Expression of the Human Androgen Receptor Gene Utilizes a ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1990 by The American Society for Biochemistry

Vol. 265, No. 23, Issue of August 15, pp. 13776-13781,199O

Printed in U.S. A.

and Molecular Biology, Inc.

Expression of the Human Androgen Receptor Gene Utilizes a Common Promoter in Diverse Human Tissues and Cell Lines* (Received for publication,

March 29, 1990)

Wayne D. TilleyS, Marco Marcelli, and Michael J. McPhaulQ From the DeDartment of Internal Medicine. The University of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75235-8857

Androgens mediate their effects through an intracellular receptor that is a member of the steroid/thyroid hormone family of receptors. The expression of this protein is tightly regulated in different tissues and among cell types within a single tissue. To define the mechanisms controlling the expression of the androgen receptor, we have isolated and characterized the promoter of the androgen receptor gene in the human prostate cell line LNCaP. The major site of transcription initiation is approximately 1.1 kilobases upstream of the initiator methionine of the androgen receptor protein. The promoter region lacks typical “TATA” and “CAAT” sequence motifs but lies in a GC-rich region and contains a putative Spl binding site characteristic of a “housekeeping” promoter and a 44-base segment composed of alternating adenosine and guanosine residues. S1 nuclease protection analyses indicate that the same promoter is employed both in human tissues (prostate, testes), in genital skin fibroblasts, in T47D and MCF-7 breast cancer cells, and in LNCaP prostate cancer cells.

The actions of the androgens testosterone and dihydrotestosterone are mediated by a specific receptor protein termed the androgen receptor. Defects that impair this protein cause abnormalities of male phenotypic sexual development in animals and in man (1). The structural defects in the androgen receptor gene that have caused androgen resistance have been defined in several patients (2-4). While the presence of a functional androgen receptor gene is essential for normal male development, the distribution and timing of androgen receptor expression are equally important. In the rat gubernaculum (5) and penis (6), changes in the tissue content of androgen receptor are correlated with specific stages in tissue growth and development. In other tissues such as the mouse prostate, androgen receptor is expressed differently in the mesenchyme and epithelium at different stages of development (7). To understand the factors that control androgen receptor expression during embryogenesis and in the adult, it is necessary to define the genetic elements that are responsible for controlling the activity of the androgen receptor gene. In the * This work was supported by a Basil O’Connor Award from the March of Dimes, the Medical Life and Health Insurance Medical Research Fund, the Welch Foundation, a grant from the Perot Family Foundation, and Grant DK03892 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Recipient of a C. J. Martin Fellowship from the National Health and Medical Research Council of Australia. § Culpeper medical scholar.

current work we have defined the promoter region of the androgen receptor gene and have examined the androgen receptor mRNA species in several human cell lines and tissues. MATERIALS

AND

METHODS’

RESULTS

Similar Species of Androgen Receptor mRNA Are Detected in Diverse Tissues That Express the Androgen Receptor-To determine whether the species of androgen receptor mRNA are similar in different human cell types, we examined RNA prepared from human tissues and cell lines by Northern analysis (Fig. 1). In all of the RNA specimens, two predominant bands are detected, one of approximately 10 kilobases and a second of approximately 6 to 7 kilobases in length. An identical pattern is detected in these samples when this analysis is performed using probes derived from segments encoding either the hormone-binding domain or the amino-terminal segment of the receptor.’ Since the pattern of androgen receptor mRNA in the prostate carcinoma cell line LNCaP was indistinguishable from that in the other human tissues and cell lines, we focused our attention first on the LNCaP cell line which expresses high levels of the androgen receptor. The Two Major Species of Androgen Receptor mRNA Have Similar Structures at Their 5’ Termini-To determine the origin of the two major androgen receptor mRNA species, we performed Northern analysis using a cDNA probe (probe 3) and segments of genomic DNA (probes 1, 2, and 4) derived from the region flanking the 5’-terminal portion of the androgen receptor coding segment (Fig. 2). Probes 2 and 3 detect both the lo- and 7-kilobase androgen receptor mRNA bands. A smaller PstI-BstBI restriction endonuclease fragment (probe 4) produced an identical result to that seen with probe 2 (results not shown). This analysis also shows that a probe derived from segment 1 does not detect either androgen receptor mRNA band. These findings suggested that either the site of initiation of androgen receptor mRNA or a splice junction within the androgen receptor gene is located very close to the SmaI restriction endonuclease cleavage site located approximately 1.1 kilobases upstream of the initiator methionine of the androgen receptor protein, as depicted schematically in Fig. 2. Primer Extension and S, Nuclease Analysis of Androgen Receptor mRNA Transcripts-To localize the transcription initiation site of the androgen receptor gene, we performed primer extension and S1 nuclease protection analyses (Fig. 3). 1 The “Materials and Methods” is presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press. * W. D. Tilley and M. J. McPhaul, unpublished observations.

