281 Down but not out? A novel protein isoform of the estrogen receptor
is expressed in the estrogen receptor
knockout mouse

M Koš, S Denger, G Reid, K S Korach1 and F Gannon European Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117 Heidelberg, Germany 1Receptor Biology Section, Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences/NIH, Research Triangle Park, North Carolina 27709, USA

(Requests for offprints should be addressed to F Gannon; Email: [email protected])

Abstract

The mouse knockout of the estrogen receptor α (ERα) gene, known as αERKO, has been extensively used for several years to study the role and function of ERα. Residual estradiol binding capacity in uterine tissue of 5–10% raised doubts if this knockout is a genuine null mutation of ERα. Although alternatively spliced ERα mRNA variants in the αERKO mouse were reported previously, the corresponding protein isoforms have not been detected to date. Here we show that a variant ERα protein, 61 kDa in size, is expressed in the uterine tissue of αERKO mice as a result of an alternative splicing. The transactivation capability of this protein is cell dependent and can be as high as 75% of the wild type ERα. Journal of Molecular Endocrinology (2002) 29, 281–286

Introduction (Couse & Korach 1999 and references therein). However, the residual estradiol binding activity in Estrogens play an important role in the develop- uterine tissue in conjunction with the fact that ment and physiological processes in vertebrates. the ER
gene was disrupted by an insertion of the Although they are primarily known as female sex neomycin gene into the first coding exon raised the hormones controlling female sex determination, possibility that alternatively spliced ER
variant(s) reproduction and pregnancy, they are also involved might be present in the
ERKO mice. Two in the development and maintenance of male alternatively spliced transcripts of the disrupted reproductive organs (Hess et al. 1997) and in other ER
gene, called E1 and E2, were described in physiological processes involving the liver, fat, bone
ERKO mice (Couse et al. 1995). The E2 mRNA metabolism and cardiovascular and neuronal contains a frame shift generating several stop activity (Norman & Litwack 1987, George & codons after the first 41 amino acids of ER
. The Wilson 1988, Auchus & Fuqua 1994). The role of E1 mRNA is generated by an alternative splicing, estrogens in several pathological processes such as which utilizes a cryptic donor splice site at the osteoporosis (Horowitz 1993), breast and endo- beginning of the neomycin gene and the endogen- metrial cancers (Henderson et al. 1988), arterio- ous acceptor splice site of coding exon 2 (Fig. 1). sclerosis and Alzheimer’s disease (Auchus & Fuqua The ER
protein isoform resulting from this 1994) is also well established. transcript would have amino acids 92 to 155 of The generation of the estrogen receptor
(ER
) wild type ER
replaced by seven amino acids from knockout mouse (
ERKO) (Lubahn et al. 1993) was the neomycin insert. This replacement would affect highly informative towards understanding the AF-1 transactivation function but not the DNA or mechanisms of estrogen action in the complexity of ligand binding domain, which contains the AF-2 the whole organism. An impressive and con- transactivation function. Such a protein isoform tinuously growing body of information on ER
has recently been observed in mouse tissues functions has been obtained from
ERKO mice (Pendaries et al. 2002). The discovery of estrogen

Journal of Molecular Endocrinology (2002) 29, 281–286 Online version via http://www.endocrinology.org 0952–5041/02/029–281 © 2002 Society for Endocrinology Printed in Great Britain

Downloaded from Bioscientifica.com at 09/25/2021 07:14:12PM via free access 282 M KOŠ and others · ER
expression in the
ERKO mouse

