- Email: [email protected]
Androgen receptor-binding regions of an androgen-responsive gene
259
Molecular and Cellular Endocrinology, 51 (1987) 259-265 Elsevier Scientific Publishers Ireland, Ltd.
MCE 01672
Androgen
receptor-binding
regions of an androgen-responsive
N.K. Rushmere,
M.G. Parker
1
gene
and P. Davies
Tenoous Institute for Cancer Research, University of Wales College of Medicine, Heath Park, Cardiff CF4 4Xx, and ’ Imperral Cancer Research Fund Laboratories, P.O. Box 123, Lincoln’s Inn Fields, London WCZA 3PX, U.K. (Received
Key words: Steroid-binding
protein
15 January
C3(1) gene; (Prostate);
1987; accepted
12 February
Sa-Dihydrotestosterone-receptor
1987)
complex
Summary The interaction of 5a-dihydrotestosterone-receptor complexes with purified DNA fragments representing upstream, coding and intervening sequences of the prostate binding protein C3(1) gene was investigated using a DNA-cellulose competition binding assay. The partially purified androgen-receptor complexes which were used in the assay had proven DNA-binding capabilities. Two fragments were identified with relatively high affinity for androgen-receptor complexes. A 300 bp fragment extending from -220 to +80 and a 500 bp fragment derived entirely from the first intron consistently competed most effectively in the system. The presence of a high affinity site or sites in or near the promoter region of the gene is consistent with current models of transcriptional activation of hormone-responsive genes by steroid receptors. High affinity sites for steroid receptors within introns may indicate a role for receptors in regulation of transcription at other stages, or in post-transcriptional modification.
Introduction Steroid hormone receptor proteins are selective regulators of gene transcription (see Yamamoto, 1985). Their exact mechanism remains unknown but centres on the fact that binding of the cognate steroid hormone induces a receptor conformation appropriate to selective interaction with intranuclear sites, presumably specific DNA regions in the vicinity of responsive genes (Parker, 1983; Yamamoto, 1985). This assumption has been vindicated by the use of purified receptors and cloned hormone-responsive genes (Payvar et al., 1981, 1983; Geisse et al., 1982; Govindan et al.,
Address for correspondence: Dr. Peter Davies, Tenovus Institute for Cancer Research, University of Wales College of Medicine, Heath, Cardiff CF4 4Xx, U.K. 0303-7207/87/$03.50
0 1987 Elsevier Scientific
Publishers
Ireland,
1982; Mulvihill et al., 1982; Compton et al., 1983: Scheidereit et al., 1983). Regions of DNA with high affinity for receptors have been located upstream of and within genes (Payvar et al., 1983; Moore et al., 1985; Slater et al., 1985). An essential function for these DNA segments in receptormodulated transcription was proven by transfer into recipient cells of genes containing or lacking the sequences (Chandler et al., 1983; Dean et al., 1983; Hynes et al., 1983; Karin et al., 1984; Renkawitz et al., 1984). The conclusion to be drawn was that the receptor associated with specific promoter elements near or inside the affected genes (see Yamamoto, 1985). Sequences with high affinity for receptors for glucocorticoids (Payvar et al., 1981, 1983; Govindan et al., 1982; Pfahl, 1982; Scheidereit et al., 1983; Karin et al., 1984), progesterone (Mulvihill et al., 1982; Bailly et al., 1983; Dean et al., 1983; Ltd
260
Renkawitz et al., 1984) and oestrogens (Jost et al., 1984; Maurer, 1985; Weisz et al., 1986) have been described. There is, however, little information regarding interactions of androgen receptors and regions of androgen-responsive genes. The major secretory product of the rat ventral prostate, prostate steroid-binding protein, PBP (Heyns and De Moor, 1977) is an oligomer containing Cl, C2 and C3 polypeptides (Heyns et al., 1978), whose rates of synthesis are stimulated markedly by androgens (Parker et al., 1978). The Cl and C2 polypeptides are encoded by unique genes (Parker et al., 1982) but there are two nonallelic C3 genes, C3(1) and C3(2) (Parker et al., 1983). Only the C3(1) gene is responsible for the production of the C3 polypeptide in the rat ventral prostate (Hurst and Parker, 1983). Earlier attempts to define regions around this gene with high affinity for androgen receptors have met with varying success (Mulder et al., 1974; Perry et al., 1985). The fact that some were unsuccessful was due probably to deficiencies in the DNA-binding capabilities of the receptor which was used (Mulder et al., 1984). This report relates investigations into receptor binding in the vicinity of the C3(1) gene using a partially purified androgen receptor with proven affinity for DNA (Davies et al., 1986). Materials and methods [1,2,4,5,6,7-‘HISa-Dihydrotestosterone (110 Ci/mmol) was provided by Amersham International PLC (Amersham, Bucks, U.K.), who also supplied the following enzymes: restriction endonucleases EcoRI, Hind11 and PuuII; T4 DNA polymerase, T4 DNA ligase, and also synthetic oligonucleotide sequences containing EcoRI cleavage sites (EcoRI linkers). Nensorb 20 nucleic acid purification cartridges were bought from NEN Research Products (Du Pont (U.K.), Stevenage, Herts, U.K.). The Boehringer Corporation (London) (Lewes, U.K.) supplied Sl nuclease and restriction endonucleases Pst I, BglII and BstEII. All other chemicals were of the highest quality commercially available. Preparation
of androgen
receptor protein
The source of androgen receptor prostate cytosol from rats castrated
was ventral18 h earlier,
for purposes of maximum yield (Davies and Thomas, 1983). Receptor in complex with [ 3H]DHT was partially purified (approx. IOOOfold) by chromatography on Cibacron F3GAagarose and heparin-Sepharose, followed by ammonium sulphate precipitation and desalting on Sephadex G25 (Davies and Thomas, 1983). Further purification resulted in poor reproducibility in the DNA-cellulose competition assay (see below), due mainly to problems of unsaturability (Davies and Thomas, 1983). Androgen receptor protein purified to the stage described above bound to both DNA and chromatin, the latter tissue-specifically, and lacked deoxyribonuclease activity as judged by its inability to alter the electrophoretic mobility of supercoiled DNA. The quantity of [ 3H]DHT-receptor complexes present in a sample was estimated by a combination of protamine sulphate precipitation and sedimentation analysis. The latter revealed the androgenreceptor complex to have a sedimentation coefficient of 4.2 + 0.2s. Under conditions of scintillation spectrometry employed (29% efficiency for 3H), 1 pmol r ec ep tor was represented by 84 x lo3 cpm assuming 1 mol receptor bound 1 mol steroid. Preparation
of cloned DNA fragments
The genomic region investigated comprised the first exon of the PBP C3(1) gene, 684 bp of 5’-flanking sequence, the whole of the first intervening sequence, and a portion of the second exon, in all 2.5 kbp. Restriction fragments derived from this sequence are shown in Fig. 1 and denoted by Roman numerals. Briefly, they are: I. (EcoRI-EcoRI), the entire sequence as cleaved from the plasmid; II. (EcoRI-BstEII) 766 bp, -685 to + 80; III. (BstEII-EcoRI) 1735 bp; IV. (EcoRI-HindII) 465 bp, - 685 to - 220; V. (HindII-BstEII) 300 bp, - 200 to + 80; VI. (BstEII-BglII) 935 bp; VII. (BglII-EcoRI) 800 bp; VIII. (BglII-PuuII) 500 bp; IX. (EcoRI-BglII) 1.7 kbp; X. (HindII-EcoRI) 2.1 kbp; XI. (HindII-BglII) 1.24 kbp. The positions of these fragments in relation to exons, the intron and upstream sequence can be obtained by reference to Fig. 1. The constitution
261 x!