13776

Androgen

9614-

-

Receptor

13777

Gene Promoter

-9614

6662 -

-f&32

& ;;;; 2.

A.

B.

c

FIG. 1. Androgen receptor mRNA expression in human cell lines and tissues. RNA samples were electrophoresed on 1% agarose gels containing 2.2 M formaldehyde and transferred to Zetaprobe membranes. Panel A, poly(A)+ RNA: prostate 10 ~g, LNCaP 10 fig; Panel B, total RNA: LNCaP 20 pg, T47D 25 pg, prostate 25 pg; Panel C; total RNA: testes 30 pg, genital skin fibroblasts (GSJ’) 30 pg. In this analysis, a labeled fragment of the human androgen receptor cDNA that encodes a segment of the hormone-binding domain of the androgen receptor was employed (nucleotides 1850-2563 in Ref. 11). Approximate sizes of the bands are indicated by the position of labeled DNA markers, shown in nucleotides in the margins. Panels A, B, and C are derived from three separate experiments. 1 I 4l I R x

2 III ‘Fj[rsfi

Lp

II P s

Genomtc ‘1 12.281 R

I--“. K

2

-“+

-

9614

-

6682

-

4361

FIG. 2. The two major species of androgen receptor mRNA have a similar structure at their 5’ termini. Three identical 20pg samples of LNCaP RNA were electrophoresed on a 0.9% formaldehyde gel and transferred to Zetaprobe membranes. Following transfer, the filter was cut into three sections and hybridized to a segment of the androgen receptor cDNA (probe 3) and portions of the 5’. flanking region of the human androgen receptor gene (probes 1, 2, and 4). R is EcoRI, X is XbaI, N is BstNI, S is SmaI, B is BstBI, P is PstI, and K is KpnI. The bored restriction endonuclease cleavage sites (N, S, B, P) are those depicted in the schematic diagram in Figs. 3 and 5. bp, base pair. The primer oligonucleotide

extension L42B

experiment and poly(A)’

was performed RNA prepared

using

the

from the LNCaP cell line. This analysis detected three major extension products of approximately 128, 132, and 234 nucleotides in length. As is also shown in Fig. 3, this same primer was used to generate a single-stranded DNA probe that was used in a simultaneous S, nuclease analysis of a sample of LNCaP RNA. As shown, this experiment detects two major products with a length of 128 and 139 nucleotides. There was close agreement between the primer extension and S, nuclease analyses for the bands labeled I and II in Fig. 3. However, there was no product detected in the S1 nuclease protection analysis that corresponded to band III produced in the primer extension assays. For this reason we undertook the construction and screening of a size-fractionated primer extended library using oligonucleotide L42B (see Fig. 3) as primer. This library (1 X lo6 recombinants) yielded six independent positive clones. The location of the 5’ termini of these clones is indicated in Fig. 4. The nucleotide sequence of all clones are consistent with initiation at the sites labeled I or II in Fig.

3. No

products

were

detected

that

< I ~

Base Pars

3

a

clone

200bp

1

Q

corresponded

~

B ,

P

L‘iza G-le 128 ““cleotldes 139 ““cieobdes II

FIG. 3. Primer extension and S, nuclease protection analysis of the human androgen receptor promoter. The positions of the oligonucleotide primers utilized are shown schematically. The experiment shown presents primer extension, Sanger dideoxy sequencing, and S1 nuclease protection all generated using the oligonucleotide primer L42B. To the left is shown the results of a primer extension experiment using L42B as primer after hybridization to 10 pg of poly(A)+ RNA prepared from the LNCaP cell line. To the right is shown an S1 nuclease protection assay in which the label in the single-stranded Ml3 probe originates from the oligonucleotide primer (L42B) end-labeled with polynucleotide kinase and [w”P]ATP. S1A was performed on 6 wg of poly(A)+ LNCaP RNA, and SI-B was performed on 30 c(g of total LNCaP RNA. The lanes A, C, G, and 7’ correspond to a sequencing ladder derived by priming DNA synthesis on the segment of genomic DNA depicted using primer L42B. N, S, B, and P represent the restriction endonuclease sites BstNI, SmaI, BstBI, and PstI boxed in schematic of genomic clone 12-2 BI shown in Fig. 2. The designations I and II, shown to the left, are aligned with the protected fragments detected in the S, nuclease protection shown to the right.

