This application is related to U. S. Provisional applications 60/053,869 and 60/054,210, which are herein incorporated by reference.

BACKGROUND OF THE INVENTION A. Estrogen Estrogens are a class of naturally occurring steroid hormones which are produced in the ovaries and other tissues of the body including the testis. Estrogens are known to directly influence the growth, differentiation and function of specific target tissues and organs in humans and animals. These specific tissues and organs also include the mammary gland, uterus, prostate, pituitary, brain and liver. Estrogens also play an important role in bone maintenance and in the cardiovascular system, where estrogens have certain cardio-protective effects. In bone, both osteoclasts and osteoblasts have been reported to respond to estrogens with estrogen withdrawal leading to increased turnover and bone loss. A variety of naturally occurring and chemically synthesized estrogens have been identified and characterized, perhaps the best known of which is the endogenous estrogen, estradiol-17 beta (also known as E2).

B. Estrogen Receptors Estrogens, as a class of hormones, act by binding to the ligand binding domain (LBD) of an intracellular protein identified as an"estrogen receptor" (ER). The presence of this intracellular ER provides and accounts for both cell proliferation and protein synthesis by estrogen-dependent cells. In the absence of the estrogen hormone,

the estrogen receptor is biologically inactive both in vivo and in vitro; and, if the cells or tissues are homogenized and fractionated into cytosol and nuclear fractions, the estrogen receptor is found in the nucleus and may also be detected the cytosol.

The known estrogen receptors are members of the well studied family of gene regulatory proteins referred to as the steroid hormone receptor family. Nuclear receptors, such as steroid hormone receptors, have a modular structure with six distinct regions. The N-terminal domain is the A/B region which includes a non-ligand dependent activation function (See Fig. la). The C region is the DNA binding domain (DBD). The D region contains nuclear localization signals. Finally, the E domain contains the ligand binding domain (LBD) and the ligand-dependent transaction function. Kuiper et a/., Endocrinology 138 (3): 863 (1997); Tremblay et a/., Mol.

Endocrin 11 (3): 333 (1997). The central DBD is typically about 100 amino acids. Like the other members of the steroid-hormone receptor family, estrogen receptors are activated by the binding of estrogen to the C-terminal LBD. The receptor proteins enable cells to respond to various lipid-soluble hormones by activating or repressing specific genes, through the interaction between the steroid hormone and its receptor.

Steroid hormone receptors are distinguishable from other nuclear receptors in a number of respects, including the nature of their ligands, their association (in the unliganded state) with a repertoire of heat-shock proteins and the fact that they may bind to hormone response elements as homodimers. Mosselman et a/., FEBS Letters 392: 49 (1996); Komm et al., Science 241: 81 (1988); Burch et a/., Mol. Cell. Biol. 8 (3): 1123 (1988).

The conventional model for steroid hormone action has assumed that steroid hormone receptors act as transcriptional regulators only when complexed with their ligands. It has, however, become evident that the majority of steroid receptors are present in the cell nucleus even in the absence of ligand. The presence of the receptors in the nucleus, despite the absence of hormone, suggests possible additional regulatory functions for the receptor in its unbound state. For example, the thyroid hormone receptors (TR) have a dual regulatory role: in the presence of hormone they function as

transcriptional activators, whereas in the absence of hormone, TRs are response element (TRE) specific transcriptional repressors.

The first estrogen receptor discovered was ERa, which was known for the past ten years merely as ER. The human ER (hER) is composed of 595 amino acids in its unbound state and is approximately 67,000 Daltons. In the absence of estrogen-binding, the ERa protein can be located in vitro within the cytosol.

Transcription of ERa occurs from two separate promoters, PO and PI, although no functional mapping has been previously published. PI represents the major ERa transcriptional start site. The PI start site is predominantly utilized in human mammary epithelial cells (HMEC) and is the major start site in ERa-positive human breast carcinomas. Multiple start sites have been identified for the PO promoter. Studies of the murine ERa gene identified 10 start sites spanning approximately 60 bases, and there is a start site at-1,994 (from the PI start site) in human cells, which would agree closely with the major murine PO start site. Transcription from the PO promoter is characteristic of human endometrial tissue and can account for 12 to 33% of ERa transcription in breast carcinoma cells. The genetic regulatory control elements of the recently discovered Eugene have yet to be delineated. Kuiper et a/., (1996 and 1997); Tremblay et al., (1997); and Mosselman et a/., (1996). It remains to be determined whether the ERg gene contains regulatory elements, such as promoters and enhancers, that are similar or function in a manner analogous to those described for ERa.

In soluble systems and under set conditions, the ERa protein can be found in various molecular forms with sedimentation coefficients of 8S, 5S or 4S as determined by sucrose density gradient analysis. The 8S form of ERa protein is believed to be the inactivated, untransformed form of ERa protein associated with the unbound, inactive state of estrogen receptor in the absence of estrogen. The 4S ERa protein is a monomeric protein molecule that can be generated from the 8S form in vitro. The 4S form binds to both nuclei and DNA-cellulose in vitro; it is generally termed the"activated but untransformed"estrogen receptor protein. The 5S form of ERa is a dimeric protein

molecule, which is created by the conversion of the 4S ERa protein via a bimolecular reaction. It is generally believed that the 5S form of ERa protein is both"activated and transformed,"and therefore is the biologically active entity which binds to the DNA within the nuclei. Moreover, it is also this 5S form which is found associated with the nuclei subsequent to the administration of estradiol in vivo. Already it has been demonstrated that both ERa and ERP can form heterodimers (Kuiper and Gustafsson, FEBS 410: 87 (1997)).

Although the precise interactions between ERa and estrogens remain poorly understood, the generally accepted mechanism of action and sequence of events is believed to be as follows: When an estrogen, such as estradiol 17 beta (E2), is introduced to the target cells and tissues, there is specific binding between the estrogen and the ERa protein which results in the formation of an estrogen/ERa protein complex.

Also, at a time subsequent to hormone binding, a process termed activation and/or transformation ensues leading to the formation of functional estrogen/hormone receptor complexes possessing a high affinity for the nuclear components, the DNA of the target cell. Once the hormone/receptor protein complex is physically formed, it binds to the chromatin at specific binding sites on the chromosomes and regulates messenger ribonucleic acid (mRNA) transcription. If transcription is up-regulated, new messenger RNA (mRNA) is synthesized, chemically modified and exported from the nucleus into the cytoplasm of the cell where ribosomes then translate the mRNA into new proteins; the hormone/receptor protein complex can also down-regulate mRNA transcription.

This constitutes the well recognized estrogenic effect that occurs within cells and involves the regulation of new protein synthesis and concomitant new cell growth/proliferation or differentiation. It remains unclear whether ERP shares the same mechanisms of action and in the same order as have been demonstrated for ERa.

Certainly, the localization of ERß along with the manner in which it modulates transcription will be at least grossly similar to ERa; however, affinities for certain DNA sequences, as well as receptor ligands likely will differ between ERa and ERß, as there

is a 97% and 60% identity respectively between the DBD and LBD sequences between the two estrogen receptors. Tremblay et al., (1997).

The clinical significance of the ER in the management of breast cancer is well known. Tamoxifen, a substituted triphenylethylene antiestrogen, is a partial antagonist that is used in the management of ER-positive breast tumors. Gallo et al., Semin.

Oncol. 24: S 1 (1997). Generally, the expression of the receptor is usually associated with a better prognosis and is less metastatic. Bonetti et al., Breast Cancer Res. Treat.

38 (3): 289 (1996). However, in many cases the tumors are either ER-negative or contain splice variants that are commonly biologically inactive. Hence, there is interest to understand how ER gene regulation, as well as the editing of the ER message, contribute to the development of mammary cancer and its clinical outcome with chemotherapy drugs such as tamoxifen. For more background, see Gallo et al., (1997); Kangas, Acta Oncol. 31 (2): 143 (1992); Evans et al., Bone 17 (4S): 1815 (1995); Safarians et al., Cancer Res. 56 (15): 3560 (1996).

More recently, estrogen receptors have been linked to bone loss associated with postmenopausal osteoporosis. Paralleling that discovery has been the fact that certain antiestrogens (e. g., tamoxifen, raloxifene, droloxifene and tamoxifen methiodide), which by definition block the actions of estrogens, stimulate only the skeletal muscle tissues and have no corresponding stimulatory effect in the uterus or mesometrial fat. Somjen et al., J. Steroid Biochem. Mol. Biol. 59: 389 (1996); Grasser et al., J. Cell Biochem. 65: 159 (1997). These antiestrogens have been termed selective estrogen receptor modulators (SERMs); they typically possess estrogen agonist-like activity on bone tissues and serum lipids, while displaying potent estrogen antagonist properties in the breast and uterus. The observed paradoxical effects observed between the different estrogen receptor agonists and antagonists most likely corresponds to response differences to the antiestrogens between ERa and ERß. Tremblay et al., 361 (1997).

Estrogen receptors are also present in human and rat prostate, as evidenced by ligand binding studies. In contrast to androgen receptors, the major part of the estrogen receptors are localized in the stroma of the rat prostate, although the epithelial cells of

the secreting alveoli contain ER. Estrogens are, in addition to androgens, implicated in the growth of the prostate, and consequently estrogens have been implicated in the pathogenesis of benign prostatic hyperplasia. Habenich et al., J. Steroid Biochem. Mol Biol. 44: 557 (1993); Kuiper et al., PNAS 93: 5930 (1996). Diethylstilbesterol (DES), a stilbene estrogen with an increased affinity for ER, is used to treat prostatic hyperplasia and carcinoma. Goethuys et al., Am. J. Clin. Oncol. 20 (1): 40 (1997); Aprikian et al., Cancer 71 (12): 3952 (1993). Therefore, identifying the tissues and diseases that express ERP likely will prove helpful in the treatment of diseases involving ERß.

Estrogen has also been demonstrated to prevent osteoporosis. Postmenopausal osteoporosis, the most common bone disease in the developed world, is associated with estrogen deficiency. This deficiency increases generation and activity of osteoclasts, large multi-nuclear cells involved with bone resorption. Estrogen has been demonstrated to down-regulate osteoclast formation and function. Tamoxifen has been demonstrated to possess estrogenic effects on bone resorption likely through tamoxifen- induced osteoclast apoptosis. Hughes et al., Nat. Med. 2 (10): 1132 (1996). Isolation of additional reagents that inhibit progression of osteoporosis would be beneficial in treating postmenopausal women suffering from the disease.

C. ER Following the cloning of estrogen receptor a (ERa) 10 years ago, there was general acceptance that only one ER gene existed and consequently only one subtype of ER, ERa. This contraste sharply with other members of the nuclear receptor superfamily, where multiple forms have been reported, e. g., thyroid hormone receptor (TR) a and ß and retinoic acid receptor (RAR) a, ß, and y.

Recently, a novel rat ER cDNA was cloned from rat prostate and ovary tissues and named ERR subtype to distinguish it from the previously cloned ER cDNA, now named the ERa subtype. ERP was partially isolated from cDNA libraries from human testis, mouse ovaries and rat prostate, which are not generally considered to be major estrogen target tissues. The estrogen receptor subtype initially discovered was termed

ERR, but for purposes of this invention will be termed the incomplete ERR (ERR,) to differentiate it from the complete ERR (ER (3 or ERR-3) of the present invention, or the three claimed alternatively spliced isoforms (ERR-1, ERR-2 and ERR-4) of this invention. Mosselman et al., (1996); Kuiper et al., (1996 and 1997); Tremblay et al., (1997)."ERR-3"refers to the sequence as isolated from mouse ovaries or its analogous sequence in other mammalian species."ER3"refers to the sequence that encodes the complete ERR, which includes the novel 192 bp at the 5'terminus of exon 1 and the newly described exon 5B; ERRe includes ERR-3, the complete sequence that encodes the nine exons of murine ERß. ERß ;, as characterized using the clones obtained from mouse ovary tissue, encodes a protein that has a molecular weight of approximately 62 kDa and has a 60 kilobase (Kb) gene size. The isolated mouse ERR, gene, called Estrb by Tremblay et al., mapped to the central region of chromosome 12; the central region of mouse chromosome 12 shares homology with human chromosome 14q, suggesting that Estrb may lie here as well. Tremblay et al., (1997).

The ERß ; cDNA encoded a predicted protein of 485 amino acids and had a calculated molecular weight of 54,200. Kuiper et al., (1997). This protein, described by Kuiper et al., as Clone 29 (herein ERß ;), displays high affinity binding of estrogens, and in a transactivation assay system, it activates expression of an estrogen response element (ERE) containing a reporter gene construct in the presence of estrogens. Kuiper et al., (1996). Alignment of the ligand binding domain (LBD) of ERa (rat, mouse and human) and ER3 ; (rat) uncovered various regions of conservation, whereas other segments are non-conserved. Kuiper et al., (1996). The DNA binding domain (DBD) and C-terminal LBD of ER3 ; is highly homologous to the rat ERa. Kuiper et al., (1997); Tremblay et al., (1997).

