486 complex in comparison with the R-TA complex, partially purified complexes were prepared as described in Materials and methods. Binding curves using these extracts are shown in Figure 4A. In these experiments binding curves are linear for all receptor concentrations measured, since other competing binding proteins have been removed from the receptor extracts. As observed earlier, the R-RU-486 complex binds less well to the DNA-cellulose than the R-TA complex. To determine whether the reduced binding of the R-RU-486 was due to a low percentage of activated R-RU-486 molecules in the extract or to a reduced affinity of the R-RU-486 for DNA, we incubated a constant amount of receptor-steroid complex with varying amounts of DNA-cellulose. Results are shown in Figure 4B. We observe a constant amount of R-RU-486 bound when the DNA-cellulose concentration is changed over an 8-fold range, whereas decreasing amounts of R-TA are bound at DNA-cellulose concentrations below 5 mg (the standard assay concentration used in the binding curve, Figure 4A). Receptor and DNA-cellulose concentrations that are equal in the experiments shown in Figure 4A and B are indicated by circles around the experimental points. The results obtained by varying the DNA-cellulose concentration clearly point out that fewer of the R-RU-486 complexes are activated than of the R-TA complexes in the same receptor extract. To establish whether the R-RU-486 complexes able to bind DNA had the same affinity for DNA-cellulose as the activated R-TA complexes, the salt sensitivity of the binding of both complexes to DNA-cellulose was investigated. Experimental results are shown in Figure 4C. The R-TA complex shows an optimal affinity for DNA-cellulose at 0.07 M salt (as reported previously by others, see Wrange et al., 1979), whereas the R-RU-486 complex is very sensitive to salt: reduced binding of R-RU-486 to DNA-cellulose is observed at ionic strengths > 0.05 M. 754 -
-
-
-
Discussion The anti-progesterone RU-486 has a strong anti-glucocorticosteroid effect in that it could be shown to be an antagonist of both dexamethasone and TA in the receptor-dependent cytolytic responses of thymoma cells and in the induction of MMTV mRNA in the T1M1b cell line. The affinity of the receptor for RU-486 is 3- to 4-fold higher than for dexamethasone, but only 400/o of the complexes formed with RU-486 are able to translocate to the nucleus (Table I). As in the case of receptor-dexamethasone complexes from ntvariants, the decreased nuclear transfer capacity observed reflects changes in receptor properties which most likely account for the anti-glucocorticosteroid effect of RU-486. Our study presents evidence for changes in two properties in receptor-RU-486 complexes as compared with receptor-TA complexes: decreased capacity to be activated by temperature (Figure 4B) and altered interaction with DNA (Figures 3 and 4). Our study shows an increased sensitivity to salt of the R-RU-486 complex binding to DNA (Figure 4C). This demonstrates that the interaction of R-RU-486 complexes with DNA sites is different from that of R-TA complexes, and suggests a decreased affinity for non-specific DNA as well as for specific DNA. Taking this into account, the decrease in affinity of R-RU-486 for specific DNA is much larger than apparent from the results in Figure 3B and C. As far as the reduced capacity of R-RU-486 complexes to be activated is concerned (Figure 4B), comparable experiments with nt- receptors have not yet been carried out, but one can probably expect that at least some nt- receptors might have that property. RU-486 does not appear to reduce the stability of the receptor and the complex formed is, therefore, definitely different from 'activation-labile' glucocorticosteroid receptor observed in a steroid-resistant variant of the CEM-C7 human lymphoid line (Schmidt et al., 1980). In summary, it can be said that the glucocorticosteroid receptor-RU-486 complex behaves, in some aspects, like an, nt- receptor-steroid complex which, under physiologicall conditions, is incapable of interacting properly with DNA sites to cause a physiological response. A similar defect of accumulation in the nucleus has recently been observed in the case of the estrogen receptor complex formed with the antiestrogen compound tamoxifen in the chick oviduct system (Lebeau et al., 1982). In prokaryotic systems, non-specific DNA binding plays a role in the association of regulatory proteins with their specific target sites (reviewed in Bourgeois and Pfahl, 1976). In eukaryotic cells, non-specific DNA binding may be regarded as a component of an 'acceptor' system favoring the nuclear accumulation of receptor-steroid complexes. The present results also suggest a role for both specifi'c and non-specific interactions of receptor with DNA in the control of specific gene expression by steroid hormones. Materials and methods Cell lines and growth W7TB is a BrdUrd-resistant derivative of the BALB/c mouse thymoma line WEHI-7 (Bourgeois and Newby, 1977). T1M1b is a subclone derived from the C57BL/C mouse thymoma line TIM14G-1.3 (Mermod et al., 1981). Cells were grown in suspension as described elsewhere (Bourgeois and Newby, 1977). To measure the cytolytic effect of hormones, the cellular material present in culture was monitored by turbidity at 660 nm after five cell doublings in the control without steroid (Bourgeois and Newby, 1977). The results are expressed as a petcentage of the optical density reached in the control without steroid.