Fig. 1. The segment of the prostate steroid-binding protein C3(1) gene (I) which was released from p62A by cleavage with EcoRI (E) was divided into smaller fragments (II to XI) using combinations of the following restriction endonucleases: B, BgIII; BE, BsrEIl; H, HindII; P, PuuII. The various fragments are designated by the Roman numeral by which they are referred in the text. The fragments (II to VIII) arranged below fragment I in the figure are those obtained by the digestion route most frequently taken; those above (IX to XI) were derived infrequently.
of the more relevant fragments will be discussed in the text. For comparison, pBR322 was digested with EcoRI and PstI, to generate fragments of 0.75 kbp and 3.6 kbp. Recombinant plasmids were purified from clear lysates on caesium chloride gradients. Fragments from restriction enzyme-cleaved plasmids were analysed by electrophoresis on 1% (w/v) agarose gels and identified in relation to molecular size markers by staining with ethidium bromide and illuminating at 300 nm. Gel slices corresponding to required fragments were excised and incubated (60 min at 37°C) in saturated sodium iodide (2 ml/g wet gel). Extracts were mixed with an equal volume of buffer A (100 mM Tris, 1 mM Tris, 1 mM EDTA, 10 mM triethylamine, pH 7.7) and applied to a pre-wetted (100% methanol) and primed (1 ml buffer A) Nensorb column. Columns were washed with 3 ml buffer A and 5-10 ml H,O, and DNA was eluted with 500 IJ.I 50% (v/v) aqueous methanol. DNA was evaporated to dryness, resuspended in 10 mM Tris-HCl, pH 7.6, containing 1 mM EDTA (buffer B), and extracted 3 times with an equal volume of water-saturated isoamyl alcohol. Residual isoamyl alcohol was removed with water-saturated diethyl ether, ether was evaporated under nitrogen, and DNA was
ethanol-precipitated. The concentration of DNA resuspended in buffer B was determined as described by Gurney and Gurney (1984). Cohesive termini were eliminated by treatment with Sl nuclease or T4 DNA polymerase. When necessary, blunted fragments were ligated to EcoRI linkers using T4 DNA ligase and subcloned into the EcoRI site of pBR322. Soluble calf thymus DNA or rat ventral prostate DNA used in competition studies was sheared to an average size of 600 bp by sonication, treated with Sl nuclease, and purified as above. DNA-cellulose competition assay Calf thymus DNA was immobilized on cellulose by the method of Alberts and Herrick (1971). Two approaches to competitive binding assays were tested: either [ 3H]DHT-receptor was bound to DNA-cellulose prior to addition of competing fragments, or fragments were mixed with DNA-cellulose prior to addition of receptor. After ensuring the methods gave similar results, the former was used more frequently. Typically, DNA-cellulose was pelleted and resuspended in 1 mM Na,PO,, pH 7.0, containing 0.1 mM EDTA, 0.25 mM dithiothreitol and 50 mM NaCl, and [3H]DHT-receptor (l-2 X 10’ cpm/ml). After incubation at 0-4°C for 60 min, aliquots (200 ~1, equivalent to l-3 pg calf thymus DNA) were pipetted into tubes containing increasing quantities of soluble competitor DNA to be tested (2-10 ~1). Assays were usually in quadruplicate. After further incubation for 30 min at 25’C, samples were chilled, diluted to 1 ml in ice-cold buffer, centrifuged at 5000 X g for 10 min, pellets were washed briefly in ice-cold buffer, repelleted, extracted twice in absolute ethanol, extracts were combined and incorporated radioactivity was assessed. Results The characteristics of retention of [ 3H]DHT-receptor complexes by calf thymus DNA immobilized on cellulose are shown in Fig. 2. Routine assay conditions were derived from these experimental data. As indicated in Fig. 2a and b, under the conditions used and with this receptor preparation, binding was saturable. Maximum binding
262
.
.
10 20 30 LO
60 Time ImInI
120
I
I
I
2 3 L 5 l:put PHIDHT-R(dpm ~10-~1
6
(c)
’ 25
I
b
I
1
#
1
35
L5
55
65
75
85
&
!