receptor  (ER) (Kuiper et al. 1996) brought directionally cloned into the BamHI and EcoRI another potential explanation of the observed sites of the pSG5 vector. The construct was binding activity. However, overlapping but distinct sequenced. phenotypes of ER knockout (Krege et al. 1998) The reporter plasmid ERE-tk-Luc containing and ER
 double knockout (Couse et al. 1999) the estrogen response element upstream of the mice do not fully support this explanation. More thymidine kinase (tk) promoter and firefly luciferase recently, ER
,ER and ER
ER knockout mice encoding sequence was kindly provided by P Webb were generated by complete deletion of the second and has been described (Paech et al. 1997) coding exon using loxP sites and crosses with previously. The pSG5-Renilla vector was kindly CMV-Cre mice (Dupont et al. 2000). These new provided by M Hentze and contains the Renilla knockout mice share many phenotypes with the luciferase coding sequence cloned between the original ER knockouts. SmaI and BamHI sites of pSG5. In this report, the presence of an ER
protein isoform, which might correspond to a putative Protein isolation and Western blot analysis translational product of the E1 transcript isolated previously by Couse et al. (1995) in
ERKO uterine The uteri were powdered in liquid nitrogen using tissue is shown. The capability of this E1 isoform a pestle and mortar. The powdered tissue was to activate a promoter containing an estrogen transferred into approximately 5 volumes of the response element (ERE) was evaluated by transient buffer containing 50 mM Tris, pH 7·5, 1 mM transfections in different cell lines. EDTA, 1 mM dithiothreitol, 10% glycerol and the protease inhibitor cocktail Complete mini (Roche). After 10 min incubation on ice, the suspension was Materials and methods pelleted in a table minicentrifuge and the supernatant was transferred into a new tube. Plasmids and constructs The protein concentration was determined by the Bradford assay (BioRad, Munich, Germany). The The mouse ER
expressing vector pSG5-mER
protein extracts were stored at 70 C. Aliquots of was constructed by subcloning the ER
coding 50 µg total proteins were analyzed by Western sequence from pMT2-MOR (kindly provided by blotting using the following antibodies: rabbit poly- M G Parker) into the EcoRI site of the pSG5 vector clonal antibody MC-20 (SantaCruz, Heidelberg, (Green et al. 1988). To create the pSG5-E1 vector Germany) raised against the last 20 amino acids of expressing the E1 ER
, the ER
coding sequences mouse ER
; rabbit polyclonal antibody H-184 upstream and downstream of the deletion were (SantaCruz) raised against the 2–185 amino acids PCR amplified from the pSG5-mER
plasmid region of human ER
; and rat monoclonal H222 using the following primer pairs: PF1 (5
-TCGAA antibody raised against the ligand binding domain TTCATGACCATGACCCTTCACACCAAAGC-3
) of human ER
(Greene et al. 1984). The H222 and PR1 (5
-ATTAGACCCGGTAGAATTCCT antibody was kindly provided by G L Greene GCAAGCGGCCGCCTCCGACCCGGGGCCG (University of Illinois, Urbana, IL, USA). The TAG-3
); and PF2 (5
-GCTTGCAGGAATTCTA rabbit polyclonal anti--actin antibody (Sigma) CCGGGTCTAATTCTGACAATCGACGCCAG was used to control for equal loading and degra- AATGGCCGAG-3
) and PR2 (5
-GCGAATTC dation. The secondary horseradish peroxidase GGGGAGCCTGGGAGCTCTCAGAT-3
). The conjugates were purchased from DiaNova PF1 spans the translation start site of ER
and (Hamburg, Germany). contains a BamHI site at its 5
end. The PR2 primer spans the termination codon of ER
and contains an EcoRI site. The PR1 and PF2 primers Cell culture, transient transfections and dual are partially overlapping and contain sequences luciferase assays spanning the deletion. The two amplified products HeLa, Ishikawa and NIH-3T3 cell lines were were isolated from the gel and were used as maintained in DMEM (Invitrogen) supplemented templates in a new PCR amplification using, again, with 10% fetal bovine serum (FBS) (Invitrogen), primers PF1 and PR2. The final PCR product was penicillin (100 U/ml; Invitrogen), streptomycin

Journal of Molecular Endocrinology (2002) 29, 281–286 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 07:14:12PM via free access ER
expression in the
ERKO mouse · M KOŠ and others 283

Figure 1 (A) Western blot analysis of uterine tissue using the antibody MC-20 recognizing the C-terminus of ERα. An arrow denotes the putative E1 ERα protein. Lanes 1–6 were loaded with samples from different αERKO animals. (B) Western blot analysis of uterine tissue using the antibody H-184 recognizing the N-terminus of ERα and anti-β-actin antibody. An arrow denotes the putative E1 ERα protein. (C) Schematic representation of the alternative splicing generating the E1 ERα transcript. Numbers below represent amino acid residues of the wild type ERα. White boxes represent exons of the ERα gene; grey boxes correspond to the inserted Neomycin sequence. Neo, neomycin.