to longer extension products (i.e. III) although the synthesis and size fractionation of the primer-extended cDNA were performed in a manner that would include all three products (I, II, and III). Additional primer extension experiments performed with a second oligonucleotide primer (L2lB) located 330 nucleotides downstream of L42B generated extension products that were in agreement with the length of the two shorter extension products (128 and 132 nucleotides, bands I and II in Fig. 3) determined using oligonucleotide L42B as primer.’ These results are consistent with the presence of androgen receptor mRNA transcripts initiated at two predominant sites, labeled I and II in Figs. 3 and 4. Nucleotide Sequence Analysis of the Promoter and 5’-Flanking Segment of the Androgen Receptor Gene-To define the genetic elements responsible for controlling androgen receptor gene expression, we determined the nucleotide sequence of the 5’-untranslated segment flanking the androgen receptor open reading frame and a 300-base pair region 5’ to the transcription initiation site (Fig. 4). We assigned the numbering of this segment by designating the adenosine residue labeled II as nucleotide +l. This 5’-untranslated segment and the region 5’ to the initiation site contains several distinctive features.

Neither

TATA

nor

CCAAT

box

elements

are located

in close proximity to the site of transcription initiation. Spl motif is located at approximately -50 to -60 (relative

A to

13778

Androgen

Receptor Gene Promoter

N . :

S -

B

P :I 189 nucleotides

t

?

-489

- 404 - 367 FIG. 4. Nucleotide sequence of the promoter and 5’-flanking segment of the human androgen receptor. The nucleotide sequence of the 1400-base pair segment of genomic DNA 5’ to the initiator methionine of the androgen receptor protein is shown. The two principal sites of transcription initiation as determined by S1 nuclease protection (labeled I and II) are shown. The 5’ termini of the six primer extended clones are indicated by dots above the terminal nucleotide. The putative Spl-binding site is boxed while the region rich in adenosine and guanosine residues and the oligonucleotide sequence of the two primers, L42B and L21B, used for primer extension analysis are underlined.

the RNA cap site), and this same region also contains a segment (between nucleotides -186 and -229) composed of alternating adenine and guanine nucleotide residues. A single ATG triplet is present in the 1126 base pairs between the transcription initiation site and the initiator methionine of the androgen receptor protein. This ATG is located at position 688 relative to the transcription start site and begins a small open reading frame that is 24 nucleotides in length. The Same Androgen Receptor Gene Promoter Is Used in Diverse Tissues and Cell Lines-The similarity of androgen

receptor mRNA species visualized in the Northern analysis of RNA from different human cell lines and tissues suggested that androgen receptor mRNA has a similar structure in diverse tissues and cell lines that express the androgen receptor. To examine whether this similarity extended to the site of transcription initiation utilized in these different preparations, we designed an S1nuclease protection assay that detects androgen receptor mRNA molecules that begin at the initiation site identified in the LNCaP cell line. This analysis utilized a uniformly labeled RNA probe that is complementary to the sense strand of genomic clone 12-2 Bl and spans the site of transcription initiation (i.e. from the BstNI restriction site to the BstBI restriction site shown in Fig. 5). Following hybridization to this probe and digestion with S1 nuclease, transcripts initiated at the transcription initiation site employed in the LNCaP cell line generate a protected fragment 189 nucleotides in length. An analysis of this type was performed on samples of RNA from human tissues and cells and is shown in Fig. 5, a and b. A protected fragment of the expected size was detected in the LNCaP cell line and confirms the assignment of II (in Fig. 4) as the principal site of transcription initiation. This same fragment is detected in all of the samples tested that express androgen receptor mRNA,