ERR, was isolated in an effort to clone and characterize novel nuclear receptors or unknown isoforms of existing receptors. Degenerate primers were designed based on conserved regions within the DBD and LBD of nuclear receptors. Using these primers in conjunction with Polymerase Chain Reaction (PCR), rat prostate mRNA was amplified. One targeted tissue was the prostate, an organ of interest given the high

incidence of prostate cancer and benign prostatic hyperplasia. Nearly all prostate tumors eventually become androgen-independent, at which point they are beyond clinical control. Kuiper et al., (1996). Therefore, the possibility existed that, apart from androgens, other (steroid) hormones or locally produced factors interacted with nuclear receptors and modulated the cell proliferation, differentiation and apoptosis of the normal prostate. For example, human testicular receptors 2 and 4 (TR2 and TR4) and the estrogen receptor (ER)-related receptors (ERR1 and ERR2) are examples of orphan nuclear genes whose receptors are expressed in the prostate. Kuiper et al., (1996).

Tissue expression of ERß ; revealed additional differences from ERa expression.

Some tissues contain exclusively ERa (i. e., uterus, pituitary, epididymis, and kidney).

Other tissues display equal or greater levels of ERß ; RNA and may be expressed preferentially in the different cell types of an organ (i. e., ovary and prostate). Kuiper et al., (1996). In brain, ER3 ; appears to be a conspicuous fraction of the ER subtype RNA.

Although Northern blots did not detect ERß ; expression in peripheral blood lymphocytes, the initial PCR fragment of ERß ; cloned by Mosselman was acquired from these cells. Thus, the ER3 ; subtype may play a significant role in estrogen action in brain, ovary, prostate, hypothalamus and possibly other tissues. Mosselman et al., at 52 (1996); Byers et al., Mol. Endocrinol. 11 (2): 172 (1997); and Shughrue et al., Steroids 61 (12): 678 (1996).

In addition to differences in tissue expression, the order of competition for physiological estrogens and stilbene estrogens, which form a diphenolic resonance structure, for ERa versus the ERp, isofbrm was also observed to vary. Kuiper et al., (1997). These differences may result from the protein sequences differences observed between ERa and ERßC, as demonstrated in the comparison of the two murine subtypes in Figure 6. However, the order of affinity for the tested triphenylethylene antiestrogens however, was the same for both subtypes: 4-OH-tamoxifen >> nafoxidine > clomifene > tamoxifen. Kuiper et al., (1997). Such ligand binding differences most likely, portend different drug therapies for ERa versus ERP dependent disease. Another incongruity between the two ER subtypes is the agonistic-antagonistic difference in response to estrogens. This observed paradoxical disparity may relate not only to binding affinity,

but, more importantly to the presence of specific activation function domains located in ERi (e. g., AF-1 or AF-2). Tremblay et al., (1997).

Substantial interest exists in determining the individual and perhaps combined roles that both ERP and ERa play in carcinomas, increased estrogen turnover and bone loss. For example, tamoxifen augments bone growth, whereas it is an antagonist to ERa positive breast cancer. Gallo et al., (1997); Delmas et al., J. Clin. Oncol. 15: 955 (1997). This incongruous observation with tamoxifen administration may arise from the drug's different interactions with ERa and ERß. The mixed agonist-antagonist or pure antagonist actions observed with antiestrogens may result more specifically from binding differences between the activation function domains, AF-1 and AF-2. For example, studies using the estrogen antagonists 4-hydroxytamoxifen (OHT) and ICI 164,384 indicated that although both compounds blocked estrogen effects, their mode of action differed: the mixed agonist/antagonist OHT inhibited only AF-2 function, while the pure antiestrogen ICI 164,384 inhibited activation by both AF-1 and AF-2. Using the mouse ERß ;, all antagonists tested effectively inhibited E2 activity. In contras to ERa, OHT displayed no agonistic activity on ERpi. Tremblay et al., (1997). Therefore, once the underlying ER subtype responsible for a particular disease state is determined (e. g., ERa positive breast cancer), one may have a more accurate means of prognosticating the estrogen receptor related disease outcome; one may accurately follow therapies; one may develop gene specific and isoform specific therapies targeting diseases influenced by ERa and/or ERP; and one may provide for opportunities for varying the aggressiveness of the therapy.

SUMMARY OF THE INVENTION The present invention is based, in part, on the isolation and identification of the complete murine (m) estrogen receptor P gene (mER, and two alternatively spliced isoforms, e. g, mERßl and mERß2 and a third isoform isolated from rat (r) ovaries, rERß4. More broadly, the invention relates to the corresponding ERpe gene (including the human gene) and to certain mammalian receptors (denoted herein as ERß-1, ERß-2,

ERß-3 and ERß-4). The ERR ; sequence has been published by other laboratories, which had prematurely claimed that ER3 ; represented the complete ERß gene (ERßc), The present invention further provides nucleic acid molecules that encode the mERß-1, mERß-2, mERß-3 and mERß-4 proteins. Such nucleic acid molecules can be in an isolated form or can be operably linked to expression control elements or vector sequences.

The present invention also provides methods of identifying other alternatively spliced forms of the mERß3, the analogous mERß3 and corresponding ERpe as expressed in different animal species or additional ER subtypes. Specifically, the nucleic acid sequence of mERß3 can be used as a probe or to generate PCR primers to identify nucleic acid molecules that encode other members of the ERi family of proteins. The nucleic acid molecules encoding MERAI, mERß-2, mERß-3 or rERß-4 can be used to identify and isolate the ERß3 gene or corresponding ER (3 in other mammalian species, and has been used to isolate the E/-3 analog in human DNA.

The present invention further provides antibodies that recognize and bind to the ERpe protein or the mERß-3 protein or its isoforms. Such antibodies can be either polyclonal or monoclonal. Particularly preferred are antibodies that are specific for the complete receptor protein, ERßc, as opposed to antibodies against the previously known receptors, e. g., ERa and ER3 ;. More specifically, the invention claims an anti-peptide antibody that distinguishes between ERß ; and ER (3. Antibodies that bind to the ERpc protein can be utilized in a variety of diagnostic and prognostic formats and therapeutic methods. Alternatively, antibodies that can distinguish between the complete form, ERßc, and its isoforms may also be useful for purposes of diagnosis and treatment of ERP subtype based disease.

The present invention further provides methods for reducing, blocking or augmenting the association of an estrogen and other agonists and antagonists with the ERpc protein. For example, the association of an ERß-3 protein with a cytoplasmic signaling partner, such as estradiol, can be blocked or reduced by contacting the ERß-3 protein with a compound that blocks the binding of estradiol or other estrogen-like

agonists or antagonists (e. g., estrogens, stilbene estrogens or triphenylethylene antiestrogens). Tora et al., Cell 59: 447 (1989); Berry et al., EMBO 9: 2811 (1990).

Additionally, as the proteins are allosteric, the association of the ligand with ERR can also be influenced, in theory, by the dimer partner. Therefore, identifying agents that modulate ERP dimerization may pose another means of manipulating ERP regulation.

Blocking the interaction between the ligand and ERR-3 or one of its isoforms can be used to modulate biological and pathological processes that require such a ligand bound complex to mediate transcription. Such methods and agents can be used to modulate cellular proliferation, differentiation, DNA synthesis or cell cycle distribution.

The present invention further provides methods for isolating ERGS or ERi protein isoforms (e. g., ERß-1, ERß-2, ERß-3 and ERß-4) that regulate transcription.

For example, ERR-3 ligand binding partners, e. g., estrogen, are isolated using the ERß-3 protein or ligand binding portions thereof. Alternatively, for example, the DNA sequences that the ERß-3 protein binds can be determined, for example, utilizing electrophoretic mobility shift assays (EMSA), yeast two hybrid assays, or by affinity selection and degenerate ERE consensus sequences using the DNA binding domains (DBDs) of ER (3 or its isoforms. Berkowitz and Evans, J. Biol. Chem. 267 (10): 7134 (1992); Nawaz et al., Gene Expr. 2 (1): 39 (1992); Mosselman et al., (1996).

The invention also describes methods to screen compounds that can distinguish between ERa and ERp, and its isoforms (e. g., ERß-1, ERß-2, ERß-3 and ERß-4).

These methods will include methods of determining whether the compound binds and either functionally acts as an agonist or an antagonist with regard to each estrogen receptor. One method to determine whether compounds act in an agonistic or antagonistic fashion would use ERßc. in a yeast two hybrid system. Such methods have been previously employed to test the interaction of certain drugs with ERa and recognized by those of ordinary skill in the art. See Ichinose et al., Gene 188: 95 (1997); Collins et al., Steroids 62: 365 (1997); Jackson et al., Mol. Endocrinol. 11: 693 (1997).

The biological and pathological processes that require eskogen/ERßc complex can be modulated further by using gene therapy methods. Additional genetic manipulation within an organism can be used to alter ERßC gene expression or the production of a ERpc protein. For example, an ERß3 gene can be introduced into a mammal deficient for ERß-3 protein to correct the genetic deficiency; peptide modulators of ERß-3 activity can be produced within a target cell using genetic transfection methods to introduce into the target cells nucleic acid molecules encoding the modulators; and the Ergs3 gene can be introduced or deleted in a non-human mammal to produce animal models expressing ERß3 gene abnormalities or delete the gene entirely (e. g., knock-out mice). The latter application, ERß3 transgenic animals, is particularly useful for identifying agents in vivo that modulate ERß-3 activity and perhaps even other genes that encode proteins that influence ERß-3 actions. The use of nucleic acids for antisense and triple helix therapies and interventions are also expressly contemplated.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 Nucleotide sequence. deduced amino acid sequence and putative domain structure of the complete murine ERßC ßc gene (mERß-3) Fig. 1 (a). Illustrates the location of each of the nine exons comprising the clone of the complete murine ERßC, mERß3, and the splicing domains that yield the different alternatively spliced isoforms of mERß3. The numbers directly above the lines signifying the exons represented by terminal nucleotides of the exon. The sizes of the nine exons in base pairs (bp) and the encoded amino acid (a. a.) sequence for each of the exons and splice variants derived from mouse ovaries is indicated. The 1,704 nucleotides of mERß3 encodes a 567 amino acid protein. The letters (A through F) refer to regions of homology shared by all members of the steroid receptor super family.

Green et al., Cold Spring Harbor Symposia on Quantitative Biology 51 (2): 751-8 (1986). Region C corresponds to the DNA binding domain (DBD). Region E is the ligand binding domain (LBD). The newly described exon 5B lies within the LBD.

Exon 5B starts with GTCCTCA and stops with CCCAAG. The shaded regions in the rendering depict the amino terminus that is included in all mERp and rERß isoforms and the additional exon (exon 5B) that is included in the full length (mERß3) as well as the alternative spliced rat isoform, rERß4. The deletion of exon 6 in rERß4 results in a frame-shift and the juxtaposition of an in-frame stop codon causing the protein to be truncated, as indicated.

Clone mERß-1 is 1,650 bp in length. It contains a previously undescribed 192 bp located in the 5'end of exon 1 as well as the 7 other described exons ; mERß-1 lacks the newly described exon 5B. The isolated isoform mERß2 is 1,533 bp and lacks both exon 3 and exon 5B. Isoform rER, Q-4, isolated from rat tissue, is 1,570 bp. Although rERß4 possesses the new exon 5B, it lacks exon 6. The loss of exon 6 results in a frame shift that causes translation to terminate at a stop codon located in exon 7.

Fig. 1 (b). The full length sequence of murine mERßc (mERß-3 clone). The additional sequence included in all mER (3 clones (mERß-1, mERß-2), as well as the alternatively spliced rat isoform, (rERß-4) is noted in underlined bold type. The sequence included in the ninth exon, exon 5B, is presented in lower case letters beginning at base 1,149.

Figure 2 Amino acid sequences of the alternatively spliced isoforms of the mERP-3 Fig. 2 (a). Deduced amino acid sequences for alternative splice variant mERR-1.

The polypeptide sequence shared by all 3 of the alternatively spliced isoforms is indicated by the underlined sequence in bold characters. The mERß-1 protein contains 549 amino acid residues.

Fig. 2 (b). Deduced amino acid sequences for alternative splice variant mERß-2.

The alternatively spliced mERß-2 is 510 amino acid residues in length.

Fig. 2 (c). Deduced amino acid sequences for a rat alternative splice variant rERß-4. This splice variant was obtained from rat ovaries. The deletion of exon 6 produces a frame shift causing a truncation that terminates 13 amino acids beyond the translated exon 5B; the resulting rERß-4 protein likely is 414 residues long. The italicized, underlined, bold characters (residues 1-64) represent the polypeptide encoded by the novel 192 nucleotides located at the 5'terminus of exon 1. The characters indicated in bold and underlined represent the polypeptide encoded by exon 5B. The "*"refers to a translated stop codon.