95 100
NoCI (mmol/L)
Fig. 2. (a) DNA-cellulose was resuspended in 1 mM Na,PO,, pH 7.0, containing 0.1 mM EDTA, 0.25 mM dithiothreitol and 50 mM NaCI, and [3H]dihydrotestosterone (DHT)-receptor complex (2 x lo5 cpm/ml), in the absence (0) and presence (0) of unlabelled [ 3H]DHT-receptor complex (250 nM). Samples were incubated at 0-4OC and, at various times, aliquots equivalent to 3 ng of calf thymus DNA were removed, washed twice with the same buffer, extracted twice in absolute ethanol, and the radioactivity contained in combined extracts assessed by liquid scintillation spectrometry. (b) DNA cellulose was resuspended in buffer as in (o) but containing varying con-
occurred within 30 min and was stable for at least 2 h at 0-4°C (Fig. 2a). Labelling of receptor with [3H]DHT in the presence of a 100-fold molar excess of nonradioactive DHT significantly reduced the radioactivity subsequently associated with DNA-cellulose. Nonspecific binding was lo-15% of total. Saturability of the system may have been due to the relatively impure receptor preparation containing other DNA-binding proteins (Chamness and McGuire, 1972; see Mulvihill et al., 1982); certainly, increased purification of the androgen receptor abrogates saturation of DNA-binding sites (Davies and Thomas, 1983). However, as chromatography on immobilized triazynyl dyes achieved significant removal of proteins with affinity for DNA, some other factors may be involved. The effects demonstrated in Fig. 2 were consistent throughout the experiments reported. An input of 2.5-4 X lo4 cpm of [3H]DHT-receptor complex per l-3 pg DNA in 200 ~1 resulted in binding of 19.4 + 4.2% (mean + SD) of total. Preformed complexes were most stable at 50-100 mM NaCl (Fig. 2~). Preliminary experiments discovered very little differences in affinity for [ ‘H]DHT-receptor complexes among a variety of DNAs. Based on their competitive performance, the following DNAs had no significantly greater affinity for androgen receptor than did calf thymus DNA: pBR322 DNA and pAT153 DNA, each either supercoiled, nicked-relaxed or linearized with either PstI or BumHI, and rat and human prostate DNA (included in the shaded area of Fig. 3a); PBP C3 double-stranded cDNA, or the large (not shown) or small fragment derived from pBR322 DNA by cleavage with both EcoRI and PstI (Fig. 3b). Plasmid p62a DNA, either circularized or linearized, was not a significantly better competitor, but the C3 genomic fragment excised from p62a by EcoRI produced a much better degree of competition (fragment I; Fig. 3a). -___
._____
centrations of [ ‘H]DHT-receptor complexes. Aliquots were incubated and processed as in (a). (c) DNA-cellulose was resuspended in 1 mM Na,PO, pH 7.0, containing 0.1 mM EDTA, 0.25 mM ditbiothreitol, [3H]DHT-receptor complex (2 x 10’ cpm/ml) and various concentrations of NaCl. Samples were incubated at O-4” C for 60 min, then processed as in (a).
263
a i
,‘\,y, P
60.
P
E
‘O 0
I 01
02
03 0 Equvalents coinpetltor DNA
01
02
03
Fig. 3. Calf thymus DNA-cellulose was resuspended in 1 mM Na,PO,, pH 7.0, containing 0.1 mM EDTA, 0.25 mM dithiothreitol, 50 mM NaCI, and [ ‘Hldihydrotestosterone (DHT)-receptor complex (1-2x IO5 cpm/ml). After incubation at 0-4’C for 60 min. aliquots (200 pl, 1-3 pg DNA) were transferred to tubes containing 0.025-0.3 equivalents by weight of soluble competitor DNA, as follows: (a) calf thymus DNA (0); pBR322 DNA, pAT153 DNA (supercoiled, nicked or linearized with either PstI or BarnHI), unfractioned rat and human prostate DNA (all included in the shaded area); fragment I (0); (b) double-stranded DNA complementary to C3 mRNA (0); the 752 bp fragment cleaved from pBR322 DNA by EcoRI and Psi1 (0). and incubated at 25 o C for 30 min. Samples were washed twice in buffer as above (without DHTreceptor complexes), extracted twice in absolute ethanol, and radioactivity in combined extracts was assessed by liquid scintillation spectrometry. Results are the means+ standard deviation (where shown) of quadruplicate experiments.
Cleavage of fragment I with BstEII yielded two fragments (II and III), both of which were improved competitors (Fig. 4a). Fragment II, comprising upstream and exon sequences, was marginally the better competitor than the large fragment III, containing mainly intronic sequences. Digestion of fragment II with Hind11 yielded two further fragments, one (fragment IV, all upstream sequence, - 685 to - 220) of much reduced competitiveness, roughly equivalent to fragment I, and another (fragment V, - 220 to + 80) of improved competitiveness (Fig. 4b). Cleavage of fragment III with BglII also yielded a fragment of reduced competing ability (fragment VI, containing the 3’ 30 bp of the first exon and approx. 900 bp of the intron) and one showing improved competition, fragment VII, which, after digestion with PuuII, allowed retrieval of fragment VIII (Bgf II-PuuII, comprising 500 bp of exclusively intronic sequence) which showed the highest affinity (with fragment V) for androgen-receptor complexes of all the fragments studied (Fig. 4~).
L, 0 Equvalents
competitor
01
02
DNA
Fig. 4. DNA-cellulose was incubated with [3H]dihydrotestosterone-receptor complexes as described in the legend to Fig. 3, and aliquots (200 pl, 1-3 pg DNA) were transferred to tubes containing 0.025-0.25 equivalents by weight of soluble competitor DNA comprising fragments of the C3(1) gene as described in Fig. 1. Aliquots were incubated and processed as described in the legend to Fig. 3. Results are the means (*SD) of quadruplicate determinations carried out using at least three representative receptor preparations.
In some experiments, fragments were obtained by different routes. The intermediate fragments produced by this means were better competitors in the system than was fragment I; fragment X, which contained both high affinity smaller fragments, V and VIII, competing more effectively than either fragment IX or XI, each of which contained only fragment V and competed similarly, the shorter XI being slightly more effective (Fig. 4d). The smaller fragment derived by PuuII cleavage of fragment VII was not obtained in sufficient quantity for testing in the system. Discussion Studies of the regulation of specific gene expression by androgen receptors have not been as
264
fruitful as similar investigations using receptors for other classes of steroid hormones. This can be attributed to the difficulties in obtaining functional androgen receptors in purified form from tissues rich in protease activities, reflected in observations of numerous proteolytic fragments (Lea et al., 1979) some of which lack specific DNAbinding activity (Mulder et al., 1984) while retaining other binding characteristics (Rowley et al., 1986). For this study, we used a receptor preparation with proven DNA-binding capability (Davies et al., 1986) and lacking nucleolytic activity and thus we have successfully located regions of an androgen-responsive gene, the rat ventral prostate steroid-binding protein subunit C3(1) gene, with comparatively high affinity for the androgen-receptor complex. In a previous report (Davies and Thomas, 1983), using a similar receptor preparation, we observed no selective retention of androgen receptor by DNA complementary to C3(1) mRNA. This is explained by the present results. The two retions of the genomic C3(1) fragment which exhibit the highest-affi~ty binding are one which comprises predo~nantly 5 ‘-flanking sequences ( - 220 to + 80, fragment V) and one which is intronic (t 1000 to + 1500, fragment VIII). Receptor-binding regions 5’ to the transcription start-sites of steroid-regulated genes have been reported frequently (see Yamamoto, 1985) and intronic receptor-binding regions have also been noted (Moore et al., 1985; Slater et al., 1985; Bailly et al., 1986). The relative increases in affinity over nonspecific (calf thymus or plasmid) DNA which can be calculated from the data reported here (17.5-fold for fragment V and 24.4-fold for fragment VIII) are of the order that would be expected, from scanning the literature, for fragments of these sizes. It is premature, due to the size of the fragments, to conjecture on the precise nature of the receptor-binding site. Previously reported data on the sequence of part of C3(1) DNA including fragment V have been compared to known regulatory elements including regions bound by receptor for other classes of steroid hormones (Hurst and Parker, 1983). Althou~ no functional significance has been ascribed to intronic receptor-binding domains (Moore et al., 1985; Slater et al.,
1985; Bailly et al., 1986) it is noteworthy that upstream regions of the C3(1) gene conferred only a disappointing level of androgen responsiveness on a heterologous coding sequence (Parker et al., 1984). A full complement of regulatory elements may be essential for complete androgen responsiveness. Using the systems described in this paper as a starting-point we are now in a position to study androgen-receptor interaction with and regulation of specific genes in precise detail.