FIG. 5. S1 nuclease protection analysis demonstrates the usage of a common transcription initiation site within the androgen receptor gene in diverse human tissues and cell lines. a, samples of total RNA prepared from different human cell lines and tissues were hybridized to a uniformly labeled antisense RNA probe that spans the transcription initiation site (probe A: which extends from the restriction endonuclease cleavage site BstNI to BstBI in the genomic clone 12-2 Bl, as indicated schematically), digested with S1 nuclease, and electrophoresed on a 6% polyacrylamide gel containing 8 M urea as described under “Materials and Methods.” b, a short exposure of the two right lanes in panel a. Sizes were estimated by comparison with end-labeled DNA markers (i.e. HpaII-digested fragments of BSKSM13+). N, S, B, and P represent, respectively, the restriction endonuclease sites BstNI, SmaI, BstBI, and PstI boxed in the schematic of genomic clone 12-2 BI shown in Fig. 2. c, samples of total RNA prepared from LNCaP cells, genital skin fibroblasts (0, and T47D, MCF-7, and PC-3 cells were hybridized to the same RNA probe @robe A) used in the experiment shown in panels a and b, digested with S, nuclease, and analyzed on a denaturing polyacrylamide gel. d, identical samples to those used in the experiment shown in panel c were hybridized to a uniformly labeled RNA probe @robe B) complementary to a segment of the androgen receptor cDNA that encodes a portion of both the DNAand hormone-binding domains of the receptor protein (nucleotides 1850-2563; see Fig. 2), digested with S1 nuclease and electrophoresed on denaturing polyacrylamide gels as described under “Materials and Methods.”

including samples from human prostate, human testes, human genital skin fibroblasts, and two human breast carcinoma cell lines. By contrast, this fragment is not detected in the prostate carcinoma cell line, PC-3, which does not express androgen receptor mRNA.3 Although these data indicate that the same initiation site is employed in each of these human cell lines and tissues, these experiments do not examine the possibility that mRNA encoding functional androgen receptor protein is produced by initiation at other promoters. To address this issue, we performed S1protection assays using two RNA probes that were uniformly labeled with [32P]UTP of the same specific activity. The first probe (probe A) was the same as that employed for the experiment depicted in Fig. 5, a and b (i.e. probe A spans the transcription initiation site). The second probe (probe B) hybridizes to a segment of the androgen receptor cDNA 3 Tilley, W. D., Wilson, C. M., (1990) Cancer Res., in press.

Marcelli,

M., and

McPhaul,

M. J.

Androgen

13779


(nucleotides 1850-2563 (11)) that encodes a segment of the DNA-binding and hormone-binding domains (i.e. the region spanned by probe 3 in Fig. 2). These two probes were used to assay the levels of androgen receptor mRNA transcripts containing these segments in RNA samples prepared from the LNCaP and PC-3 prostate carcinoma cell lines, the T47D and MCF-7 breast cancer cell lines, and genital skin fibroblasts. As shown in Fig. 5, c and d, a single protected fragment of the predicted size was obtained for each sample using both probes (i.e. approximately 189 nucleotides for probe A, and 713 nucleotides for probe B). Neither protected fragment was detected using these probes in assays of the RNA samples prepared from the PC-3 cell line which lacks detectable androgen receptor mRNA. As the two probes were synthesized using radiolabeled UTP of identical specific activity, the intensities of the bands detected by autoradiography reflect the relative abundance of androgen receptor mRNA transcripts complementary to these segments. Our results suggest that the bulk of transcripts that contain the coding segment of the androgen receptor are initiated at the same promoter in genital skin fibroblasts, the LNCaP prostate carcinoma cell line, and the T47D and MCF-7 breast cancer cell lines. In separate experiments we have found that much of the androgen receptor mRNA present in samples prepared from human prostate is derived from this same promoter.’ DISCUSSION