Figure 3 Tissue Specific Expression of mERß-3 Protein Detected bv Western Blot Fig. 3 (a). Using anti-peptide antibodies (Antibody 1068) raised against N- CSSEDPHWHVAQTKSAVPR-OH (the sequence encoded by exon 5B), the mERß-3 protein was observed in a Western blot of human ovary, mouse ovary, rat ovary, ROS 17/2.8 cells, and murine primary osteoblasts protein extracts.

Fig. 3 (b). Western blot of human ovary, mouse ovary, rat ovary, ROS 17/2.8 cells, and murine primary osteoblasts protein extracts probed with antibody 1068 pre- immune sera. The protein extracts of each lane of both Figures 3 (a) and 3 (b) are: lane 1, human ovary; lane 2, mouse ovary; lane 3, rat ovary; lane 4, ROS 17/2.8 cells ; lane 5, ROS 17/2.8 cells treated with 100 nM estradiol for 16 hours; lane 6, murine primary osteoblasts.

Figure 4 Tissue Specific Expression of rERß DNA Detected bv Southern Blot of RT-PCR Products Fig. 4 (a). Total RNA from rat ovarian and ROS 17/2.8 cells amplified for 35 cycles using an oligo that can detect rERß. Each lane in Fig. 4 (a) contains PCR products derived from the following types of RNA: lane 1, control, no RNA; lane 2, rat ovarian RNA (0.1 ug); lane 3, ROS 17/2.8 cells (0.1 pg); lane 4, rat ovarian RNA

control (0. 1 llg), no reverse transcriptase (RT); and lane 5, ROS 17/2.8 total RNA (0.1 gag), no RT.

Fig. 4 (b). Total RNA from rat ovarian, ROS 17/2.8 cells, bone marrow RNA treated with estradiol and total RNA from primary osteoblasts in co-culture amplified for 25 cycles using an oligo derived from rERp (Accession #U57439). Each lane in Fig. 4 (b) contains the following types and amounts of RNA: lane 1, control, no RNA; lane 2, rat ovarian RNA (2 ng); lane 3, ROS 17/2.8 total RNA (0.1 pg), lane 4, total (cultured) bone marrow RNA (0.1 [tg); lane 5, total cultured bone marrow RNA (0.1 ug) where the cells had been treated with estradiol for 16 hours; lane 6, total RNA from primary osteoblasts in co-culture (0.1 ug); lanes 7-11, control reactions without reverse transcriptase (RT) for lanes 2-6, respectively.

Figure 5 Gel Shift Assay Fig. 5 (a). Gel shift analysis of mER (3-3. The receptor-DNA complex was disrupted using the anti-peptide antibody 1067, which recognizes polypeptides encoded by exon 5B.

Fig. 5 (b). Gel shift analysis of the human alpha form of the estrogen receptor (ERa). Disruption of the ERa-DNA complex was assayed using the two anti-peptide antibodies specific to exon 5B.

Both Fig. 5 (a) and (b) contain the following: lanes 1 and 2, extract alone ; antibody 1067, lanes 3 and 4; antibody 1067 pre-immune serum, lanes 5 and 6; antibody 1068, lanes 7 and 8; antibody 1068 pre-immune serum, lanes 9 and 10; lanes 11 and 12 are control lanes that contain 16 pg of untransfected COS-7 nuclear extract.

Figure 6 Comparison of mERß-3 protein with the murine ERa

The upper sequence is the protein sequence of mERß-3, whereas the lower sequence is that of the mouse (m) mERa. The"l"between the matched sequences indicates residue identity. The" :" between the matched sequences represents similar amino acids. The"."observed in the sequences is a"gap"added by the sequence alignment program. The lines bisecting the paired sequences delineate the six domains (A-F) found in ER (3 and ERa. There is 99% similarity and 97% identity between the C domains, which contain the DBD, of the two murine estrogen receptor subtypes. There is 79% similarity and 59% identity between the E domains, which contains the LBD.

Figure 7 Comparison between mERß-3 and mERßi nucleotide sequences The upper paired sequence (which starts at nucleotide 151) is the nucleotide sequence of mERß3, whereas the lower sequence is the nucleotide sequence of mERßj published by Tremblay et al., (1997). There is a one nucleotide difference between the mERj¢3 sequence (an adenine at 1,244) and mERßj (a guanine at 1,009). This nucleotide difference results in an aspartic acid (D) residue in mERß-3 and a glycine (G) residue in mers.

Figure 8 Activity of ERß Isoforms in the presence of various estrogens Reporter constructs expressing ERß-1 (B1), ERß-3 (B3), ERa (alpha), or both ERß-1 and ERß-3 (B1+B3) were exposed to clomiphene, diethylstilbesterol (DES), 4 OH-tamoxifen (4-OHT), or 17 P estradiol (E2). Expression was standardized to ERa response to 100 nM drug.

Figure 9 Transactivation Profiles-cV2ERE The four panels display the ability of the different estrogen receptors to transactivate cV2ERE. The cellular response of ERa (ER Alpha), murine ERß-1 (mER-

B 1), murine ERß-3 (mER-B3), or coexpression of both murine ERß-l and ERß-3 isoforms (mER B I +B3) in COS-7 cells to E2, clomid, DES and 4-OHT were compared.

Figure 10 In situ Hvbridization of Various Estrogen Target Tissues Ovaries (upper panels), uteri (middle panels), and E-15 rat embryos were serially sectioned and probed using anti-sense (left panels) and sense (right panels) probes from ERR and ERa. Cervical spine is shown in the lower panels.

I. General Description Estrogen receptors are members of the nuclear hormone receptor family.

Biologically, these proteins are intracellular receptors which mediate the effects of steroid hormones. Upon hormone binding, estrogen receptors control the transcriptional expression of certain hormone-responsive genes. This involves the binding of the receptors, often in homo-or heterodimeric form, to specific sequences, hormone response elements, located in the target gene promoter.

The compositions and methods of this invention provide for the screening of candidate compounds to be used to treat ERpc related diseases. The compositions are based on the isolation of an ER (3 sequence, Ergs-3, and the three alternatively spliced isoforms, ERß-l, ERß2 and ERß-4. Additionally, these compositions can be used to screen for ERpe based disease to facilitate disease prognosis and to monitor disease- related aberrant expression of ERßC or its isoforms.

II. Specific Embodiments The specific embodiments disclosed in this invention relate to the isolation of the nucleic acid sequence that encodes the ERpe gene, ERß-3. The murine (m) form of ERß3 is composed of 1,704 base pairs (bp) from the ATG start codon to TGA (Fig. la

and b) and encodes a 567 amino acid protein; this sequence contains nine exons, including the newly described exon 5B, which is located in the region encoding the LBD. Also isolated were three alternatively spliced isoforms: mERßl, mERß2 and rERß4. mERßl is 1,650 bp and encodes a 549 residue long polypeptide; ERß-1 lacks exon 5B (see Figs. la and 2a). mER, is composed of 1,533 base pairs (bp); it lacks both exon 5B and exon 3, which contains 117 bp (see Figs. 1 a and 2b). The sequence encoding rERß4, an alternatively spliced isoform isolated from rat ovaries, is composed of 1,570 bp; it contains exon 5B, and the 54 bp it comprises, but exon 6, which contains 134 bp, has been deleted (see Figs. 1 a and 2c).

Also described herein are methods of making and using the nucleic acid sequences corresponding to ERß-3, its isoforms and to the proteins encoded by these nucleic acid sequences. The methods of using the nucleic acid sequences of ERß-3 or its isoforms include determination of what tissues express ERRc and its isoforms (e. g., ERß-1, ERß-2 and ERß-4), function characterization for the proteins and nucleic acid sequences of ERß-3 and its isoforms, development of methods to recombinantly express ERßC nucleic acid molecules and their associated protein products, development of an ERß-3 reporter system, identification of ERß-3 ligands such as estrogen that influence ERß-3 or its isoforms and identification of compounds that modulate the influence exerted by ERß-3 or an isoform thereof on transcriptional regulation of other genes and determining the corresponding physiological effects of such influence.

A. Isolation of the complete (ER (3 DNA and protein Through such methods as reverse transcriptase (RT)-PCR and/or 5'RACE (rapid amplification of cDNA ends), the complete estrogen receptor ß and three isoforms were isolated. Genomic primers wee used for RT-PCR on mouse ovary RNAs to clone murein (m) mERß-1, mERß-2 and mERß-3. The sequences for mERßl, mERß3 and the rat (r) isoform rERß4 were obtained by 5'RACE using the Marathon system and a different set of primers. The primers and vectors chosen to isolate and clone these sequences would have been commonly known to an individual skilled in the art.

Using these techniques, a novel upstream sequence was found at the 5'end of exon 1. In mice and rats, this is an 192 bp sequence which is located at the 5'end of exon 1 of the mERP-3 gene, two alternatively spliced isoforms of mERß3 and the alternatively spliced isoform isolated from rat ovaries (rERß4), as a result of an additional open reading frame (ORF) located upstream of the published clones. Kuiper et al., (1996 and 1997); Mosselman et al., (1996); Tremblay et al., (1997). Further analysis by RT-PCR of mRNA derived from osteoblast and bone marrow co-cultures revealed a ninth exon, exon 5B, comprised of 54 bp and located within the LBD, as depicted in Figure la and b. As a result, the previously published human, rat and mouse sequences, all of which are referred to herein as ERßj, are probably 5'truncated splice variants of this larger complete ERpe form, which in the murine system is mERß3 (see Fig. la and b). The nucleic acid sequence information for ERß3 predicts a 567 amino acid protein with a molecular weight of approximately 63 kDa, instead of 54 kD predicted for ERp ;.

To obtain the analogous sequences in other mammals, such as humans, the heretofore unknown mER, gene or portions thereof can be used as probes. These probes should be of at least 18 nucleotides and preferably should be redundant for one or more sequences encoding the ERß-3 protein; the probes are to be designed from the ERpc amino acid sequence and should account for the degenerate genetic code. An appropriate cDNA library, such as that for ovary, testes or prostate cells, may then be screened with the probes for cDNAs which hybridize under standard conditions to one or more of the probe compositions. For examples of such general methods, see Sambrook et al., MOLECULAR CLONING: A LABORATORY MANUAL (1989). The cDNAs may then be isolated and sequenced to determine whether they code for the ER (3 protein. In this manner, the cDNA encoding the human ERpe protein or other mammalian ER, BC genes and their respective species specific isoforms may be isolated.

A method of isolating other ER (3 related genes is also described herein. Briefly, the nucleic acid sequences can be isolated by probing a DNA library such as that for prostate, ovary or testes, which is comprised of either genomic DNA or cDNA.

Libraries may be from either commercial sources or prepared from mammalian tissue by techniques known to those skilled in the art. The preferred cDNA libraries are human cDNA libraries which are available from commercial sources such as Stratagene.

The DNA libraries can be probed by plaque hybridization using oligonucleotide probes of at least 20 nucleic acid residues in length, which are complementary to unique sequences of murine or other ER, &3 genes. The preferred probes are the sequences for Primer 1 and Primer 2. The nucleic acid probes may be labeled to facilitate isolation of the hybridized clones. Labeling can be by any of the techniques known to those skilled in the art, but typically the probes are labeled with [32p] using terminal deoxynucleotidyl-transferase as disclosed in Sambrook et al., (1989).

Alternatively, those of skill may use polymerase chain reaction (PCR) technology to amplify nucleic acid sequences of the E/3 gene directly from mRNA, cDNA, genomic libraries or cDNA libraries. PCR or other in vitro amplification methods may also be useful, for example, to clone nucleic acid sequences that code for proteins to be expressed, to make nucleic acids to use as probes for detecting the presence of ERß3 DNA or ERß3 mRNA in tissue samples, for nucleic acid sequencing, or for other purposes.

Appropriate primers and probes for identifying ER3-3 from alternative mammalian tissues are generated from comparisons of the sequences provided herein.

For a general overview of PCR techniques, see PCR PROTOCOLS: A GUIDE TO METHODS AND APPLICATIONS (Innis, M, Gelfand, D., Sninsky, J. and White, T., eds.), Academic Press, San Diego (1990), incorporated herein by reference.

The present invention further provides nucleic acid molecules that encode ER, z 1, and the related ERp-3 isoform proteins herein described, preferably in isolated form.

As used herein,"nucleic acid"is defined as RNA or DNA that encodes a ERp-3 polypeptide, or is complementary to nucleic acid sequence encoding such peptides, or hybridizes to such nucleic acid and remains stably bound to it under appropriate stringency conditions, or encodes a polypeptide sharing at least 75% sequence identity, preferably at least 80%, and more preferably at least 85%, with the peptide sequences.

Specifically contemplated are genomic DNA, cDNA, mRNA and antisense molecules, as well as nucleic acids based on alternative backbone or including alternative bases whether derived from natural sources or synthesized. Such a hybridizing or complementary nucleic acid, however, is defined further as being novel and nonobvious over any prior art nucleic acid including that encodes, hybridizes under stringent conditions or other appropriate stringency conditions, or is complementary to a nucleic acid encoding an ERß-3 protein according to the present invention.