(100 µg/ml; Invitrogen) and glutamin (2 mM, (Berthold Technologies, Bad Wildbad, Germany).  Invitrogen) at 37 Cina5%CO2 incubator. The All experiments were performed at least three times cells were split 48 h before transfection into 24-well in triplicate. plates and the next day were transferred to the medium without phenol red and supplemented with 2·5% charcoal striped FBS. Transfections Results were performed using Fugene6 reagent (Roche) following the recommendations of the manufac- A novel ER
protein isoform is expressed in turer. In each transfection, 200 ng ERE-tk-Luc
ERKO uteri reporter construct, 50 ng of either pSG5-mER
or In order to investigate if ER
protein isoform(s) are pSG5-E1 vector and 5 ng pSG5-Renilla vector expressed in
ERKO mice, uterine samples from were used. Immediately after transfection the several animals of different ages were analyzed by vehicle, 17-estradiol (108 M) or ICI 182,780 Western blotting. A protein of approximately (108 M) were added to the media. Twenty hours 61 kDa was detected by the MC-20 antibody that after transfection cells were lysed in passive lysis recognizes the last 20 amino acids of the mouse buffer (Promega) and dual luciferase assays ER
receptor protein (Fig. 1A). This antibody (Promega) were performed following the recom- produces an unspecific doublet at the 55 kDa mendations of the manufacturer. Ten microliters of range. The other proteins of approximately 52, 40 cell lysate were transferred into a 96-well plate and and 37 kDa are degradation products of ER
,as the light emission was measured for 10 s with 2 s they can be partially prevented by inclusion of delay after the injection of 50 µl luciferase assay iodoacetamide or molybdate in the extraction reagent or Stop&Glow reagent on an EG&G buffer (Horigome et al. 1987 and our observation). Berthold Microplate Luminometer LB 96V Nevertheless, the presence of the same degradation www.endocrinology.org Journal of Molecular Endocrinology (2002) 29, 281–286

Downloaded from Bioscientifica.com at 09/25/2021 07:14:12PM via free access 284 M KOŠ and others · ER
expression in the
ERKO mouse

pattern in the
ERKO samples indicates that the observed 61 kDa protein is likely to be an ER
isoform. The same protein was also detected by the H-184 antibody specific to the N-terminus of ER
(Fig. 1B). However, the H222 monoclonal anti- body, which is specific to the ligand binding domain of the human ER
but cross-reacts with murine ER
(Greene et al. 1984, Horigome et al. 1987) failed to detect any ER
protein in
ERKO animals by Western blots (data not shown). Although a ligand binding peak from
ERKO uterine tissue analyzed by sucrose gradient centrifugation was shifted by H222 (Couse et al. 1995), a small fraction of animals did not express this protein (
ERKO sample 2 in Fig. 1A). In total, uterine tissue from 13 animals was analyzed and, with the exception of two cases, the expression of this 61 kDa ER
protein was detected. The levels of expression estimated from Western blots with H-184 and anti--actin antibodies (Fig. 1B) varied between animals. Figure 2 (A) The transactivation capabilities of E1 ERα The size and the fact that the 61 kDa protein isoform and wild type (wt) ERα were compared in was detected by both N-terminal-specific and transient transfection using three different cell lines HeLa, NIH-3T3 (3T3) and Ishikawa. The activity of the C-terminal-specific anti-ER
antibodies indicate wild type ERα in the presence of estradiol (E2) was set that this protein isoform might be a putative to 100%. The activity was abolished by ICI 182,780 translation product of the E1 ER
transcript (ICI). (B) Western blot analysis (H-184 antibody) of the observed previously in
ERKO uteri (Couse et al. expression of wild type and E1 ERα in transiently 1995). The transactivation capability of E1 ER
transfected HeLa cells. was previously characterized in COS-1 cells to be 25–30% of the wild type level (Couse et al. 1995). of
ERKO animals. This protein might corres- However, as ER
acts in COS-1 cells mainly pond to the putative E1 ER
described previously. through the AF-1 (El Tanani & Green 1997) we The activity of the E1 ER
variant is cell evaluated the capability of the E1 ER
variant to dependent and would have the capability to transactivate a promoter containing an estrogen significantly contribute to ER
signaling in certain response element by transient transfections in other cell types or tissues. cell lines: HeLa, NIH-3T3 and an ER
negative variant of the Ishikawa cell lines (Fig. 2A). The induction of the luciferase reporter gene expression Discussion by E1 ER
compared with wild type ER
varied between 30% (Ishikawa cells) and 75% (HeLa When the
ERKO mouse was created the current cells). The activity was completely abolished by technology utilized at the time was the insertional antiestrogen ICI 182,780 (ICI). The levels of disruption of genes. The 1·8 kbp neomycin expression of both E1 and wild type ER
proteins encoding sequence was inserted into the NotI site in transiently transfected HeLa cells were compar- of the first coding exon of the murine ER
gene able (Fig. 2B). Both protein isoforms were also (Lubahn et al. 1993). As all the coding exons of the down-regulated by estradiol and ICI treatment as ER
gene are still present in the genome of previously reported (Read et al. 1989, Gyling et al.
ERKO mouse, the possibility of alternative 1990, Gibson et al. 1991, Dauvois et al. 1992, splicing that generates mRNA variants encoding Wijayaratne & McDonnell 2001) (Fig. 2B). truncated or deletion versions of ER
protein In summary, these results indicate that a novel exists. Such alternative splicing was observed, for ER
protein isoform is expressed in the majority example, in a knockout mouse in which the