Similar species of androgen receptor mRNA are detected in RNA samples prepared from two human androgen target tissues (testes and prostate), three human cell lines (LNCaP, T47D, MCF-7), and genital skin fibroblasts. The human prostate carcinoma cell line LNCaP, which expresses high levels of androgen receptor mRNA and protein, was utilized to characterize the promotor for the androgen receptor gene. Initial experiments using Northern analysis with genomic DNA fragments suggested that the transcription initiation site is localized to a region approximately 1.1 kilobases upstream of the initiation methionine of the androgen receptor open reading frame. The precise localization of the transcription initiation site required primer extension, S1 nuclease analyses, and isolation of cDNA clones from a primer extended library. While primer extension identified three principal elongation products of 128, 132, and 234 nucleotides in length, S1 nuclease mapping with a single-stranded end-labeled Ml3 probe synthesized using the same oligonucleotide primer (L42B) as that employed for primer extension analysis identified only two major products of 128 and 139 nucleotides in length. Thus, good agreement exists between the shortest primer extension and S1 nuclease protection products and fixed one site of transcription initiation at approximately 128 nucleotides upstream of the oligonucleotide primer, L42B. However, agreement between the two larger primer extension products and the larger S, protection product was less convincing and raised the possibility that other androgen receptor mRNA transcripts contain a 5’-untranslated exon. We examined this possibility by preparing and screening a primer extended library using the same oligonucleotide primer employed in the primer extension analysis depicted in Fig. 3. Notably, the method of cDNA synthesis, fractionation, and screening was designed to permit the isolation of all of the products visualized in Fig. 3. Despite this, the nucleotide sequence analysis of six independent clones isolated from this primer extended library revealed structures consistent with initiation only at the sites detected by S1 nuclease protection, that is, sites I and II (in Figs. 3 and 4). Additional primer extension experiments with a primer (L21B) located 330

nucleotides 3’ to L42B supported the conclusion that the major site of transcription initiation is located in the region of sites I and II defined by S1 protection.’ Furthermore, S1 nuclease protection studies are consistent with the adenosine residue designated II (Fig. 4) as the principal site of transcription initiation, although some transcripts may also be initiated as the cytosine residue designated I (Fig. 4) as well. A similar heterogeneity of transcription initiation has been described for other genes which lack typical “TATA” or “CAAT” box motifs (17). These results also suggest that the longest primer extension product (labeled as III in Fig. 3) is derived from cross-hybridization with another mRNA and not derived from hybridization to androgen receptor mRNA. This conclusion is reinforced by the detection of primer extension products similar in size to band III in analyses performed using RNA prepared from cell lines (e.g. the PC-3 cell line) known not to express androgen receptor mRNA.2 The nucleotide sequence of the region surrounding the promoter of the human androgen receptor indicates that it is organized in a fashion analogous to the class of genes that contain elements rich in guanosine and cytosine found in the regulatory elements for many genes (17). This same type of organization has been described for the chicken progesterone receptor (18) and differs from that reported for the human estrogen receptor which contains both TATA and CCAATlike sequences upstream of the transcription initiation site (19). In this respect, the 5’-flanking segments of the chicken progesterone receptor and the human androgen receptor genes both contain GC-rich elements, including a Spl binding motif (GGGCGG) characteristic of a “housekeeping” promoter and purine-rich elements composed of alternating adenosine and guanosine residues. While the functional significance of these latter segments is not known, these similarities emphasize that the progesterone and androgen receptor gene promoters are similar in their organization, a finding that is consistent with the similarities evident between the genomic organization (3, 18, 20) and the predicted protein sequences (11, 2124) of the human androgen and chicken progesterone receptors. Our studies also confirm that the 5’untranslated region of the human androgen receptor mRNA contains a small open reading frame that would encode a protein with a molecular weight of 947. An open reading frame of similar length and composition is present in the 5’-untranslated segment of the rat androgen receptor mRNA (25) as well, suggesting that this segment may encode a functional product of the androgen receptor locus. The function of this peptide is unknown. The detection of similar androgen receptor mRNA species by Northern analysis in RNA samples prepared from human tissues and cell lines suggested a common structure of the androgen receptor mRNA. The definition of the androgen receptor gene promoter in the LNCaP prostatic carcinoma cell line allowed us to examine whether this promoter was employed in other human tissues. We found that the activity of this promoter correlates with the level of androgen receptor mRNA expression detected in each of the tissues and cell lines that we have examined. Quantitative S1 nuclease protection assays using uniformly labeled RNA probes that recognize portions of the androgen receptor coding region and that span the promoter region establish that much of the mRNA encoding the androgen receptor can be accounted for by initiation at this single promoter in the prostate carcinoma cell line LNCaP, genital skin fibroblasts, the human prostate, and the breast cancer cell lines MCF-7 and T47D. Acknowledgments-We Hennis for expert technical

thank Judith and secretarial

A. Gruber assistance,

and Brenda respectively.