"Stringent conditions"are those that (1) employ low ionic strength and high temperature for washing, for example, 0.015 M NaCl, 0.0015 M sodium titrate, 0.1 % SDS at 50°C; or (2) employ during hybridization a denaturing agent such as formamide, for example, 50% (vol/vol) formamide with 0.1% bovine serum albumin (BSA), 0.1% Ficoll, 0.1% polyvinylpyrrolidone, 50 mM sodium phosphate buffer at pH 6.5 with 750 mM NaCI, 75 mM sodium citrate at 42°C. Another example is use of 50% formamide, 5 x SSC (0.75 M NaCI, 0.075 M sodium citrate), 50 mM sodium phosphate (pH 6.8), 0.1% sodium pyrophosphate, 5 x Denhardt's solution, sonicated salmon sperm DNA (50 gg/ml), 0.1% SDS, and 10% dextran sulfate at 42°C, with washes at 42°C in 0.2 x SSC and 0. 1% SDS. A skilled artisan can readily determine and vary the stringency conditions appropriately to obtain a clear and detectable hybridization signal.

B. Characterization of the new sequences The complete estrogen receptor ß, such as mERß3, contains nine exons. The three isoforms that have been isolated include mERßl, mERß2, mERß3 and the alternatively spliced isoform from rat ovaries, rERß4. mERßl is 1,650 bp; it contains the previously identified eight exons, lacks the new exon 5B, but contains the previously undescribed 192 bp located at the 5'end of exon 1 (see Figs. 1 a and 2a). In addition to having the novel 192 bp sequence located in the 5'terminus of exon 1 and the newly described exon 5B, there is a one nucleotide difference between the sequence published by Tremblay et al., (1997) and the sequence disclosed here: nucleotide 1,244 in exon 6

of the mERß-3 sequence is an adenine whereas in the sequence by Tremblay et al., (1997) it is a guanine (nucleotide 1,009). mERß2 contains 1,533 base pairs (bp); /2 lacks both exon 3 and exon 5B (see Figs. la and 2b). rERß4 includes 1,570 bp and has exon 5B, but exon 6 is deleted (see Figs. la and 2c). The full length mERß3 contains a previously unidentified 192 nucleotides at its 5'terminus as well as the sequences of exon 5B and exon 6. All three isoforms, as well as mERß3, contain the novel 192 bp located at the 5'terminus of exon 1.

One embodiment of this invention includes using ERß3 nucleic acid sequences containing the heretofore unknown 192 bp or 54 bp (exon 5B) domains or portions thereof and placing these sequences in appropriate vectors for purposes of replication.

Such vectors can then be introduced into the appropriate cell expression systems to express the proteins for use either in an assay system or to help to characterize the function of particular portions of the ERO, gene or its corresponding protein.

Characterization of the ERpc protein can be performed by creating mutants, using antibodies that recognize specific domains on ERp and using polypeptide sequences to specific regions of the protein to determine their function through competition assays. This invention proposes using such techniques to characterize the specific functions of the sequences or isoforms containing the novel 192 bp and/or exon 5B (54 bp) sequences.

Another method of characterizing ERR, ; and its isoform proteins includes the use of antibodies to map out specific functional domains on the ER (3 protein, including the LBD, the dimerization site, and the DNA binding domain (DBD) of the ERp protein.

Antibodies could also be utilized to determine whether the ER (3 or its isoforms is in a functional or non-functional conformation.

C. Creating antibodies to ER (3 protein sequences Antibodies are useful in several areas, including determining tissue expression of ERßC, such as ERß-3 or its isoforms (e. g., ERR-1, ERß-2 or ERß-4), and determining the functional domains of ERß-3 or its isoforms. Once the amino acid sequence of ERß-3 or its isoforms is known, another embodiment of this invention includes using polypeptides to create antibodies. Polypeptide sequences can be assessed using computer software to determine the antigenicity of certain polypeptide sequences for the purpose of creating antibodies to these ER3 specific polypeptides. Hopp and Woods, PNAS 78: 3824 (1981); Garnier et al., J. Mol. Bio. 120: 97 (1978).

One antibody that has been created is an anti-peptide antibody that can distinguish between the mERß-3 and ERßi, Other antibodies can be created to distinguish between the ERß-3 isoforms, in addition to being able to distinguish between the active and inactive states of ERß resulting from allosteric-induced ligand interactions with the receptor. The anti-peptide antibodies that distinguish between ERß-3 and ER3 ; were prepared using conventional methods and were raised to the polypeptide sequence encoded by exon 5B with a cysteine group at the amino terminus: N-CSSEDPHWHVAQTKSAVPR-OH (Antibodies 1067 and 1068). This antibody contains all of the exon 5B polypeptide. The Jameson-Wolf antigenicity program determined that this polypeptide possesses a high degree of antigenicity. Garnier et al., (1978). This program or the Hopp and Wood algorithm can also be employed to determine sequences of antigenicity in the novel amino terminus of ERß-3 and its isoforms to develop additional antibodies.

Two other antibodies were created that recognize both ERß-3 and ER (3 ;. These antibodies (Antibodies 1069 and 1070) were created against the following sequence: N- CSSTEDSKNKESSQ-OH. This polypeptide sequence is located in the carboxy terminus of the published rat ER (3 ;. Kuiper et al., (1996 and 1997). Antibodies 1067 and 1068 or 1069 and 1070 were obtained from the eggs of different chickens.

An alternative method to create antibodies to ERß-3 polypeptide sequences involves isolating ERß-3 proteins and digesting them with various proteases. The

cleavage fragments can then be purified by size and used to raise antibodies against specific portions of ERß-3. Finally, ERß-3 polypeptide sequences can be created recombinantly through fusion protein techniques. ERß-3 polypeptide sequences can be expressed by fusing the desired E/-3 nucleotide sequence to, for example, the gene expressing glutathione S-transferase (GST). The expressed ERß-3 polypeptide sequences created as a fusion ERß-3/GST fusion product can then be used to create antibodies to the specific portion of ERß-3 encoded in the ER3 containing fusion gene construct. Antibodies raised to such recombinant proteins can be either monoclonal or polyclonal and such preparation techniques are generally known.

Polyclonal antibodies 1067,1068,1069 and 1070 were raised in chickens. Other animals could also be utilized. Pre-immune sera was purified from 2-3 eggs collected prior to hen immunization. Immunizations were prepared with 2 mg of antigen conjugated to 2 mg Imject Keyhole limpet hemocyanin (KLH) via maleimide to the extra cysteine residue located at the amino terminus of each peptide as recommended in the manufacturer's (Pierce) instructions. The coupled carrier-antigen complex (0.5 ml) was emulsified with Complete Freund's adjuvant (0.5 ml) and 1.0 ml was used for the initial injection. The chickens were subsequently boosted every 2 weeks with coupled immunogen as described by Aves Laboratory, except that Incomplete Freund's Adjuvant was used. Six eggs were collected and the IgY was purified from the yolks. Other immunoglobulin isotypes and isotype subclasses can also be used (e. g., IgG,, IgG2, IgM).

Monoclonal antibodies may be obtained by various techniques familiar to those skilled in the art. Briefly, spleen cells from an animal immunized with a desired antigen are immortalized, commonly by fusion with a myeloma cell (see, Kohler and Milstein, Eur. J. Immunol. 6: 511 (1976)). Alternative methods of immortalization include transformation with Epstein Barr Virus, oncogenes, or retroviruses or other methods well known to those of ordinary skill in the art. Colonies arising from single immortalized cells are screened for production of antibodies of the desired antigen specificity and affinity. The yield of the monoclonal antibodies produced by such cells

may be enhanced by various techniques, including injection into the peritoneal cavity of a vertebrate host.

Alternatively, peptide specific antibodies, such as antibodies 1067 and 1068, are prepared by immunizing suitable mammalian hosts (e. g., chickens or rabbits) under appropriate immunization protocols using the peptide haptens alone, if they are of sufficient length, or if required to enhance immunogenicity, conjugated to suitable carriers. Methods for preparing immunogenic conjugates with carriers such as bovine serum albumin (BSA), keyhole limpet hemocyanin (KLH) or other carrier proteins are well known in the art. In some circumstances, direct conjugation using, for example, carbodiimide reagents may be effective. In other instances, linking reagents, such as those supplied by Pierce Chemical Co., Rockford, IL, may be desirable to provide accessibility to the hapten. The hapten peptides can be extended at either the amino or carboxy terminus with a cysteine (Cys) residue or interspersed with cysteine residues, for example, to facilitate linking to a carrier. Administration of the immunogens is conducted generally by injection over a suitable time period and with use of a suitable adjuvant, as is generally understood in the art. During the immunization schedule, titers of antibodies are taken to determine adequacy of antibody formation. For more information, refer to Harlow and Lane, ANTIBODIES: A LABORATORY MANUAL, Cold Spring Harbor Pubs., N. Y. (1988), which is incorporated herein by reference.

While the polyclonal antisera produced in this way may be satisfactory for some applications, for pharmaceutical compositions, the use of monoclonal preparations is preferred. Immortalized cell lines which secrete the desired monoclonal antibodies may be prepared using the standard method of Kohler and Milstein or with modifications which effect immortalization of lymphocytes or spleen cells, as is generally known.

Kohler and Milstein, (1976). The immortalized cell lines secreting the desired antibodies are screened by immunoassay in which the antigen is the peptide hapten or is the ERp protein itself. When the appropriate immortalized cell culture secreting the desired antibody is identified, the cells can be cultured either in vitro or by production from ascites fluid.

The desired monoclonal antibodies are then recovered from the culture supernatant or from the ascites supernatant. Fragments of the monoclonal or the polyclonal antisera which contain the immunologically significant portion can be used as antagonists, as well as the intact antibodies. Use of immunologically reactive fragments, such as the Fv, Fab, Fab', or F (ab') 2 fragments, is often preferable, especially in a therapeutic context, as these fragments are generally less immunogenic than the whole immunoglobulin. The antibodies or fragments may also be produced, using current technology, by recombinant means. Regions that bind specifically to the desired regions of the receptor can also be produced in the context of chimeras with multiple species origin. Alternatively, the ERp specific antibody can be humanized antibodies or human antibodies, as described in U. S. Patent No. 5,585,089 by Queen et al. See also Riechmann et al., Nature 332: 323 (1988). Finally, given that ERp protein may be involved in certain cancers, it would be useful to create bispecific antibodies capable protein and, for example, cytotoxic T cells to facilitate the killing of tumor cells which may be useful in treating cancer. Berg et al., PNAS 88: 4723 (1991).

The antibodies thus produced are useful not only as modulators of ERßc-estrogen interaction, but are also useful in immunoassays to detect ERp protein or its isoforms and for the purification of ERßC protein or its protein isoforms. One can use immunoassays to detect the ERp protein or its alternatively spliced isoforms.

Immunoassays can be used to qualitatively and quantitatively analyze the ER (3 protein.

A general overview of the applicable technology can be found in Harlow and Lane, (1988). In brief, ER (3 protein or a fragment or isoform thereof is expressed in transfected cells, preferably bacterial cells, and purified as generally described above and in the examples. The product is then injected into a mammal capable of producing antibodies. Either monoclonal or polyclonal antibodies specific for the gene product can be used in various immunoassays; such assays include enzyme linked immunoabsorbant assays (ELISAs), competitive immunoassays, radioimmunoassays, Western blots (Fig.

3), indirect immunofluorescent assays, gel shift assays (Fig. 5) and the like.

D. Creating polypeptides that interfere with the binding domains of ERßC One embodiment of this invention utilizes ERßC polypeptide sequences to assay their ability to interfere with ERpc protein mediated transcription regulation. Such interference can be created by preventing ERp activating agents, such as estradiol, from binding protein. Alternatively, a polypeptide could be designed to inhibit dimerization and subsequent signaling from occurring. Such polypeptides could be created using peptide synthesizers or by creating fusion protein expressing gene constructs or other expression systems for either prokaryotic or eukaryotic cell systems.

In brief summary, the expression of natural or synthetic nucleic acids encoding mammalian ER, 6, will typically be achieved by operably linking the gene or cDNA to a promoter (which is either constitutive or inducible) and incorporating it into an expression vector. The vectors preferably are suitable for replication and integration in either prokaryotes or eukaryotes. Typical cloning vectors contain transcription and translation terminators, initiation sequences and promoters useful for regulation of the expression of the ERO, gene. The vectors may also comprise generic expression cassettes containing at least one independent terminator sequence, sequences permitting replication of the plasmid in both eukaryotes and prokaryotes, i. e., shuttle vectors and selectable markers for both prokaryotic and eukaryotic systems.

Methods for the expression of cloned genes in bacteria are also well known. In general, to obtain high level expression of a cloned gene in a prokaryotic system, it is essential to construct expression vectors which contain, at a minimum, a strong promoter to direct DNA replication. The inclusion of selectable markers in DNA vectors transformed in E. coli is also useful. Examples of such markers include genes specifying resistance to ampicillin, tetracycline or chloramphenicol.