The authors are grateful to the Tenovus Organization for their generous financial support and to Professor Keith Griffiths for encouragement and enthusiasm. References Alberts, B. and Herrick, G. (1971) Methods Enzymol. 21, 198-217. BaiIIy, A., Atger, M., Atger, P., Cerbon, M.A., Al&on, M.. VuHai, M.T., Logeat, F. and Milgrom, E. (1983) J. Biol. Chem. 258,10384-10389. Bailfy, A., Le Page, C., Rauch. M. and Milgrom, E. (1986) EMBO J. 5, 3235- 3241. Chamness, G.C. and McGuire, W.L. (1972) Biochemistry 11, 2466-2472. Chandler, V.L., Maler, B.A. and Yamamoto, K.R. (1983) Cell 33, 4x9-499. Compton, J.G., Schrader, W.T. and O’Malley, B.W. (1983) Proc. Nat]. Acad. Sci. U.S.A. 80, 16-20. Davies, P. and Thomas, P. (1983) J. Steroid B&hem. 20, 57-65. Davies, P., Thomas, P. and Manning, D.L. (1986) Prostate 8, 151-166. Dean, D.C., Knoll, B.J., Riser, M.E. and O’Malley, B.W. (1983) Nature 305, 551-554. Geisse, S., Scheidereit, C., Westphal, H.M., Hynes, N.E., Groner, B. and Beato, M. (1982) EMBO J. 1, X13-1619. Govindan, M.V., Spiers, E. and Majors, J. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 5157-5161. Gurney, T, and Gurney, E.G. (1984) In: Methods in Molecular Biology, Vol. 2, Nucleic Acids, Ed.; J.M. Walker (Humana Press, Clifton, NJ) pp. 5-12. Heyns, W. and De Moor, P. (1977) Eur. J. Biochem. 78, 221-230. Heyns. W., Peeters, B., Mous, J., Rombounts, W. and De Moor, P. (1978) Eur. .I. B&hem. 89, 181-186. Hurst, H.C. and Parker, M.G. (1983) EMBO J. 2, 769.-774. Hynes, N.E., van Ooyen, A.J.J., Kennedy, N., Herrlich, P., Ponta, H. and Groner, B. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 3637-3641.
265 Jost, J.-P., Seldran, M. and Geiser, M. (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 429-433. Karin, M., Haslinger, A., Holtgreve, H., Richards, RI., Krauter, P., Westphal, H.M. and Beato, M. (1984) Nature 308, 513-519. Lea, O.A.. Petrusz, P. and French, F.S. (1979) J. Biol. Chem. 254, 6169-6173. Maurer, R.A. (1985) DNA 4, 1-8. Moore, D.D., Marks, A.R., Buckley, D.I., Kapler, G., Payvar, F. and Goodman, H.M. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 699-702. Mulder, E., Vrij, A.A., Brinkmann. A.O., van der Molen, H.J. and Parker, M.G. (1984) Biochim. Biophys. Acta 781, 121-129. Mulvihill, E.R., Le Pennec, J.-P. and Chambon, P. (1982) Cell 24, 621-632. Parker, M.G. (1983) Nature 304, 687-688. Parker, M.G., Scrace, G.T. and Mainwaring, W.I.P. (1978) B&hem. J. 170, 115-121. Parker, M.G., Needham, M., White, R., Hurst, H. and Page, M. (1982) Nucleic Acids Res. 10, 5121-5132. Parker, M.G., White, R., Hurst, H.. Needham, M. and Tilly, R. (1983) J. Biol. Chem. 258, 12-15. Parker, M.G., Needham, M., White, R., Hurst, H. and Page, M. (1982) Nucleic Acids Res. 10, 5121-5132.
Payvar, F., Wrange, O., Carlstedt-Duke, J., Okret, S., Gustafsson, J.A. and Yamamoto, K.R. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 6628-6632. Payvar, F., De France, D., Firestone, G.L., Edgar, B., Wrange, O., Okret, S., Gustafsson, J.A. and Yamamoto, K.R. (1983) Cell 35, 381-392. Perry, ST., Viskochil, D.H., Ho, K.-C., Fong. K., Stafford, D.W.. Wilson, E.M. and French, F.S. (1985) In: Regulation of Androgen Action, Eds.: N. Bruchovsky, A. Chapdelaine and F. Neumann (Congressdruck R., Bruckner, Berlin) pp. 1677173. Pfahl, M. (1982) Cell 31, 475-482. Renkawitz, R., Schutz, G., van der Ahe, D. and Beato, M. (1984) Cell 37, 503-510. Rowley, D.R., Premont, R.T., Johnson, M.P., Young, C.Y.F. and Tindall, D.J. (1986) Biochemistry 25, 6988-6995. Scheidereit, C., Geisse, S., Westphal, H.M. and Beato, M. (1983) Nature 304, 749-752. Slater. E.P., Rabenau, O., Karin, M., Baxter, J.D. and Beato, M. (1985) Mol. Cell. Biol. 5, 2984-2992. Weisz, A., Coppola, L. and Bresciani, F. (1986) Biochem. Biophys. Res. Commun. 139, 396-402. Yamamoto, K.R. (1985) Annu. Rev. Genet. 19, 209-252.