Journal of Molecular Endocrinology (2002) 29, 281–286 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 07:14:12PM via free access ER
expression in the
ERKO mouse · M KOŠ and others 285 transforming growth factor
gene was disrupted by AF-2 activities are cell type and promoter insertion (Luetteke et al. 1993). Similarly, alterna- dependent (Tora et al. 1989, Berry et al. 1990, tively spliced ER
mRNAs, named E1 and E2, Metzger et al. 1995, Flouriot et al. 2000). As the E1 were also observed in
ERKO mice (Couse et al. ER
has impaired AF-1 function but conserved 1995). While the E2 mRNA encodes only the AF-2 function, it can be anticipated that only small first 41 amino acids of ER
, the E1 mRNA differences in transactivation capabilities between encodes a potential ER
isoform with a deletion E1 and wild type ER
would be observed in cells in the B domain containing the AF-1 function. with strong AF-2 context such as HeLa (Tora et al. However, the corresponding E1 ER
protein has 1989). The cell context most likely depends on the been detected in
ERKO tissues only recently blend of AF-1 or AF-2 specific coactivators and/or (Pendaries et al. 2002). The fact that the H222 corepressors present in the cell. Therefore, it is antibody was originally used to analyze the difficult to predict the activity of E1 ER
in various presence of ER
protein isoforms in
ERKO cell types and especially in tissues. The high uterine tissue by Western blot techniques (Couse similarity in the reproductive and general pheno- et al. 1995) might explain the failure to detect the types of
ERKO and the recently generated ER
E1 ER
isoform which is expressed in
ERKO at knockout mice (Dupont et al. 2000) might indicate lower levels than ER
in wild type mice. In fact, that E1 ER
plays only a minor role in the the mRNA for the E1 variant was only detected by physiology of the reproductive tract in
ERKO RT-PCR techniques. General experience suggests mice. Interestingly, use of the
ERKO mice has that although the H222 antibody readily interacts shown a lack of ligand independent growth factor with ER
from various species (Greene et al. 1984) signaling in vivo with either epidermal growth factor in native conditions, such as in sucrose gradients (Curtis et al. 1996) or insulin-like growth factor-I where E1 binding was demonstrated, it may not (Klotz et al. 2002), which would be a clear perform well under denaturing conditions, such as demonstration that ER
AF-1 contributes to the Western blot involving low amounts of protein. In cross talk signaling in vivo. contrast, the rabbit polyclonal antibodies used in To conclude, our demonstration that a novel this study readily detect ER
in either murine or ER
protein isoform is expressed in the
ERKO human tissues in Western blots, therefore allowing mouse indicates that data obtained from studies the detection of E1 ER
in
ERKO uteri. using
ERKO mice should be further considered The penetrance of the E1 variant expression with respect to AF-1 and AF-2 functions. However, differed between animals and varied from the variable penetrance of this phenotype in undetectable (in two of thirteen animals tested in
ERKO animals means that ER
isoform total) to varying levels in most of the animals. The status can serve as a model to study the AF-1 and influence of genetic background or generation AF-2 functions within individual tissues in the number on the penetrance of a mutant phenotype mouse. has been documented in several cases and might also account for the observed differences in E1 ER
expression in
ERKO (Dunn et al. 1997, References Herrera et al. 1999, Wawersik et al. 1999, Kume et al. 2000, Nadeau 2001). Nevertheless, incomplete Auchus RJ & Fuqua SAW 1994 The oestrogen receptor. In Bailliere’s penetrance might be an advantage, as approxi- Clinical Endocrinology and Metabolism: Hormones, Enzymes and Receptors, mately 15% of
ERKO animals seem to represent pp 433–449. London: Baillière Tindall. Berry M, MetzgerD&ChambonP1990 Role of the two activating genuine ER
-null mutants while the rest can be domains of the oestrogen receptor in the cell-type and promoter- regarded as AF-2 hypomorphic AF-1 knockouts. context dependent agonistic activity of the antioestrogen Thus the
ERKO mouse might represent a good 4-hydroxytamoxifen. EMBO Journal 9 2811–2818. Couse JF & Korach KS 1999 Estrogen receptor null mice: what model system to investigate the comparative have we learned and where will they lead us. Endocrine Reviews 20 physiological roles of AF-1 and AF-2 in ER
358–417. function. Couse JF, Curtis SW, Washburn TF, Lindzey J, Golding TS, The dramatic difference in the activity of E1 LubahnDB,SmithiesO&Korach KS 1995 Analysis of transcription and estrogen insensitivity in the female mouse after ER
obtained in transfection experiments using targeted disruption of the estrogen receptor gene. Molecular various cell lines is not surprising, as AF-1 and Endocrinology 9 1441–1454. www.endocrinology.org Journal of Molecular Endocrinology (2002) 29, 281–286