DNA binding of glucocorticosteroid receptors bound to a steroid antagonist Steroid binding assays The specific binding of [3H]dexamethasone (sp. act. 42 Ci/mmol; Amersham) or [3H]RU-486 (sp. act. 37.5 Ci/mmol; Roussel-Uclaf) to whole cells, and the nuclear transfer of receptor-steroid complexes were measured as described elsewhere (Pfahl et al., 1978b). Isolation and assay of MMTV RNA RNA was purified as described by Chirgwin et al. (1979), and poly(A) + RNA was selected on an oligo(dT)-cellulose column. Slot-blots onto nitrocellulose were prepared as described by Palmiter et al. (1982) and probed with a mixture of nick-translated MMTV sequences. The MMTV fragments used were a 1.8-kb Pstl fragment containing the 3' end of the env gene and a 0.9-kb Pstl fragment of the LTR region (Pfahl, 1982). The specific activities after nicktranslation were 7.5 x 106 c.p.m.4tg of env fragment and 107 c.p.m./4g of LTR fragment. Hybridization was carried out in standard saline-phosphateEDTA buffer containing 100o dextran sulfate. After autoradiography, slot intensities were estimated by densitometer scanning and integration of the peak areas. DNA-cellulose binding assay The DNA-cellulose binding assay and the preparation of activated steroidreceptor complex have been described (Pfahl, 1982). Freshly prepared extract (containing 15 mg of protein/ml) was incubated for 30 min at 280C with [3H]TA (sp. act: 26 Ci/mmol; Amersham) at a final concentration of 4 x 10-8 M or [3H]RU486 (sp. act. 37.5 Ci/mmol) at a final concentration of 4 x 10-8 M. The extract was then cooled to 00C in an ice bucket and the unbound hormone was removed with activated charcoal (Norit A charcoal:dextran 500 = 9:1) using 12 mg of charcoal per ml of extract. In some experiments, rat liver receptor was partially purified from other DNA binding proteins, by passing the cytosol receptor extract over a phosphocellulose column prior to incubation with labeled steroid. In the standard assay, 3.75 mg DNA-cellulose are incubated with labeled receptor extract for 30 min at room temperature (with vortexing every 5 min) in 20 mM phosphate buffer pH 6.9 containing 50 mM NaCl, 1007o glycerol (w/v), and 2 mM 2-hydroxyethylmercaptan. After centrifugation, the amount of receptor bound to DNA cellulose is determined by measuring the amount of radioactivity associated with the DNA-cellulose pellet. When desired, the ionic strength of the incubation buffer was altered by diluting with H20 or by adding NaCl.