Molecular and Cellular Endocrinology, 51 (1987) 259-265 Elsevier Scientific Publishers Ireland, Ltd.
MCE 01672
Androgen
receptor-binding
regions of an androgen-responsive
N.K. Rushmere,
M.G. Parker
1
gene
and P. Davies
Tenoous Institute for Cancer Research, University of Wales College of Medicine, Heath Park, Cardiff CF4 4Xx, and ’ Imperral Cancer Research Fund Laboratories, P.O. Box 123, Lincoln’s Inn Fields, London WCZA 3PX, U.K. (Received
Key words: Steroid-binding
protein
15 January
C3(1) gene; (Prostate);
1987; accepted
12 February
Sa-Dihydrotestosterone-receptor
1987)
complex
Summary The interaction of 5a-dihydrotestosterone-receptor complexes with purified DNA fragments representing upstream, coding and intervening sequences of the prostate binding protein C3(1) gene was investigated using a DNA-cellulose competition binding assay. The partially purified androgen-receptor complexes which were used in the assay had proven DNA-binding capabilities. Two fragments were identified with relatively high affinity for androgen-receptor complexes. A 300 bp fragment extending from -220 to +80 and a 500 bp fragment derived entirely from the first intron consistently competed most effectively in the system. The presence of a high affinity site or sites in or near the promoter region of the gene is consistent with current models of transcriptional activation of hormone-responsive genes by steroid receptors. High affinity sites for steroid receptors within introns may indicate a role for receptors in regulation of transcription at other stages, or in post-transcriptional modification.
Introduction Steroid hormone receptor proteins are selective regulators of gene transcription (see Yamamoto, 1985). Their exact mechanism remains unknown but centres on the fact that binding of the cognate steroid hormone induces a receptor conformation appropriate to selective interaction with intranuclear sites, presumably specific DNA regions in the vicinity of responsive genes (Parker, 1983; Yamamoto, 1985). This assumption has been vindicated by the use of purified receptors and cloned hormone-responsive genes (Payvar et al., 1981, 1983; Geisse et al., 1982; Govindan et al.,
Address for correspondence: Dr. Peter Davies, Tenovus Institute for Cancer Research, University of Wales College of Medicine, Heath, Cardiff CF4 4Xx, U.K. 0303-7207/87/$03.50
0 1987 Elsevier Scientific
Publishers
Ireland,
1982; Mulvihill et al., 1982; Compton et al., 1983: Scheidereit et al., 1983). Regions of DNA with high affinity for receptors have been located upstream of and within genes (Payvar et al., 1983; Moore et al., 1985; Slater et al., 1985). An essential function for these DNA segments in receptormodulated transcription was proven by transfer into recipient cells of genes containing or lacking the sequences (Chandler et al., 1983; Dean et al., 1983; Hynes et al., 1983; Karin et al., 1984; Renkawitz et al., 1984). The conclusion to be drawn was that the receptor associated with specific promoter elements near or inside the affected genes (see Yamamoto, 1985). Sequences with high affinity for receptors for glucocorticoids (Payvar et al., 1981, 1983; Govindan et al., 1982; Pfahl, 1982; Scheidereit et al., 1983; Karin et al., 1984), progesterone (Mulvihill et al., 1982; Bailly et al., 1983; Dean et al., 1983; Ltd
260
Renkawitz et al., 1984) and oestrogens (Jost et al., 1984; Maurer, 1985; Weisz et al., 1986) have been described. There is, however, little information regarding interactions of androgen receptors and regions of androgen-responsive genes. The major secretory product of the rat ventral prostate, prostate steroid-binding protein, PBP (Heyns and De Moor, 1977) is an oligomer containing Cl, C2 and C3 polypeptides (Heyns et al., 1978), whose rates of synthesis are stimulated markedly by androgens (Parker et al., 1978). The Cl and C2 polypeptides are encoded by unique genes (Parker et al., 1982) but there are two nonallelic C3 genes, C3(1) and C3(2) (Parker et al., 1983). Only the C3(1) gene is responsible for the production of the C3 polypeptide in the rat ventral prostate (Hurst and Parker, 1983). Earlier attempts to define regions around this gene with high affinity for androgen receptors have met with varying success (Mulder et al., 1974; Perry et al., 1985). The fact that some were unsuccessful was due probably to deficiencies in the DNA-binding capabilities of the receptor which was used (Mulder et al., 1984). This report relates investigations into receptor binding in the vicinity of the C3(1) gene using a partially purified androgen receptor with proven affinity for DNA (Davies et al., 1986). Materials and methods [1,2,4,5,6,7-‘HISa-Dihydrotestosterone (110 Ci/mmol) was provided by Amersham International PLC (Amersham, Bucks, U.K.), who also supplied the following enzymes: restriction endonucleases EcoRI, Hind11 and PuuII; T4 DNA polymerase, T4 DNA ligase, and also synthetic oligonucleotide sequences containing EcoRI cleavage sites (EcoRI linkers). Nensorb 20 nucleic acid purification cartridges were bought from NEN Research Products (Du Pont (U.K.), Stevenage, Herts, U.K.). The Boehringer Corporation (London) (Lewes, U.K.) supplied Sl nuclease and restriction endonucleases Pst I, BglII and BstEII. All other chemicals were of the highest quality commercially available. Preparation
of androgen
receptor protein
The source of androgen receptor prostate cytosol from rats castrated
was ventral18 h earlier,
for purposes of maximum yield (Davies and Thomas, 1983). Receptor in complex with [ 3H]DHT was partially purified (approx. IOOOfold) by chromatography on Cibacron F3GAagarose and heparin-Sepharose, followed by ammonium sulphate precipitation and desalting on Sephadex G25 (Davies and Thomas, 1983). Further purification resulted in poor reproducibility in the DNA-cellulose competition assay (see below), due mainly to problems of unsaturability (Davies and Thomas, 1983). Androgen receptor protein purified to the stage described above bound to both DNA and chromatin, the latter tissue-specifically, and lacked deoxyribonuclease activity as judged by its inability to alter the electrophoretic mobility of supercoiled DNA. The quantity of [ 3H]DHT-receptor complexes present in a sample was estimated by a combination of protamine sulphate precipitation and sedimentation analysis. The latter revealed the androgenreceptor complex to have a sedimentation coefficient of 4.2 + 0.2s. Under conditions of scintillation spectrometry employed (29% efficiency for 3H), 1 pmol r ec ep tor was represented by 84 x lo3 cpm assuming 1 mol receptor bound 1 mol steroid. Preparation
of cloned DNA fragments
The genomic region investigated comprised the first exon of the PBP C3(1) gene, 684 bp of 5’-flanking sequence, the whole of the first intervening sequence, and a portion of the second exon, in all 2.5 kbp. Restriction fragments derived from this sequence are shown in Fig. 1 and denoted by Roman numerals. Briefly, they are: I. (EcoRI-EcoRI), the entire sequence as cleaved from the plasmid; II. (EcoRI-BstEII) 766 bp, -685 to + 80; III. (BstEII-EcoRI) 1735 bp; IV. (EcoRI-HindII) 465 bp, - 685 to - 220; V. (HindII-BstEII) 300 bp, - 200 to + 80; VI. (BstEII-BglII) 935 bp; VII. (BglII-EcoRI) 800 bp; VIII. (BglII-PuuII) 500 bp; IX. (EcoRI-BglII) 1.7 kbp; X. (HindII-EcoRI) 2.1 kbp; XI. (HindII-BglII) 1.24 kbp. The positions of these fragments in relation to exons, the intron and upstream sequence can be obtained by reference to Fig. 1. The constitution
261 x!