Suitable eukaryote hosts may include plant cells, insect cells, mammalian cells, yeast and filamentous fungi. In a preferred embodiment of this invention, the baculovirus/insect cell system is used for gene expression.

The protein encoded by the ERßC gene, which can be produced by recombinant DNA technology, may be purified by standard techniques well known to those of skill in

the art. Recombinantly produced ERp can be directly expressed or expressed as a fusion protein. The protein is then purified by a combination of cell lysis (e. g., sonication) and affinity chromatography. For fusion products, subsequent digestion of the fusion protein with an appropriate proteolytic enzyme releases the desired ERßC, its isoforms or a fragment thereof.

The purified ERpe, when described as"isolated"and or"substantially pure", describes a protein that has been separated from components which naturally accompany it. Typically, a monomeric protein is substantially pure when at least about 85% or more of a sample exhibits a single polypeptide backbone. Minor variants or chemical modifications may typically share the same polypeptide sequence. Depending on the purification procedure, purities of 85%, and preferably over 95% pure are possible.

Protein purity or homogeneity may be indicated by a number of means well known in the art, such as polyacrylamide gel electrophoresis of a protein sample, followed by visualizing a single polypeptide band on a polyacrylamide gel upon staining. For certain purposes, high resolution will be needed and high performance liquid chromatography (HPLC) or a similar means for purification utilized.

The ERp protein or its isoforms of this invention may be purified to substantial purity by standard techniques well known in the art, including selective precipitation with such substances as ammonium sulfate, column chromatography, immunopurification methods, and others. See, for instance, R. Scopes, PROTEIN PURIFICATION: PRINCIPLES AND PRACTICE, Springer-Verlag: New York (1982).

The ERp polypeptides or isoform polypeptides could then be used in various assays, such as gel shift assays or yeast two hybrid systems wherein these polypeptide sequences can be tested to observe their binding ability to the hormone response elements (HRE) on DNA sequences, dimerization binding ability, and agonist/antagonist binding ability.

E. Determining tissue localization of ERp or its isoforms by nucleic acid hybridization

Using portions of the newly isolated ER, (3 gene, probes can be synthesized either using polymerase chain reaction (PCR) techniques or using in vitro transcription, of which both techniques are known to skilled artisans. These probes, which are typically radiolabeled, can be utilized to determine which tissues express a particular ER, 6, transcript either via Northern blot analysis or dot blots of RNA samples or by Southern blots wherein the mRNA has been reverse transcribed into DNA, which is then further amplified using polymerase chain reaction (PCR) as demonstrated in Fig. 4. Southern analysis of DNA is also useful in determining whether the ERßC gene is present or disrupted. For example, it is known that ERa is disrupted in certain breast tumors; such information may in turn be beneficial in determining the course of chemotherapy to be utilized on a patient. Using nucleic acid sequences unique to the ERßC, it can be readily determined what tissues express the gene.

The present invention also provides methods for detecting the presence, absence and or abnormal expression of ERßC gene products in a physiological specimen, as well as in other tissue samples. One method for evaluating the presence or absence of ER (3 in a sample involves a Southern transfer and is well known to those of skill in the art (Fig. 4). Briefly, the digested genomic DNA is run on agarose slab gels in buffer and transferred to membranes. Hybridization is carried out using the probes discussed above. Visualization of the hybridized portions allows the qualitative determination of the presence or absence of the ERßC gene or its isoforms. Southern blotting will also distinguish, depending on the stringency conditions used for hybridization, whether the ERßC gene is normal or contains gene deletions or rearrangements.

Similarly, a Northern transfer may be used for the detection of ERßC messenger RNA (mRNA) in tissue samples of mRNA. This procedure is also well known in the art. See Maniatis et al., MOLECULAR CLONING: A LABORATORY MANUAL, Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. (1982). In brief, the mRNA is isolated from a given cell sample using an acid guanidinium-phenol-chloroform extraction method. The mRNA is then electrophoresed to separate the mRNA species, and the mRNA is transferred from the gel to a nitrocellulose membrane. Labeled probes are used to identify the presence or absence of the ERßC transcript.

An alternative means for determining the level of expression of the ERßC gene is in situ hybridization. In situ hybridization assays are well known and are generally described in Angerer et al., Methods Enzymol., 152: 649 (1987). This hybridization technique has already been used to study ER3 ; expression in rat hypothalamus.

Shughrue et al., (1997). In an in situ hybridization, cells are fixed to a solid support, typically a glass slide. If DNA is to be probed, the cells are denatured with heat or alkali. The cells are then contacted with a hybridization solution at a moderate temperature to permit annealing of ERßC specific probes that are labeled. The probes are preferably labeled with radioisotopes or fluorescent reporters.

F. Recombinant nucleic acid molecules containing ERi sequences This invention relates to recombinant sequences that express the entire ERßC gene, its isoforms or portions thereof that include the newly described 5'terminus or the newly described 54 bp of exone 5B, as described in Figures 1 and 2. The invention includes all methods of expressing such recombinant constructs both in prokaryotic and eukaryotic replication systems, which would have been known to one skilled in the art.

The methods for using recombinant deoxyribonucleic acid (DNA) or recombinant ribonucleic acid (rRNA) sequences would include sequences that form triple helixes or sequences that are antisense to the ERO, MRNA or isoform mRNA.

Additional methods would include expressions of these recombinant nucleic acid sequences to express the encoded protein.

G. Host cells containing an exogenously supplied jEjR. encoding nucleic acid molecule The invention also relates to a method of introducing the recombinant full length form of ERßC, such as E-3 or one of the other isolated isoforms, ERßl, ERß2 or ERß4 into non-ERp-3 expressing cells and assaying the effect said rDNA and its associated protein product have on transcriptional regulation. Cells transfected with either the full-length (ERß3) or alternatively spliced isoforms of E/ (e. g., ERßl,

ER, I-2 or ER, Q-4) can then be utilized to assay the transfected cells'ability to form colonies in soft agar, different rates of DNA synthesis, differences in cell-cycle distribution in cells expressing different ERß3 isoforms and altered morphology of the transfected cells.

The present invention further provides host cells transformed or transfected with a nucleic acid molecule encoding an ERß-3 protein. The host cell can be either prokaryotic or eukaryotic. Eukaryotic cells useful for expression of a ERß-3 protein are not limited, so long as the cell line is compatible with cell culture methods and with the propagation of the expression vector and expression of the ERß3 gene product. Preferred eukaryotic host cells include, but are not limited to, yeast, insect and mammalian cells, preferably vertebrate cells such as those from a mouse, rat, monkey or human fibroblastic cell line. Particularly preferred eukaryotic host cells include insect cells. Any prokaryotic host can be used to express an ERß3 encoding recombinant DNA (rDNA) molecule. The preferred prokaryotic host is E. coli.

Transformation of appropriate cell hosts with a rDNA molecule of the present invention is accomplished by well known methods that typically depend on the type of vector used and host system employed. With regard to transformation of prokaryotic host cells, electroporation and salt treatment methods are typically employed, see, for example, Cohen et al., PNAS 69: 2110 (1972); and Maniatis et al., (1982); Sambrook et al., (1989); or CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, F. Ausubel et al., ed. Greene Publishing and Wiley-Interscience, New York (1987). With regard to transformation of vertebrate cells with vectors containing rDNAs, electroporation, cationic lipid or salt treatment methods are typically employed, see, for example, Graham et al., Virologv 52: 456 (1973); Wigler et al., PNAS 76: 1373 (1979); and Sambrook et al., (1989).

Successfully transformed cells, ie., cells that contain a rDNA molecule of the present invention, can be identified by well known techniques. For example, cells resulting from the introduction of an rDNA of the present invention can be cloned to produce single colonies. Cells from those colonies can be harvested, lysed and their DNA content examined for the presence of the rDNA using a method such as that described by Southern,

J. Mol. Biol. 98: 503 (1975) or Berent et al., Biotech 3: 208 (1985) or the proteins produced from the cell assayed via an immunological method, as discussed above.

Utilization of the full-length, fragments, or isoforms of ERß3 to determine their ability to regulate the formation of colonies in soft agar is useful in assessing whether a particular isoform of the ER, &3 gene is responsible for cellular proliferation and/or differentiation. The ability of a particular E/-3 isoform to spur proliferation and/or differentiation may in turn correspond to the gene's involvement in certain ERP associated diseases.

Alternatively, the isoforms may be used in transfected cell lines to assay [3EI1- thymidine incorporation to test the effect of a particular ERl isoform on DNA synthesis.

Fluorescent activated cell sorting (FACS) could be utilized to determine differences in cell growth between cells bearing one isoform over another. Finally, transfected cells could be examined for morphological changes due to the expression of different ERßC isoforms.

Once the properties of ERßC are determined with respect to impact on DNA expression, changes in morphology, and effects on cellular proliferation and/or differentiation, the same assays can be implemented to identify compounds that regulate the observed effects induced by isoforms of ER (3. Identification of putative drugs, which are discussed in greater detail herein, would be valuable in modulating concentrations of ERßC proteins or its isoforms in diseases involving such proteins.

H. Production of ERße protein using recombinant methods This invention also describes the methods used to express the ERßC protein, such as by using recombinant DNA (rDNA) of the E/ ?-3 gene, such as using its novel isoforms (ERß-1, ERß-2 and Ergs4) or portions thereof. Methods for generating rDNA molecules are well known in the art, for example, see Sambrook etal., (1989). The preferred rDNA molecules would contain an ERß3 encoding DNA or a DNA encoding one of its isoforms operably linked to expression control sequences and/or vector sequences.

The choice of vector and/or expression control sequences to which one of the ERE nucleic acid molecules of the present invention is operably linked depends directly, as is well

known in the art, on the functional properties desired, e. g., protein expression, and the host cell to be transformed. Any vector contemplated by the present invention should be at least capable of directing replication, insertion into the eukaryote's chromosome or replicating extrachromasomally in a prokaryote, and preferably also expression of the ERß-3 protein encoded in the rDNA molecule.

Expression control elements that are used for regulating the expression of an operably linked protein encoding sequence are known in the art and include, but are not limited to, inducible promoters, constitutive promoters, secretion signals, and other regulatory elements. Preferably, the inducible promoter is readily controlled, such as being responsive to a nutrient in the host cell's medium.

In one embodiment, the vector containing a ERß-3 encoding nucleic acid molecule will include a prokaryotic replicon, i. e., a DNA sequence having the ability to direct autonomous replication and maintenance of the recombinant DNA molecule extrachromosomally in a prokaryotic host cell, such as a bacterial host cell, transformed therewith. Such replicons are well known in the art. In addition, vectors that include a prokaryotic replicon may also include genes which confer such detectable markers as a drug resistance marker. Typical bacterial drug resistance genes are those that confer resistance to ampicillin or tetracycline. Vectors that include a prokaryotic replicon can further include a prokaryotic or viral promoter capable of directing the expression (transcription and translation) of the Ergs3 gene sequences in a bacterial host cell, such as E. coli. A promoter is an expression control element formed by a DNA sequence that permits binding of RNA polymerase and transcription to occur. Promoter sequences compatible with bacterial hosts are typically provided in plasmid vectors containing convenient restriction sites for insertion of a DNA segment of the present invention. Typical of such vector plasmids are pUC8, pUC9, pBR322 and pBR329 available from Biorad Laboratories, (Richmond, CA), pPL and pKK223 available from Pharmacia, Piscataway, N. J.

Expression vectors compatible with eukaryotic cells, preferably those compatible with vertebrate cells, can also be used to form rDNA molecules that contain ERß-3 sequences. Eukaryotic cell expression vectors are well known in the art and are available

from several commercial sources. Typically, such vectors are provided containing convenient restriction sites for insertion of the desired DNA segment. Typical of such vectors are pSVL and pKSV-10 (Pharmacia), pBPV-l/pML2d (International Biotechnologies, Inc.), pTDT1 (ATCC, #31255), the vector pCDM8 described herein, and like eukaryotic expression vectors.

Eukaryotic cell expression vectors used to construct the rDNA molecules of the present invention may further include a selectable marker that is effective in an eukaryotic cell, preferably a drug resistance selection marker. A preferred drug resistance marker is the gene for which expression results in neomycin resistance, ie., the neomycin phosphotransferase (neo) gene as described by Southern et al., J. Mol. Anal. Genet. 1: 327 (1982).

1. Diagnostic technique for measuring ERßC mRNA transcript levels Another embodiment of the present invention is the use of ERßC nucleic acid sequences to measure changes in cells'mRNA concentrations. Methods of quantitatively and/or qualitatively assessing mRNA levels includes Northern blotting, in situ hybridization, nucleic acid hybridization and RT-PCR. Raval, J. Pharmacol. Toxicol.