Palmiter,R.D., Brinster,R.L., Hammer,R.E., Trumbauer,M.E., Rosenfeld, M.G., Birnberg,N.C. and Evans,R.M. (1982) Nature, 300, 611-615. Payvar,F., Wrange,O., Carlstedt-Duke,J., Okret,S., Gustafsson,J.A. and Yamamoto,K. (1981) Proc. Natl. Acad. Sci. USA, 78, 6628-6632. Payvar,F., Firestone,G.L., Ross,S.R., Chandler,V.L., Wrange,O., CarlstedtDuke,J., Gustafsson,J.-A. and Yamamoto,K.R. (1982) J. Cell. Biochem., 19, 241-247. Pfahl,M. (1982) Cell, 31, 475-482. Pfahl,M., Kelleher,R.J. and Bourgeois,S. (1978a) Mol. Cell. Endocrinol., 10, 193-207. Pfahl,M., Sandros,T. and Bourgeois,S. (1978b) Mol. Cell. Endocrinol., 10, 175-191. Pfahl,M., McGinnis,D., Hendricks,M., Groner,B. and Hynes,N.E. (1983) Science (Wash.), 222, 1341-1343. Philibert,D., DeReadt,R. and Teutsch,G. (1981) 8th International Congress of Pharmacology, Tokyo, Abstract No. 1463. Rousseau,G.G., Baxter,J.D. and Tomkins,G.M. (1972) J. MoL Biol., 67, 99115. Scheidereit,C., Geisse,S., Westphal,H.M. and Beato,M. (1983) Nature, 304, 749-752. Schmidt,T.J., Harmon,J.M. and Thompson,E.B. (1980) Nature, 286, 507510. Wrange,O., Carlstedt-Duke,J. and Gustafsson,J.A. (1979) J. Biol. Chem., 254, 9284-9290. Received on 7 December 1983; revised on 12 January 1984
Acknowledgements We thank Roussel-Uclaf for providing us with RU-486 and its tritiated derivative, and Thomas Ryden and David McGinnis for their excellent technical help. E.-E.B. is grateful to Dr. Roger Guillemin for his hospitality in the Neuroendocrinology Laboratory during the course of this work. This work was supported by NIH grant GM20868 and CA36146, NSF grant PCM8104184, and by a grant from the Whitehall Foundation.
References Baulieu,E.E., Atger,M., Best-Belpomme,M., Corvol,P., Courvalin,J.-C., Mester,J., Milgrom,E., Robel,P., Rochefort,H. and DeCatalogne,D. (1975) Vitamins Hormones, 33, 649-736. Bourgeois,S. and Pfahl,M. (1976) Adv. Protein Chem., 30, 1-99. Bourgeois,S. and Newby,R. (1977) Cell, 11, 423-430. Carlstedt-Duke,J, Okret,S., Wrange,O. and Gustafsson,J.A. (1982) Proc. Natl. Acad. Sci. USA, 79, 4260-4264. Chirgwin,J.M., Przybyla,A.E., MacDonald,R.J. and Rutter,W.T. (1979) Biochemistry (Wash.), 18, 5294-5299. Compton,J.G., Schrader,W.T. and O'Malley,B.W. (1982) Biochem. Biophys. Res. Commun., 105, 96-104. Govindan,M.V. and Manz,B. (1980) Eur. J. Biochem., 108, 47-53. Govindan,M.V., Spiess,E. and Majors,J. (1982) Proc. Natl. Acad. Sci. USA, 79, 5157-5161. Groner,B., Hynes,N.E., Rahmsdorf,U. and Ponta,H. (1983) Nucleic Acids Res., 11, 4713-4725. Herrmann,W., Wyss,R., Riondel,A., Philibert,D., Teutsch,G., Sakiz,E. and Baulieu,E.E. (1982) CR. Hebd. Seances Acad. Sci., Paris, Serie III, 294, 933-938. Hynes,N.E., Kennedy,N., Rahmsdorf,U. and Groner,B. (1981) Proc. NatI. Acad. Sci. USA, 78, 2038-2042. Lebeau,M.-C., Massol,N. and Baulieu,E.-E. (1982) Biochem. J., 204, 653662. Mermod,J.J., lntriere,L., MacLeod,C. and Bourgeois,S. (1981) J. Steroid Biochem., 15, 25-34. Mermod,J.J., Bourgeois,S., Defer,N. and Crepin,M. (1983) Proc. Natl. Acad. Sci. USA, 80, 110-114. Mulvihill,E.R., LePennec,J.-P. and Chambon,P. (1982) Cell, 24, 621-632.
755