Fig. 1. The segment of the prostate steroid-binding protein C3(1) gene (I) which was released from p62A by cleavage with EcoRI (E) was divided into smaller fragments (II to XI) using combinations of the following restriction endonucleases: B, BgIII; BE, BsrEIl; H, HindII; P, PuuII. The various fragments are designated by the Roman numeral by which they are referred in the text. The fragments (II to VIII) arranged below fragment I in the figure are those obtained by the digestion route most frequently taken; those above (IX to XI) were derived infrequently.
of the more relevant fragments will be discussed in the text. For comparison, pBR322 was digested with EcoRI and PstI, to generate fragments of 0.75 kbp and 3.6 kbp. Recombinant plasmids were purified from clear lysates on caesium chloride gradients. Fragments from restriction enzyme-cleaved plasmids were analysed by electrophoresis on 1% (w/v) agarose gels and identified in relation to molecular size markers by staining with ethidium bromide and illuminating at 300 nm. Gel slices corresponding to required fragments were excised and incubated (60 min at 37°C) in saturated sodium iodide (2 ml/g wet gel). Extracts were mixed with an equal volume of buffer A (100 mM Tris, 1 mM Tris, 1 mM EDTA, 10 mM triethylamine, pH 7.7) and applied to a pre-wetted (100% methanol) and primed (1 ml buffer A) Nensorb column. Columns were washed with 3 ml buffer A and 5-10 ml H,O, and DNA was eluted with 500 IJ.I 50% (v/v) aqueous methanol. DNA was evaporated to dryness, resuspended in 10 mM Tris-HCl, pH 7.6, containing 1 mM EDTA (buffer B), and extracted 3 times with an equal volume of water-saturated isoamyl alcohol. Residual isoamyl alcohol was removed with water-saturated diethyl ether, ether was evaporated under nitrogen, and DNA was
ethanol-precipitated. The concentration of DNA resuspended in buffer B was determined as described by Gurney and Gurney (1984). Cohesive termini were eliminated by treatment with Sl nuclease or T4 DNA polymerase. When necessary, blunted fragments were ligated to EcoRI linkers using T4 DNA ligase and subcloned into the EcoRI site of pBR322. Soluble calf thymus DNA or rat ventral prostate DNA used in competition studies was sheared to an average size of 600 bp by sonication, treated with Sl nuclease, and purified as above. DNA-cellulose competition assay Calf thymus DNA was immobilized on cellulose by the method of Alberts and Herrick (1971). Two approaches to competitive binding assays were tested: either [ 3H]DHT-receptor was bound to DNA-cellulose prior to addition of competing fragments, or fragments were mixed with DNA-cellulose prior to addition of receptor. After ensuring the methods gave similar results, the former was used more frequently. Typically, DNA-cellulose was pelleted and resuspended in 1 mM Na,PO,, pH 7.0, containing 0.1 mM EDTA, 0.25 mM dithiothreitol and 50 mM NaCl, and [3H]DHT-receptor (l-2 X 10’ cpm/ml). After incubation at 0-4°C for 60 min, aliquots (200 ~1, equivalent to l-3 pg calf thymus DNA) were pipetted into tubes containing increasing quantities of soluble competitor DNA to be tested (2-10 ~1). Assays were usually in quadruplicate. After further incubation for 30 min at 25’C, samples were chilled, diluted to 1 ml in ice-cold buffer, centrifuged at 5000 X g for 10 min, pellets were washed briefly in ice-cold buffer, repelleted, extracted twice in absolute ethanol, extracts were combined and incorporated radioactivity was assessed. Results The characteristics of retention of [ 3H]DHT-receptor complexes by calf thymus DNA immobilized on cellulose are shown in Fig. 2. Routine assay conditions were derived from these experimental data. As indicated in Fig. 2a and b, under the conditions used and with this receptor preparation, binding was saturable. Maximum binding
262
.
.
10 20 30 LO
60 Time ImInI
120
I
I
I
2 3 L 5 l:put PHIDHT-R(dpm ~10-~1
6
(c)
’ 25
I
b
I
1
#
1
35
L5
55
65
75
85
&
!