Methods 32 (3): 125 (1994). One may use the coding sequence of ERßC or its isoforms, particularly the sequences of ER, 6,, not found in ERßj, to determine the level of mRNA present in the cell. By lysing cells obtained by a biopsy, under conditions which inhibit RNases in accordance with conventional methodologies, the mRNA may be reverse transcribed into DNA and the DNA expanded using PCR. The expanded DNA may then be quantified. Less conveniently, Northern blot analysis may be used as described above.

Reverse transcription PCR (RT-PCR) has been used to ascertain specific mRNA concentrations in breast cancer cells and would be commonly known to individuals skilled in the art. See for example, Bartlett et aL, Br. J Cancer 73 (12): 1538 (1996). The benefits of using RT-PCR is that sample sizes do not have to be great to obtain valuable and sensitive results, as was observed in a study looking at mRNA levels of heart muscle biopsies.

Engelhardt et al., J. Am. Coll. Cardiol. 27 (1): 146 (1996).

Determination of ERßC mRNA expression can also be assessed using in situ hybridization. This in situ labeling technique, which would employ labeled nucleic acid sequences capable of hybridizing to ERßC mRNA or its alternatively spliced isoforms and subsequent detection by a imaging device, would be useful in localizing tissues that have increased or decreased expression of ERßC or its isoforms'mRNA. This technique also would be commonly known to individuals skilled in the art. For example, see Guldenaar et al., Brain Res. 700 (1-2): 107 (1995).

By employing any of the above diagnostic techniques, the presence and amount of transcription and expression of ER3 or its isoforms may be determined, as a measure of the expression of ERß-3 protein, as well as other proteins for which transcription is regulated by the ERß-3 protein. This information is related to the aggressive nature of a particular cancer, the change in the nature of the cancer in relation to treatments, such as irradiation, chemotherapy, or surgery, the metastatic nature of the cancer, as well as the aggressiveness of metastases, and the like. For example, see Maas et al., Cancer Lett.

97 (1): 107 (1995), which discussed changes of specific mRNA levels in breast cancer cells using RT-PCR after treatment with different anti-cancer agents. This relationship may be useful for determining the level of therapeutic treatment, monitoring the response of the tumor (or other ERpe related diseases) to the therapeutic treatment, and in providing a prognosis for the patient concerning the course of the disease.

J. Methods to identify agents that block ERpc transcriptional regulation Another embodiment of the present invention provides methods for identifying agents that inhibit or block the association of an estrogen or estrogen-like agonists/antagonists with ERp protein. For example, estrogen can be mixed with the ERp protein or a cellular extract containing the ERßC, in the presence and absence of the compound to be tested. After mixing under conditions that allow association of the estrogen or estrogen-like agonist/antagonist with ERßC, the two mixtures are analyzed and compared to determine if the compound augmented, reduced or completely blocked the association of the estrogen or estrogen-like agonist/antagonist with the ERp protein or its isoforms.

Agents that block or reduce the association of an estrogen or estrogen-like agonist/antagonist with the ERpc protein will be identified as decreasing the concentration of estrogen-ERßc binding present in the sample containing the tested compound.

The receptor protein likely must undergo allosteric change in its conformation before the estrogen-ERpc complex has the ability to bind to DNA. Once inside the nucleus, the activated receptor initiates transcription of genetic information from the DNA to mRNA, which is in turn a template for the linking of amino acids into proteins.

The antiestrogen effects produced by drugs such as tamoxifen (Nolvadex Registered TM) appear to be one of preventing the estrogen receptor from interacting with DNA in the nucleus to stimulate RNA and protein synthesis. This action initiates a block in the synthesis of macromolecules such as proteins, causing cell damage and the ultimate death of the cell.

Antiestrogens are believed to be lipophilic molecules having a portion of the molecule which resembies naturally occurring estrogens. This portion of the antiestrogen selectively binds to the estrogen receptors. The antiestrogens, however, have a side chain arm (e. g, dimethylaminophenyl ethoxy) which distorts the three-dimensional configuration of the estrogen receptor preventing translocation of the receptor to the nucleus. Morgan, U. S.

Patent No. 4,732,904 (1988). Another method of determining whether candidate reagents inhibit estrogen action on the complete estrogen receptor P subtype would be by determining whether ERR has undergone an allosteric transformation as a result of interacting with a candidate reagent such that ER (3 or its isoforms can no longer combine with the native substrate, estrogen. Changes in the conformation of ER (3 or homodimers of ERßC can be detected using antibodies, either monoclonal or polyclonal, to conformational epitopes that exist on ERßC or homodimers of the receptor. Antibodies were used to determine the functional state of ERa and a similar method could be used in determining whether compounds augment transformation into the activated allosteric conformation or inhibit the conformation all together. See Wotiz et al., U. S. Patent No. 5,312,752 (1994).

Antibodies can not only be used to determine whether the ER (3 is functionally in an active or inactive state. Antibodies could also be screened to determine whether their binding to either the ligand or to the receptor itself enhanced the binding of the ligand to the

receptor. Methods of determining said enhancement are known to the art. See Aguilar et al., Mol. Cell. Biochem. 136 (1): 35 (1994). Another method of determining whether a particular reagent augments or inhibits dimerization of ERßC or augments or inhibits ERpe from assuming the activated state would be to utilize a yeast two hybrid system. Yeast two hybrid systems have been successfully used to determine whether ERa dimerization is ligand-dependent (Wang et al., J. Biol. Chem. 270 (40): 23,322 (1992)); to isolate agents such as proteins or antibodies that enhance transcriptional activity of hormone receptors (Onate et al., Science 270 (5240): 1354 (1995)); to isolate compounds that are antagonistic to ERp action in a manner comparable to what has been done with ERa (Ichinose et al., (1997) and Collins et al., (1997)); and to determine whether ERp can form heterodimers in a manner analogous to what has been observed for retinoic acid receptors. See for further discussion Forman et al., Cell 81 (4): 541 (1995) and Walfish et al., PNAS 94 (8): 3697 (1997).

Another method to screen agents is to use a reporter gene such as ß-galactosidase (ß ga or luciferase. These transactivation experiments can be performed in yeast or in mammalian cell lines. The cells would contain ERp along with an appropriate estrogen responsive element (ERE) upstream of the reporter gene (e. g., luciferase or ßgal), such as cV2ERE. Both antagonists and agonists of ERP can be assayed in this fashion. Gaido et al., Toxicol. Appl. Pharmacol. 143 (1): 205 (1997); Hafner et al., J. Steroid Biochem. Mol.

Biol. 58 (4): 385 (1996); Muhn et al., Ann. N. Y. Acad. Sci. 761: 311 (1995). Such assays can also be utilized to determine whether cross-talk occurs for ERp and progesterone (PR) as has been demonstrated for ERa and PR. Kraus et al., Mol. Cell. Biol. 15 (4): 1847 (1995).

Identifying novel ER (3 responsive elements can be done rapidly using libraries of degenerate oligonucleotides. The protocol requires no purified protein and specifically selects for functional response elements. Nawaz et al., (1992).

Compounds that are assayed by the above methods can be randomly selected or rationally selected or designed. As used herein, an agent is said to be randomly selected when the agent is chosen arbitrarily, without considering the specific sequences involved in the association of the estrogen or estrogen-like agonist/antagonist to the ERp, protein. An

example of such randomly selected agents is the use a chemical library, a peptide combinatorial library or a growth broth of an organism.

As used herein, an agent is said to be rationally selected or designed when the agent is chosen on a non-random basis which takes into account the sequence of the target site and/or its conformation in connection with the agent's action. Agents can be rationally selected or rationally designed by utilizing the peptide sequences that recognize and bind to either the estrogen or estrogen-like agonist/antagonist or to the steroid hormone binding site on the ERpe protein.

The agents of this embodiment can be, by way of example, peptides or other small molecules, antibodies (e. g, monoclonal or polyclonal), fragments of antibodies (e. g, Fv), or drugs with antiestrogenic or estrogenic activity (e. g., narigenin, kaempferide, phloretin, biochanin A, flavone, ICI 182,780, raloxifene, tamoxifen, [6-hydroxy-3- [4- [2- (1- piperidinyl) ethoxy] phenoxy]-2-] 4-hydroxybenzo [b] thiophene, raloxifene HC1, and ethynyl estradiol). Collins et al., (1998); Palkowitz et al., J. Chem. Med. 40 (10): 1407 (1997). A skilled artisan can readily recognize that there is no limit as to the structural nature of the agents of the present invention. One class of compounds of the present invention includes polypeptide agents whose amino acid sequences are chosen based on the amino acid sequence of the ER3 LBD.

The peptide agents of the invention can be prepared using standard solid phase (or solution phase) peptide synthesis methods, as is known in the art. In addition, rDNAs encoding these polypeptides may be synthesized using commercially available oligonucleotide synthesis instrumentation and produced recombinantly using standard recombinant production systems. These rDNA molecules can then be utilized to recombinantly express polypeptides that bind to the ERß-3 protein or its isoforms. The production using solid phase peptide synthesis is necessitated if non-recombinantly produced polypeptide sequences are to be used.

K. Administration of agents that affect ERp signaling

The agents of the present invention can be provided alone, or in combination with additional agents that modulate a particular pathological process. For example, an agent of the present invention that reduces or otherwise modulates ER transcriptional regulation, by blocking estrogen or other agonist/antagonists from binding and transforming the ER (3 protein or its isoforms into an active state, can be administered in combination with other similar agents. As used herein, two agents are said to be administered in combination when the two agents are administered simultaneously or are administered independently in a fashion such that the agents will act at the same time.

The agents of the present invention can be administered via parenteral, subcutaneous, intravenous, intramuscular, intraperitoneal, transdermal, or buccal routes.

Alternatively, or concurrently, administration may be by the oral route. The dosage administered will be dependent upon the age, health, and weight of the recipient, kind of concurrent treatment, if any, frequency of treatment, and the nature of the effect desired.

The present invention further provides compositions containing one or more agents which block transcriptional regulation by the ERß-l protein. While individual needs may vary, determination of optimal ranges of effective amounts of each component is within the skill of the art.

In addition to the pharmacologically active agents, the compositions of the present invention may contain pharmaceutically acceptable carriers comprising excipients and auxiliaries which facilitate processing of the active compounds into preparations which can be used pharmaceutically for delivery to the site of action. Suitable formulations for parenteral administration include aqueous solutions of the active compounds in water- soluble form, for example, water-soluble salts. In addition, suspensions of the active compounds as appropriate oily injection suspensions may be administered. Suitable lipophilic solvents or vehicles include fatty oils (e. g., sesame oil) or synthetic fatty acid esters (e. g., ethyl oleate or triglycerides). Aqueous injection suspensions may contain substances which increase the viscosity of the suspension include, for example, sodium carboxymethyl cellulose, sorbitol, and/or dextran. Antiestrogens are typically characterized as having limited solubility, therefore the use of agents such as dimethylformamide increases

the solubility of such agonists/antagonists thus increasing their effect on, in this instance, ER (3 or its isoforms. Sasson et al., J. Steroid Biochem. 29 (5): 491 (1988). Optionally, the suspension may also contain stabilizers. Liposomes can also be used to encapsulate the agent for delivery into the cell. For certain drugs such as tamoxifen, agents (e. g., acetone and polyethylene glycol 4000) may be required to enhance the drug's solubility. Cotreau- Bibbo et al., Phaim. Sci. 85 (11): 1180 (1996).

The pharmaceutical formulation for systemic administration according to the invention may be formulated for enteral, parenteral or topical administration. Indeed, all three types of formulations may be used simultaneously to achieve systemic administration <BR> <BR> <BR> <BR> of the active ingredient. Suitable formulations for oral administration include hard or soft : gelatin capsules, pills, tablets, including coated tablets, elixirs, suspensions, syrups or inhalations and controlled release forms thereof.

In practicing the methods of this invention, the compounds of this invention may be used alone or in combination with other therapeutic or diagnostic agents. In certain preferred embodiments, the compounds of this invention may be co-administered along with other compounds typically prescribed for these conditions according to generally accepted medical practice.

L. Gene therapy expression of ERpc The ER,3 gene, for example ERß-3 gene and the ERß-3 protein can also serve as a target for gene therapy in a variety of contexts. For example, in one application, ER (3-3 deficient animals can be generated using standard knock-out procedures to inactivate a ER, & 3 gene. In such a use, a non-human mammal (e. g., a mouse or a rat) is generated in which the ERß3 gene is inactivated or deleted. This can be accomplished using a variety procedures known in the art, such as targeted recombination. Once generated, the E/ deficient animal can be used to (1) identify biological and pathological processes mediated by the ER, a-3 gene; (2) identify proteins and other genes that interact with Ergs3 ; (3) identify agents that can be exogenously supplied to overcome ERß3 deficiency ; and

(4) serve as an appropriate screen for identifying mutations within ERß3 gene that increase or decrease activity.

In addition to animal models, human ERßc deficiencies or mutations can be corrected by supplying to a patient a genetic construct encoding the necessary ERpe protein.