H.

13780

Androgen


REFERENCES 1. Wilson, J. D., Griffin, J. E., Leshin, M., and MacDonald, P. C. (1983) in The Metabolic Basis of Inherited Disease (Stambury, J. B., Wyngaarden, J. G., Fredrickson, D. S., et al, eds) 5th Ed, pp. 1001-1025, McGraw-Hill Book Co., New York 2. Brown, T. R., Lubahn, D. B., Wilson, E. M., Joseph, D. R., French, F. S., and Migeon, C. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 86,8151-8155 3. Lubahn, D. B., Brown, T. R., Simental, J. A., Higgs, H. N., Migeon, C. J., Wilson, E. M., and French, F. S. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,9534-9538 4. Marcelli, M., Tilley, W. D., Wilson, C. M., Wilson, J. D., Griffin, J. E., McPhaul, M. J. (1990) J. Clin. Znuest. 85, 1522-1528 5. George, F. W., and Peterson, K. G. (1988) Biol. Reprod. 39, 536539 6. Takane, K. K., George, F. W., and Wilson, J. D. (1990) Am. J. Physiol. 258, E45-E50 7. Cunha, G. R., Chung, L. W., Shannon, J. M., Taguchi, O., and Fujii, H. (1983) Recent Prog. Horm. Res. 39,559-597 8. Chirgwin, J. M., Przybyla, A. E., McDonald, R. J., and Rutter, W. J. (1979) Biochemistry l&5294-5299 9. Aviv, H., and Leder, P. (1972) Proc. Natl. Acad. Sci. U. S. A. 69, 1408-1412 10. Feinberg, A. P., and Vogelstein, B. (1983) Anal. Biochem. 132, 6-13 11. Tilley, W. D., Marcelli, M., Wilson, J. D., and McPhaul, M. J. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 327-331 12. Burke, J. F. (1984) Gene (Am&.) 30,63-68 13. Gubler, U., and Hoffman, B. J. (1983) Gene (Amst.) 25,263-269 14. Huynh, T. V., Young, R. A., and Davis, R. W. (1985) in DNA

MATERIALS

*ND

METHODS

15. 16.

17. 18.

19. 20.

21.

22. 23.

24.

25.

Cloning (Glover, D. M., ed) Vol. I, pp. 49-78, RL Press, Oxford Sanger, F., Coulson, A. R., Barrell, B. G., Smith, A. J. H., and Roe, B. A. (1980) J. Mol. Biol. 143, 161-178 Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY Maniatis, T., Goodbourn, S., and Fischer, J. A. (1987) Science 236, 1237-1245 Huckaby, C. S., Conneely, 0. M., Beattie, W. G., Dobson, A. D. W., Tsai, J.-J., and 0-Malley, B. W. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,8380-8384 Ponglikitmongkol, M., Green, S., and Chambon, P. (1988) EMBO J. 7,3385-3388 Kuiper, G. G. J. M., Faber, P. W., van Rooij, H. C. J., van der Korput, J. A. G. M., Ris-Stalpers, C., Klaassen, P., Trapman, J., and Brinkman, A. S. (1989) Mol. Endocrinol. 2, Rl-R4 Lubahn, D. B., Joseph, D. R., Sar, M., Tan, J., Higgs, H. N., Larson, R. E., French, F. S., and Wilson, E. M. (1988) Mol. Endocrinol. 2, 1265-1275 Chang, C., Kokontis, J., and Liao, S. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,7211-7215 Trapman, J., Klaassen, P., Kuiper, G. G. J. M., van der Korput, J. A. G. M., Faber, P. W., van Rooij, H. C. J., Geurts van Kessel, A., Voorhorst, M. M., Mulder, E., and Brinkmann, A. 0. (1988) Biochem. Biophys. Res. Commun. 153,241-248 Misrahi, M., Atger, M., d’Aurio1, L., Loosfelt, H., Meriel, C., Fridlansky, F., Guiochon-Mantel, A., Galibert, F., and Milgrom, E. (1987) Biochem. Biophys. Res. Commun. 143,740-748 Tan, J., Joseph, D. R., Quarmby, V. E., Lubahn, D. B., Sar, M., French, F. S., and Wilson, E. M. (1988) Mol. Endocrinol. 2, 1276-1285

Androgen