95 100
NoCI (mmol/L)
Fig. 2. (a) DNA-cellulose was resuspended in 1 mM Na,PO,, pH 7.0, containing 0.1 mM EDTA, 0.25 mM dithiothreitol and 50 mM NaCI, and [3H]dihydrotestosterone (DHT)-receptor complex (2 x lo5 cpm/ml), in the absence (0) and presence (0) of unlabelled [ 3H]DHT-receptor complex (250 nM). Samples were incubated at 0-4OC and, at various times, aliquots equivalent to 3 ng of calf thymus DNA were removed, washed twice with the same buffer, extracted twice in absolute ethanol, and the radioactivity contained in combined extracts assessed by liquid scintillation spectrometry. (b) DNA cellulose was resuspended in buffer as in (o) but containing varying con-
occurred within 30 min and was stable for at least 2 h at 0-4°C (Fig. 2a). Labelling of receptor with [3H]DHT in the presence of a 100-fold molar excess of nonradioactive DHT significantly reduced the radioactivity subsequently associated with DNA-cellulose. Nonspecific binding was lo-15% of total. Saturability of the system may have been due to the relatively impure receptor preparation containing other DNA-binding proteins (Chamness and McGuire, 1972; see Mulvihill et al., 1982); certainly, increased purification of the androgen receptor abrogates saturation of DNA-binding sites (Davies and Thomas, 1983). However, as chromatography on immobilized triazynyl dyes achieved significant removal of proteins with affinity for DNA, some other factors may be involved. The effects demonstrated in Fig. 2 were consistent throughout the experiments reported. An input of 2.5-4 X lo4 cpm of [3H]DHT-receptor complex per l-3 pg DNA in 200 ~1 resulted in binding of 19.4 + 4.2% (mean + SD) of total. Preformed complexes were most stable at 50-100 mM NaCl (Fig. 2~). Preliminary experiments discovered very little differences in affinity for [ ‘H]DHT-receptor complexes among a variety of DNAs. Based on their competitive performance, the following DNAs had no significantly greater affinity for androgen receptor than did calf thymus DNA: pBR322 DNA and pAT153 DNA, each either supercoiled, nicked-relaxed or linearized with either PstI or BumHI, and rat and human prostate DNA (included in the shaded area of Fig. 3a); PBP C3 double-stranded cDNA, or the large (not shown) or small fragment derived from pBR322 DNA by cleavage with both EcoRI and PstI (Fig. 3b). Plasmid p62a DNA, either circularized or linearized, was not a significantly better competitor, but the C3 genomic fragment excised from p62a by EcoRI produced a much better degree of competition (fragment I; Fig. 3a). -___
._____
centrations of [ ‘H]DHT-receptor complexes. Aliquots were incubated and processed as in (a). (c) DNA-cellulose was resuspended in 1 mM Na,PO, pH 7.0, containing 0.1 mM EDTA, 0.25 mM ditbiothreitol, [3H]DHT-receptor complex (2 x 10’ cpm/ml) and various concentrations of NaCl. Samples were incubated at O-4” C for 60 min, then processed as in (a).
263
a i
,‘\,y, P
60.
P
E
‘O 0
I 01
02
03 0 Equvalents coinpetltor DNA
01
02
03
Fig. 3. Calf thymus DNA-cellulose was resuspended in 1 mM Na,PO,, pH 7.0, containing 0.1 mM EDTA, 0.25 mM dithiothreitol, 50 mM NaCI, and [ ‘Hldihydrotestosterone (DHT)-receptor complex (1-2x IO5 cpm/ml). After incubation at 0-4’C for 60 min. aliquots (200 pl, 1-3 pg DNA) were transferred to tubes containing 0.025-0.3 equivalents by weight of soluble competitor DNA, as follows: (a) calf thymus DNA (0); pBR322 DNA, pAT153 DNA (supercoiled, nicked or linearized with either PstI or BarnHI), unfractioned rat and human prostate DNA (all included in the shaded area); fragment I (0); (b) double-stranded DNA complementary to C3 mRNA (0); the 752 bp fragment cleaved from pBR322 DNA by EcoRI and Psi1 (0). and incubated at 25 o C for 30 min. Samples were washed twice in buffer as above (without DHTreceptor complexes), extracted twice in absolute ethanol, and radioactivity in combined extracts was assessed by liquid scintillation spectrometry. Results are the means+ standard deviation (where shown) of quadruplicate experiments.
Cleavage of fragment I with BstEII yielded two fragments (II and III), both of which were improved competitors (Fig. 4a). Fragment II, comprising upstream and exon sequences, was marginally the better competitor than the large fragment III, containing mainly intronic sequences. Digestion of fragment II with Hind11 yielded two further fragments, one (fragment IV, all upstream sequence, - 685 to - 220) of much reduced competitiveness, roughly equivalent to fragment I, and another (fragment V, - 220 to + 80) of improved competitiveness (Fig. 4b). Cleavage of fragment III with BglII also yielded a fragment of reduced competing ability (fragment VI, containing the 3’ 30 bp of the first exon and approx. 900 bp of the intron) and one showing improved competition, fragment VII, which, after digestion with PuuII, allowed retrieval of fragment VIII (Bgf II-PuuII, comprising 500 bp of exclusively intronic sequence) which showed the highest affinity (with fragment V) for androgen-receptor complexes of all the fragments studied (Fig. 4~).
L, 0 Equvalents
competitor
01
02
DNA
Fig. 4. DNA-cellulose was incubated with [3H]dihydrotestosterone-receptor complexes as described in the legend to Fig. 3, and aliquots (200 pl, 1-3 pg DNA) were transferred to tubes containing 0.025-0.25 equivalents by weight of soluble competitor DNA comprising fragments of the C3(1) gene as described in Fig. 1. Aliquots were incubated and processed as described in the legend to Fig. 3. Results are the means (*SD) of quadruplicate determinations carried out using at least three representative receptor preparations.
In some experiments, fragments were obtained by different routes. The intermediate fragments produced by this means were better competitors in the system than was fragment I; fragment X, which contained both high affinity smaller fragments, V and VIII, competing more effectively than either fragment IX or XI, each of which contained only fragment V and competed similarly, the shorter XI being slightly more effective (Fig. 4d). The smaller fragment derived by PuuII cleavage of fragment VII was not obtained in sufficient quantity for testing in the system. Discussion Studies of the regulation of specific gene expression by androgen receptors have not been as
264
fruitful as similar investigations using receptors for other classes of steroid hormones. This can be attributed to the difficulties in obtaining functional androgen receptors in purified form from tissues rich in protease activities, reflected in observations of numerous proteolytic fragments (Lea et al., 1979) some of which lack specific DNAbinding activity (Mulder et al., 1984) while retaining other binding characteristics (Rowley et al., 1986). For this study, we used a receptor preparation with proven DNA-binding capability (Davies et al., 1986) and lacking nucleolytic activity and thus we have successfully located regions of an androgen-responsive gene, the rat ventral prostate steroid-binding protein subunit C3(1) gene, with comparatively high affinity for the androgen-receptor complex. In a previous report (Davies and Thomas, 1983), using a similar receptor preparation, we observed no selective retention of androgen receptor by DNA complementary to C3(1) mRNA. This is explained by the present results. The two retions of the genomic C3(1) fragment which exhibit the highest-affi~ty binding are one which comprises predo~nantly 5 ‘-flanking sequences ( - 220 to + 80, fragment V) and one which is intronic (t 1000 to + 1500, fragment VIII). Receptor-binding regions 5’ to the transcription start-sites of steroid-regulated genes have been reported frequently (see Yamamoto, 1985) and intronic receptor-binding regions have also been noted (Moore et al., 1985; Slater et al., 1985; Bailly et al., 1986). The relative increases in affinity over nonspecific (calf thymus or plasmid) DNA which can be calculated from the data reported here (17.5-fold for fragment V and 24.4-fold for fragment VIII) are of the order that would be expected, from scanning the literature, for fragments of these sizes. It is premature, due to the size of the fragments, to conjecture on the precise nature of the receptor-binding site. Previously reported data on the sequence of part of C3(1) DNA including fragment V have been compared to known regulatory elements including regions bound by receptor for other classes of steroid hormones (Hurst and Parker, 1983). Althou~ no functional significance has been ascribed to intronic receptor-binding domains (Moore et al., 1985; Slater et al.,
1985; Bailly et al., 1986) it is noteworthy that upstream regions of the C3(1) gene conferred only a disappointing level of androgen responsiveness on a heterologous coding sequence (Parker et al., 1984). A full complement of regulatory elements may be essential for complete androgen responsiveness. Using the systems described in this paper as a starting-point we are now in a position to study androgen-receptor interaction with and regulation of specific genes in precise detail.