A variety of techniques are presently available, and others are being developed, for introducing nucleic acid molecules into human subjects to correct genetic deficiencies and mutations. Such methods can be readily adapted to employ the ERßC encoding nucleic acid molecules of the present invention.

In another embodiment, genetic therapy can be used as a means for modulating an ERpe mediated biological or pathological process. For example, during osteoporosis, it may be desirable to introduce into the patient a genetic expression unit that encodes a modulator of ERßC mediated transcriptional regulation, such as a nucleic acid molecule that is antisense to the ERp (3 mRNA. Alternatively, tissue specific co-activators or co-repressors could be identified and introduced into a recipient to augment modulation of ERßC or its isoforms.

Such a modulator can either be constitutively produced or inducible within a cell or specific target cell. This allows a continual or inducible supply of a therapeutic agent within the patient.

M. Conformational analysis of ERßC using antibodies Using site-specific polyclonal and monoclonal antibodies against the DNA binding domain (DBD) of the ERpc protein, one can determine the active state of the protein. On this basis, the user is able to determine whether the DBD of ERßC is present in a functional or non-functional altered state and whether the ERp protein has been activated or not. The invention, therefore, includes specifically prepared immunogens, polyclonal antisera and monoclonal antibodies which bind specifically to the DBD of ERßC or its isoforms, and immunoassays employing these site-specific antibodies with cellular samples on a functional and correlative test basis, as described above. Similar procedures and methods have been utilized in determining whether ERa is in its active or inactive state. Wotiz et al.

Without further description, it is believed that one of ordinary skill in the art can, using the preceding description and the following illustrative examples, make and utilize the compounds of the present invention and practice the claimed methods; additionally, all the preceding description involving E-3 or, alternatively spliced isoforms of the complete gene, (e. g., ER, Ergs2 and ERß4) can be applied to their analogs in other mammalian species. The following working examples therefore, specifically point out preferred embodiments of the present invention, and are not to be construed as limiting in any way the remainder of the disclosure. Other generic configurations will be apparent to one skilled in the art.

EXAMPLES Example 1 Cloning of the complete murine mERß3 cDNA The mERß3 clone was twice isolated using two separate procedures: (1) reverse transcriptase PCR (RT-PCR) of mRNA, and (2) amplification from a mouse embryonic stem (ES) cell genomic DNA library. A mouse ES cell genomic DNA library was screened using a cDNA probe and RT-PCR. The oligonucleotides chosen, corresponded to regions in the D and E domains of rat ERß ; published by Kuiper et al., (1996). These oligonucleotides were: 5'-ATG ACA TTC TAC AGT CCT GCT GTG ATG-3' (Primer 1) and 5'-GAA GTG AGC ATC CCT CTT TGC GTT TGG-3' (Primer 2). Using these oligonucleotides five clones were obtained. Two primers were chosen in these genomic DNAs, one around the first ATG, which is 192 bp upstream from the published ATG (Kuiper et al., (1996); Mosselman et al., (1996); and Tremblay et al., (1997)), 5'-TCT CTG AGA GCA TCA TGT CC-3' (Primer 3), and one around the TGA, 5'-CAG CCT GGC CGT CAC TGT GA-3' (Primer 4). The RT-PCR was performed on 10 and 100 ng samples of mouse ovary RNA using the Titan RT-PCR System of Boehringer Mannheim according to manufacturer's instructions. The amplified products obtained using Primers 3 and 4 underwent a second amplification using: 5'-TGC TCT AGA CCA CCA TGT CCA TCT

GTG CCT CT-3' (Primer 5) and 5'-CCG GAA TTC TCA CTG TGA CTG GAG GTT CTG 3' (Primer 6). The products obtained using Primers 6 and 7 were then inserted into Bluescript) vector. The same conditions were used to clone mERßl, mERß2 and mERß 3.

The mER, ß3 clone was also isolated from mRNA using the Marathon RT-PCR system from Clontech. For 5'RACE, poly A+ RNA was prepared from total RNA derived from mouse ovaries according to the methods described in Sambrook et al., (1989).

Approximately 0.5 gg of the poly A+ RNA was reverse transcribed using 200 U Superscript II exogenase- (exo-) using the Marathon cDNA synthesis primer, 5'-TTC TAG AAT TCA GCG GCC GC (T3o)-3', according to manufacturer instructions (GIBCO). The second strand synthesis and all subsequent steps, except PCR, were performed according to the conditions described by Marathon. The cDNA (0.5 Ill of a 10 p1 reaction) was then amplified using the Marathon adaptor primer, 5'-CCA TCC TAA TAC GAC TCA CTA TAG GC-3', with one of two gene specific reverse primers in the presence of Advantage Taq polymerase: 5'-GCA GTA GCT CCT TCA CCC G-3' (Primer 7) or 5'-GCA CTT CAT GCT GAG CAG-3' (Primer 8). The following four step thermocycling program was used to amplify the two products: (1) 5 cycles, 30 sec at 94°C, 4 min at 72°C; (2) 5 cycles, 30 sec at 94°C; (3) 25 cycles, 4 min at 70°C; and (4) 20 sec at 94°C, 4 min at 68°C. Single, predominant amplicons corresponding to the 5'end of the cDNA were then digested with restriction enzymes, cloned and sequenced. The clone was then inserted into a Bluescript vector as described above.

Once the mERß3 gene was cloned, both the nucleic acid and the amino acid sequences were deduced for the complete estrogen receptor P sequence (see Figs. la and b).

In addition to having the novel 192 bp sequence located in the 5'terminus of exon 1 and the newly described exon 5B, nucleotide 1,244, an adenine, in exon 6 of the mERß3 sequence differs from the guanine (nucleotide 1,009) found in the sequence by Tremblay et al., (1997). The mERß3 gene is 1,704 nucleotides long and encodes a 567 amino acid protein.

Using these techniques, it was ascertained that the 5'end of mERß3 included an additional open reading (192 bp) frame as well as a ninth exon, exon 5B (54 bp); both of these sequences are not found in ERß ;. Kuiper et al., (1996); Mosselman et al., (1996); and Tremblay et al., (1997). Exon 5B is located in the ligand binding domain (LBD) of mERß3 and likely plays a significant role in mERß-3 function.

Example 2 Isolation of three alternatively spliced isoforms In addition to the complete murine estrogen receptor (3 gene sequence, mERß-3, two other alternatively spliced murine forms of mERß3 were identified (mERßl and mERß2,) as well as a fourth alternatively spliced isoform isolated from rat ovaries, rERß4. The first alternatively spliced form of mERß-3, mERß-1, contains the novel 192 bp at the 5'terminus of exon 1, but lacks the 54 bp of exon 5B; it is 1,650 nucleotides in length and putatively encodes a 549 amino acid long polypeptide (Fig. 2a). Preliminary data indicates that the mERßl isoform may be more active than the full length mERß3. The mERßl isoform was isolated using both methods described for the isolation of mERß3.

Isoform mERß2 is composed of 1,533 bp, which would encode 510 amino acids (Fig. 2b); mERß2 lacks exon 3, which contains 117 bp. The mERß2 isoform was isolated only from the mouse ES cell genomic library.

Isoform rERß4 was obtained from rat (r) ovaries whereas mER, and mERß2 as well as the full length mERß-3 were obtained from mouse (m) ovaries; it is 1,570 nucleotides in length and contains exon 5B, but exon 6 is deleted. Exon 6 is comprised (as shown in Fig. la) of 134 bp. The putative protein product of rERß4 would be 414 amino acids (Fig. 2c).

All the nucleic acid sequences discussed relate to the coding regions and sequences for the corresponding mRNAs would be longer in both their 5'and 3'regions. It is likely that the published incomplete estrogen receptor P genes (ERßi) isolated from human, rat and mouse libraries are splice variants of this complete form, which in mice is mERß3, and contains the 54 bp of exon 5B and the 192 bp located at the 5'terminus of exon 1.

Mosselman et al., (1996); Kuiper et al., (1996); and Tremblay et al., (1997). All four

sequences contain the 192 bp located at the 5'terminus of exon 1 and not described in the previously published sequences. Id. A sequence similar to the novel 192 bp region located in the 5'terminus of exon 1 may also exist in human ER (3 and its isoforms.

The alternatively spliced isoforms (e. g, mERßl, mERß2 and rERß4) of the full length murine En., gene, mERß3, were twice isolated using the same two different procedures used to acquire mERß3. The primers used in both Examples 1 and 2 were selected based on the assumption that variants, if any, would occur within the boundaries of these selected primers. Historically, similar primers have produced analogous results with ERa.

Once the isoforms were isolated the DNA sequences could be sequenced and the amino acid sequence encoded by each could be determined. The proteins for the three alternatively spliced isoforms are shown in Figures 2a, 2b and 2c.

Example 3 Tissue specific expression of mERß-3 protein using Western Blotting Anti-peptide antibodies raised against a sequence specific to the mouse ERRe (mER (3-3) specifically recognized a protein of 64 kDa in ovary and in bone, as well as in other tissues. Two anti-peptide antibodies were raised in chickens to N- CSSEDPHWHVAQTKSAVPR-OH (Antibodies 1067 and 1068); this polypeptide is encoded by exon 5B and recognizes mERß-3 as well as the isoforms that express the exon 5B coding region. Antibody 1067 and 1068 were obtained from the eggs of two different chickens, as were antibodies 1069 and 1070. These antibodies recognize the protein produced by mER, but not the ERp, protein discovered by Kuiper et al., (1996 and 1997), which lacks exon 5B.

Total proteins (60 µg) obtained from ovarian tissue or bone tissue samples were resolved by electrophoresis in 10% SDS acrylamide gels; the gels were electrophoresed for 16 hours at 40 V. The proteins were transferred from the gels onto nitrocellulose membranes; the transfer was done for 4 hours at 100 mA. The blots were probed using a 1: 1,000 dilution of the chicken antisera to mERß-3 (Antibody 1068) in conjunction with a 1: 1,000 dilution of a secondary antibody conjugated to horseradish peroxidase (Promega).

The proteins were visualized using the ECL chemiluminescent substrate, and exposed to film (BMR film, Kodak) for one minute.

Figure 3 shows the results of the Western blot obtained using Antibody 1068, which detects the polypeptide encoded by exon 5B. Total protein (60 ug) was resolved by electrophoresis. The proteins were transferred to nitrocellulose membrane and probed with a 1: 1,000 dilution of Antibody 1068 (Fig. 3a). Figure 3 (b) is the blot probed with antibody 1068 pre-immune sera. The protein extracts of each lane of both Figures 3 (a) and 3 (b) are: lane 1, human ovary; lane 2, mouse ovary; lane 3, rat ovary; lane 4, ROS 17/2.8 cells; lane 5, ROS 17/2.8 cells treated with 100 nM estradiol for 16 hours; lane 6, murine primary osteoblasts. The antibody specifically recognizes a 64 kDa protein, which closely approximates the predicted size of mERß-3. The question mark refers to a protein migrating at approximately 58 kDa that may be immune specific but is otherwise unidentified. ROS 17/2.8 cells are a line characterized by Gideon Rodan; it is a rat osteoblast-like osteosarcoma cell line (ROS).

Example 4 Tissue specific expression of rat ER., U determined by Southern Blotting of RT-PCR Products Cell expression of rat (r) ER, (rERß) mRNA was examined Southern blotting of RT-PCR products. Total RNA (2-100 ng) obtained from rat ovary, rat total bone marrow (100 ng), and ROS 17/2.8 cells (100 ng) were reverse transcribed using 200 U of Superscript (exo-) reverse transcriptase (Gibco-BRL) and 100 pmol random hexamer probe according to the manufacturer's recommended conditions. ROS 17/2.8 cells are a rat osteoblast-like osteosarcoma cell line (ROS). The rat cDNA was then amplified by PCR in 100 ul reactions using 2 U Taq polymerase and 1 RM 5'-GTC AAG TGT GGA TCC AGG-3' (Primer 9; beginning at base 924 of Accession U57439 and corresponding to base 700 of mERß3) and 5'-GCT CAC TAG CAC ATT GGG-3' (Primer 10; beginning at base 1,130 of rERßj by Kuiper et al., Accession U57439, and corresponding to base 906 of mER (3-3) per each individual reaction. Products were amplified using 25-40 cycles of the following amplification program: 90°C x 1 min; 55°C x 45 sec; 72°C x 2 min. The product was allowed to be extended at 72°C x 5 min at the end of the program

Following amplification, the PCR products were resolved in a 4% NuSieve agarose (FMP)/TBE gel; the DNA was transferred to nylon membranes (Boehringer Mannheim) and cross-linked by UV irradiation for Southern analysis. Ten pmol of an oligonucleotide internal to the predicted amplicon 5'-AGC AGG TAC ACT GCC TGA GCA AAG CCA AGA-3' (Primer 11; beginning at base 991 of Accession U57439 and corresponding to bases beginning at 767 of mER>3) was end-labeled using T4 polynucleotide and used to probe the immobilized DNA amplicon. Following pre-hybridization at 58°C in Quick-Hyb hybridization solution (Stratagene), the probe was added and allowed to hybridize for 1 hr.