Downloaded from Bioscientifica.com at 09/25/2021 07:14:12PM via free access 286 M KOŠ and others · ER
expression in the
ERKO mouse

Couse JF, Hewitt SC, Bunch DO, Sar M, Walker VR, Davis BJ & Klotz DM, Hewitt SC, Ciana P, Raviscioni M, Lindzey JK, Foley J, Korach KS 1999 Postnatal sex reversal of the ovaries in mice Maggi A, DiAugustine RP & Korach KS 2002 Requirement of lacking estrogen receptors
and . Science 286 2328–2331. estrogen receptor-alpha in insulin-like growth factor-I (IGF-I)- Curtis SW, Washburn T, Sewall C, DiAugustine R, Lindzey J, induced uterine responses and in vivo evidence for IGF-I/estrogen Couse JF & Korach KS 1996 Physiological coupling of growth receptor cross-talk. Journal of Biological Chemistry 277 8531–8537. factor and steroid receptor signaling pathways: estrogen receptor Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, knockout mice lack estrogen-like response to epidermal growth Sar M, Korach KS, Gustafsson JA & Smithies O 1998 factor. PNAS 93 12626–12630. Generation and reproductive phenotypes of mice lacking estrogen Dauvois S, Danielian PS, WhiteR&Parker MG 1992 Antiestrogen receptor . PNAS 95 15677–15682. ICI 164,384 reduces cellular estrogen receptor content by Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S & Gustafsson JÅ increasing its turnover. PNAS 89 4037–4041. 1996 Cloning of a novel receptor expressed in rat prostate and Dunn NR, Winnier GE, Hargett LK, Schrick JJ, Fogo AB & Hogan ovary. PNAS 93 5925–5930. ffi BL 1997 Haploinsu cient phenotypes in Bmp4 heterozygous null KumeT,DengK&Hogan BL 2000 Minimal phenotype of mice mice and modification by mutations in Gli3 and Alx4. homozygous for a null mutation in the forkhead/winged helix Developmental Biology 188 235–247. gene, Mf2. Molecular and Cellular Biology 20 1419–1425. Dupont S, Krust A, Gansmuller A, Dierich A, ChambonP&Mark Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS & M 2000 Effect of single and compound knockouts of estrogen Smithies O 1993 Alteration of reproductive function but not receptors alpha (ERalpha) and beta (ERbeta) on mouse prenatal sexual development after insertional disruption of the reproductive phenotypes. Development 127 4277–4291. mouse estrogen receptor gene. PNAS 90 11162–11166. El Tanani MK & Green CD 1997 Two separate mechanisms for ff ligand-independent activation of the estrogen receptor. Molecular Luetteke NC, Qiu TH, Pei er RL, Oliver P, SmithiesO&LeeDC 1993 TGF alpha deficiency results in hair follicle and eye Endocrinology 11 928–937. Cell Flouriot G, Brand H, Denger S, Metivier R, Kos M, Reid G, abnormalities in targeted and waved-1 mice. 73 263–278. Sonntag-BuckV&GannonF2000Identification of a new Metzger D, Ali S, Bornert JM & Chambon P 1995 Characterization isoform of the human estrogen receptor-alpha (hER-alpha) that is of the amino-terminal transcriptional activation function of the encoded by distinct transcripts and that is able to repress human estrogen receptor in animal and yeast cells. Journal of hER-alpha activation function 1. EMBO Journal 19 4688–4700. Biological Chemistry 270 9535–9542. George FW & Wilson JD 1988 Sex determination and sex Nadeau JH 2001 Modifier genes in mice and humans. Nature