The authors are grateful to the Tenovus Organization for their generous financial support and to Professor Keith Griffiths for encouragement and enthusiasm. References Alberts, B. and Herrick, G. (1971) Methods Enzymol. 21, 198-217. BaiIIy, A., Atger, M., Atger, P., Cerbon, M.A., Al&on, M.. VuHai, M.T., Logeat, F. and Milgrom, E. (1983) J. Biol. Chem. 258,10384-10389. Bailfy, A., Le Page, C., Rauch. M. and Milgrom, E. (1986) EMBO J. 5, 3235- 3241. Chamness, G.C. and McGuire, W.L. (1972) Biochemistry 11, 2466-2472. Chandler, V.L., Maler, B.A. and Yamamoto, K.R. (1983) Cell 33, 4x9-499. Compton, J.G., Schrader, W.T. and O’Malley, B.W. (1983) Proc. Nat]. Acad. Sci. U.S.A. 80, 16-20. Davies, P. and Thomas, P. (1983) J. Steroid B&hem. 20, 57-65. Davies, P., Thomas, P. and Manning, D.L. (1986) Prostate 8, 151-166. Dean, D.C., Knoll, B.J., Riser, M.E. and O’Malley, B.W. (1983) Nature 305, 551-554. Geisse, S., Scheidereit, C., Westphal, H.M., Hynes, N.E., Groner, B. and Beato, M. (1982) EMBO J. 1, X13-1619. Govindan, M.V., Spiers, E. and Majors, J. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 5157-5161. Gurney, T, and Gurney, E.G. (1984) In: Methods in Molecular Biology, Vol. 2, Nucleic Acids, Ed.; J.M. Walker (Humana Press, Clifton, NJ) pp. 5-12. Heyns, W. and De Moor, P. (1977) Eur. J. Biochem. 78, 221-230. Heyns. W., Peeters, B., Mous, J., Rombounts, W. and De Moor, P. (1978) Eur. .I. B&hem. 89, 181-186. Hurst, H.C. and Parker, M.G. (1983) EMBO J. 2, 769.-774. Hynes, N.E., van Ooyen, A.J.J., Kennedy, N., Herrlich, P., Ponta, H. and Groner, B. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 3637-3641.
265 Jost, J.-P., Seldran, M. and Geiser, M. (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 429-433. Karin, M., Haslinger, A., Holtgreve, H., Richards, RI., Krauter, P., Westphal, H.M. and Beato, M. (1984) Nature 308, 513-519. Lea, O.A.. Petrusz, P. and French, F.S. (1979) J. Biol. Chem. 254, 6169-6173. Maurer, R.A. (1985) DNA 4, 1-8. Moore, D.D., Marks, A.R., Buckley, D.I., Kapler, G., Payvar, F. and Goodman, H.M. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 699-702. Mulder, E., Vrij, A.A., Brinkmann. A.O., van der Molen, H.J. and Parker, M.G. (1984) Biochim. Biophys. Acta 781, 121-129. Mulvihill, E.R., Le Pennec, J.-P. and Chambon, P. (1982) Cell 24, 621-632. Parker, M.G. (1983) Nature 304, 687-688. Parker, M.G., Scrace, G.T. and Mainwaring, W.I.P. (1978) B&hem. J. 170, 115-121. Parker, M.G., Needham, M., White, R., Hurst, H. and Page, M. (1982) Nucleic Acids Res. 10, 5121-5132. Parker, M.G., White, R., Hurst, H.. Needham, M. and Tilly, R. (1983) J. Biol. Chem. 258, 12-15. Parker, M.G., Needham, M., White, R., Hurst, H. and Page, M. (1982) Nucleic Acids Res. 10, 5121-5132.
Payvar, F., Wrange, O., Carlstedt-Duke, J., Okret, S., Gustafsson, J.A. and Yamamoto, K.R. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 6628-6632. Payvar, F., De France, D., Firestone, G.L., Edgar, B., Wrange, O., Okret, S., Gustafsson, J.A. and Yamamoto, K.R. (1983) Cell 35, 381-392. Perry, ST., Viskochil, D.H., Ho, K.-C., Fong. K., Stafford, D.W.. Wilson, E.M. and French, F.S. (1985) In: Regulation of Androgen Action, Eds.: N. Bruchovsky, A. Chapdelaine and F. Neumann (Congressdruck R., Bruckner, Berlin) pp. 1677173. Pfahl, M. (1982) Cell 31, 475-482. Renkawitz, R., Schutz, G., van der Ahe, D. and Beato, M. (1984) Cell 37, 503-510. Rowley, D.R., Premont, R.T., Johnson, M.P., Young, C.Y.F. and Tindall, D.J. (1986) Biochemistry 25, 6988-6995. Scheidereit, C., Geisse, S., Westphal, H.M. and Beato, M. (1983) Nature 304, 749-752. Slater. E.P., Rabenau, O., Karin, M., Baxter, J.D. and Beato, M. (1985) Mol. Cell. Biol. 5, 2984-2992. Weisz, A., Coppola, L. and Bresciani, F. (1986) Biochem. Biophys. Res. Commun. 139, 396-402. Yamamoto, K.R. (1985) Annu. Rev. Genet. 19, 209-252.