Reviews. differentiation. In The Physiology of Reproduction, pp 3–26. Eds E Genetics 2 165–174. Knobil, JD Neil, LL Ewing, GS Green-Wals, CL Market & DW Norman AW & Litwack G 1987 Estrogens and progestins. In Pfaff. New York: Raven Press. Hormones, pp 550–560. Ed. G Litwack. London: Academic Press. Gibson MK, Nemmers LA, Beckman WC Jr, Davis VL, Curtis SW Paech K, Webb P, Kuiper GG, Nilsson S, Gustafsson JÅ, Kushner & Korach KS 1991 The mechanism of ICI 164,384 anti- PJ & Scanlan TS 1997 Differential ligand activation of estrogen estrogenicity involves rapid loss of estrogen receptor in uterine receptors ER
and ER at AP1 sites. Science 277 1508–1510. tissue. Endocrinology 129 2000–2010. Pendaries C, Darblade B, Rochaix P, Krust A, Chambon P, Korach Green S, Issemann I & Sheer E 1988 A versatile in vivo and in vitro KS, BayardF&Arnal JF 2002 The AF-1 activation-function of eukaryotic expression vector for protein engineering. Nucleic Acids ERalpha may be dispensable to mediate the effect of estradiol on Research 16 369. endothelial NO production in mice. PNAS 99 2205–2210. Greene GL, Sobel NB, King WJ & Jensen EV 1984 Immunochemical Read LD, Greene GL & Katzenellenbogen BS 1989 Regulation of studies of estrogen receptors. Journal of Steroid Biochemistry 20 51–56. estrogen receptor messenger ribonucleic acid and protein levels in Gyling M, LeclercqG&HeusonJC1990Estrogenic and human breast cancer cell lines by sex steroid hormones, their antiestrogenic down-regulation of estrogen receptor levels: antagonists, and growth factors. Molecular Endocrinology 3 295–304. ff evidence for two di erent mechanisms. Journal of Receptor Research Tora L, White R, Brou C, Tasset D, Webster N, Scheer E & 10 217–234. Chambon P 1989 The human estrogen receptor has two Henderson BE, RossR&Bernstein L 1988 Estrogen as a cause of nonacidic transcriptional activation functions. Cell 59 477–487. human cancer: the Richard and Hinda Rosenthal Foundation Wawersik S, Purcell P, Rauchman M, Dudley AT, Robertson EJ & Cancer Research award lecture. 48 246–253. Maas R 1999 BMP7 acts in murine lens placode development. Herrera E, SamperE&Blasco MA 1999 Telomere shortening in Developmental Biology 207 176–188. mTR/embryos is associated with failure to close the neural Wijayaratne AL & McDonnell DP 2001 The human estrogen tube. EMBO Journal 18 1172–1181. receptor-alpha is a ubiquitinated protein whose stability is affected Hess RA, Bunick D, Lee KH, Bahr J, Taylor JA, Korach SK & differentially by agonists, antagonists, and selective estrogen Lubahn DB 1997 A role for oestrogens in the male reproductive receptor modulators. Journal of Biological Chemistry 276 system. Nature 390 509–512. 35684–35692. Horigome T, Golding TS, Quarmby VE, Lubahn DB, McCarty K & Korach KS 1987 Purification and characterization of mouse uterine estrogen receptor under conditions of varying hormonal status. Endocrinology 121 2099–2111. Horowitz MC 1993 Cytokines and estrogen in bone: anti-osteoporotic Received 2 July 2002 effects. Science 260 626–627. Accepted 7 August 2002

Journal of Molecular Endocrinology (2002) 29, 281–286 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/25/2021 07:14:12PM via free access