The blot was then washed twice with 2x SSC containing 0.1% SDS at room temperature, and then twice with 0.1 X SSC and 0.1% SDS at 58°C. The blot was then exposed to film.

Figure 4 is an autoradiograph of Southern blot of rat ERP (rERß) products amplified by RT-PCR. Total RNA from a variety of tissues was reverse transcribed, amplified by PCR, transferred to nylon membranes and probed using a 32P labeled mER (3-3 oligonucleotide.

Figure 4 (a) was amplified for 35 cycles. Each lane in Figure 4 (a) contains the following types and amounts of RNA: lane 1, control, no RNA; lane 2, rat ovarian RNA (0.1 ug); lane 3, ROS 17/2.8 cells (0.1 ug); lane 4, rat ovarian RNA control (0.1 µg), no reverse transcriptase (rat); and lane 5, ROS 17/2.8 total RNA (0.1 gag), no RT.

Figure 4 (b) is a Southern blot of total RNA. The ERP products were amplified for 25 cycles by RT-PCR. Each lane in Figure 4 (b) contains the following types and amounts of RNA: lane 1, control, no RNA; lane 2, rat ovarian RNA (2 ng); lane 3, ROS 17/2.8 total RNA (0.1 ug), lane 4, total (cultured) bone marrow RNA (0.1 ug); lane 5, total cultured bone marrow RNA (0.1 ug) where the cells had been treated with estradiol for 16 hours; lane 6, total RNA from primary osteoblasts in co-culture (0.1 ug); lanes 7-11, control reactions for lanes 2-6, respectively.

Discrimination analysis for the relative expression of ERßC isoforms may be done utilizing random primers and reverse transcriptase (RT) to synthesize the cDNA from various rat or mouse or other mammalian tissues. The cDNAs so obtained are then amplified by PCR using the completely homologous rat and mouse primers 5'-GTC AAG TGT GGA TCC AGG-3' (Primer 9), which corresponds to base 700 of mERß-3 or base 924 of Rattus norvegicus estrogen receptor P mRNA, accession U57439 (Kuiper et al., 1996),

and 5'-GCA CTT CAT GCT GAG CAG-3' (Primer 8) corresponding to base 1,554 of mERß-3 and 1,724 of accession U57439. Following amplification, the PCR products are purified and digested with Fsp I, a restriction endonuclease with a consensus site within exon 5B (TGCGCA at base 1,176 of mERß and also present in rERß-4). Digestion of the mouse or rat amplicons bearing the exon 5B sequence thus yields smaller products. The digested PCR products are resolved by agarose gel electrophoresis, transferred to nylon membranes, and probed with complementary oligonucleotide probes specific to either rat or murine sequences, or both. The specific sizes of the hybridized DNA present determines what isoform is present in a particular tissue or cell sample. Additionally, the intensity of the band allows quantitation of the relative abundance of the isoform (s) in a particular sample.

Example 5 Gel Shift Assavs Gel shift analysis of mERß-3 is demonstrated in Figure 5 (a). The results obtained by the mERß-3 gel shift (Fig. 5a) were compared to that obtained for the human estrogen receptor alpha (ERa) form, as displayed in Fig. 5 (b). The receptor-DNA complexes formed were then disrupted using anti-peptide antibodies directed toward the novel exon 5B (Antibodies 1067 and 1068). Nuclear extracts (16 ug) derived from COS-7 cells transfected with expression plasmids containing mERß-3 (Fig. 5a) or human alpha estrogen receptor (pHEGO) (Fig. 5b) were incubated with 5 fmols of 27 base pair perfect ERE end-labeled with 32p isotope. The description for the lanes in both figures are the same. The lanes for both Fig. 5 (a) and (b) contain the following: lanes 1 and 2, extract alone; antibody 1067, lanes 3 and 4; antibody 1067 pre-immune serum, lanes 5 and 6; antibody 1068, lanes 7 and 8; antibody 1068 pre-immune serum, lanes 9 and 10; lanes 11 and 12 are control lanes that contain 16 ug ofuntransiected COS-7 nuclear extract.

Example 6 Relative Affinities of Various Estrogens for ER Subtypes

The experimental data displayed in Table I demonstrates the different affinities that various estrogens have for the ER ? subtypes (e. g., ERa, ERß-1 and ERß-3).

E2 binding affinity was determined by incubating transfected COS-7 cell cytosol with different concentrations of [3H]-E2 (0-200nM) and with or without unlabeled E2 500X for 4 h at 4°C in 40 mM Tris HCl pH 7.4,150 mM KCI, PMSF 0.1 mM, DTT 2 mM.

COS-7 cells were transfected as described in Example 7. Bound receptor was separated by the hydroxy apatite method (Obourn et al., Biochemistrv 32 (24): 6229-6236 (1993)) or the ligand was removed by the dextran coated charcoal method (Garcia et al., Mol. Endocrinol.

6 (12): 2071-2078 (1992)), and bound hormone measured by liquid scintillation counting.

Dissociation constants (kd) were obtained by Scatchard plots. Similarly, the relative affinities of different estrogenic ligands were determined using 1 nM receptor and 5 nM (3H]-E2 with or without various concentrations of the described competitor steroids. The concentration of ligand necessary to displace 50% of the bound, labeled [3H]-17p-estradiol from the receptor, was used in the denominator to express the values shown in table (with a constant value of 1 nM in the numerator).

The differences in the relative affinities (results of the IC-50 experiments) show that the mERß-1 and mER (3-3 receptors have different affinities for different ligands. This suggests that the transcriptional responsiveness to different ligands is a function of both the expression pattern of the receptor sub-type, and the estrogenic ligand used to stimulate transcription.

Table I shows the different affinities of estrogens to human ERa, mouse ERß-1 and mouse ERß-3 (which contains exon 5B). As indicated, the affinity of the different estrogens varies as to the receptor. The larger the number, the greater the affinity the estrogen has for the estrogen receptor target. Diethylstilbestrol (DES) has a greater affinity for the ERP isoforms than for ERa.

TABLE 1. Relative Affinities of Various Estrogens for ER Subtypes

gen tRO-3 M66 17 Estradiol 100 100 100 (standard) Kd = 0. 40 Kd = 0. 57 Kd = 7.14 Diethylstilbestrol 283 432 1009 Ethynyl Estradiol 216 34 189 RU58668 208 111 475 Raloxifene 143 22 279 RU39411 126 147 900 Moxestrol 87 16 132 4 OH-Tamoxifen 49 94 291 Estriol 5 14 38 17-a-estradiol 2. 5 0. 8 7 Tamoxifen 2 5 10 Estrone 1. 3 2. 3 21. 5 The method used in this experiment can be utilized for screening reagents with different affinities for each of the ERP isoforms and comparing them to the ERa for determination of the affinity a particular drug may have for the other estrogen receptor proteins and their isoforms.

Example 7 Transactivation Profiles of ERß-1 and ERß-3 Isoforms This experiment assessed the effect of ERß-1 and ERß-3 isoforms when expressed both individually and when expressed together as compared to the effect of ERa.

The ability of estrogens to stimulate transcription via an estrogen response element (ERE) functionally linked to tk-CAT (a construct described by Metzger et al., J. Biol. Chem.

270 (16): 9535 (1995)) was measured by transient transfection of the expression vectors for mER (3-1 and mER (3-3 in COS-7 cells. For transfection, COS-7 cells were seeded into six-well plates in phenol red free, low glucose DMEM. At approximately 50-80% confluency, the cells were transfected using lipofectamine according to the manufacturer's instructions (GIBCO-BRL). The expression constructs were transfected with a total of 2 Rg DNA containing 500 ng of reporter, 100-500 ng expression plasmid, and the remainder (1-1.4 ug) as pBluescript as a carrier DNA. After 24 h, the cells were washed with DMEM and replaced with fresh medium containing drug (17-P estradiol, 4-hydroxy tamoxifen, clomiphene or DES at 1-300 nM concentrations) or vehicle (ethanol). After 24 h the cells were lysed, and the CAT activity determined by liquid scintillation counting of converted chloramphenicol (as described in CURRENT PROTOCOLS IN MOLECULAR BIOLOGY). For co-transfection of both mERß-1 and mERß-3 isoforms, equivalent amounts of expression constructs were transfected (usually 100 ng each) (Fig. 8). To analyze the effect of other drugs on transcription, similar experiments were performed as described above, with the exception that drug concentrations were varied (Fig. 9).

The results of transfection analysis (Fig. 8) show that mER (3 is capable of stimulating transcription from a reporter containing a canonical responsive element. The mER (3-1 can stimulate transcription to approximately 50-70% of that observed in similar cells transfected with the ERa construct, pHEGO, at estradiol concentrations of 100 nM.

The mERß-3 isoform is capable of stimulating transcription to only approximately 40% of that observed in pHEGO at the 100 nM drug concentration.

As can be seen in Figures 8 and 9, the mERß-1 responsiveness is similar to that observed in cells transfected with pHEGO, which encodes ERa. mERß-3 only stimulated transcription at very high estrogen concentrations (100-200 nM). By contrast, when both receptors are co-expressed together, the magnitude of the response is ablated, showing that mERp-3 functions as a dominant negative modulator of the action of mERß-1. (Similar

results were corroborated in an article published by Maruyama et al., Biochemical Biophysical Research Communications 246 (1) : 142-147 (1998)).

The results in Figure 8 show that mERp-1 has a transactivation profile similar to ERa when exposed to E2, clomiphene (clomid), diethylstilbestrol (DES) and 4-OHT. The mERß-3 isoform has a decreased ability to transactivate cV2ERE as compared to either ERa or mERß-1. However, the transactivation activity is reduced when the isoforms are co- expressed (Fig. 8, panel indicated as mER B 1+B3). The assay utilized in this example can be similarly used to determine what agents can modulate homodimers of ERR isoforms, as well as heterodimers of the ERP isoforms or heterodimers composed of ERR and ERa isoforms.

Figure 9 demonstrates that ERß-1 (displayed as B1 in Fig. 9) and ERß-3 (B3) both possess similar activity when exposed to clomiphene, DES, 4-OHT, and E2. However when ERß-l and ERß-3 are co-expressed in a reporter system, their activity is down regulated as compared to individual expression of the ERP isoforms or to ERa. This assay system can be utilized to screen other estrogens or compounds that modulate the activity of the various ER (3 isoforms.

Example 8 In Situ Hybridization of various tissues In situ hybridization analysis was performed using anti-sense cRNA probes to both the ERa and ERR to localize the message for each of the ER subtypes. The tissue was treated with 0.1 M TEA, pH 8.0, plus 0.25% acetic anhydride for 10 min at room temperature, rinsed three times in 2X SSC, dehydrated through a series of alcohols and air dried. cRNA riboprobes corresponding to the ERa or ER (3-3 isoforms were prepared and used to probe tissue sections. The hybridization solution was removed, the sections washed and air dried. For riboprobes, an 801 base pair insert corresponding to the ligand binding domain of the mERß-1 plasmid (bases 931-1731 of the rat sequence) was linearized using the restriction enzyme ApaLI and transcribed using RNA polymerase in vitro in the presence of [35S]-UTP and [35S]-CTP according to methods of Goldstein et al., Neuroscience 71 (1): 243 (1996). The riboprobes were purified by ethanol precipitation, and the dried tissue

sections hybridized with probe in hybridization buffer overnight at 55°C. The hybridization solution was removed, the sections were incubated briefly with RNase, then washed, dehydrated, and air dried. The dried sections were exposed to film for normalization of subsequent exposure times and dipped in NTB3 emulsion to determine the cellular and anatomical localization of each mRNA.

The results demonstrate abundant and wide spread distribution of the ERP message within the developing ovarian follicle (Fig. 10, top panel) and in the lung, kidney cortex, and specific regions of the brain (not shown). The pattern of distribution of ERa was quite different and was highly expressed only in the uterus (Fig. 10, middle panel), in the medullary regions of the kidney (not shown) and specific regions of the brain. Preliminary data also indicated that ERP is expressed in ossification center that appear to correspond with mesenchymal condensation zones in developing rat bone (12 days), especially in the spine (Fig. 10).

The ERP message is observed in developing Graffian follicles (GA), but not in resorbing follicles (FA) undergoing atresia (Fig. 10, top panel, antisense). In the ovary, the ERa message receptor was only abundant within the uterine tube (not shown). ERa mRNA was observed to be widely expressed throughout the uterus (Fig. 10, middle panel, antisense). In the cervical spine, the ERP mRNA was localized to zones of mesenchymal ossification (M) similar to the expression patter of Osf/Cbfal, an osteoblast differentiation protein (Fig. 10, bottom panel, antisense, arrows). Controls corresponding to serial sections hybridized using sense riboprobe controls are also shown in the panels on the right.

Example 9 Methods of Screening for Drugs A. Phosphorylation of ERß. Most of the members of the steroid receptor superfamily, including ERa, undergo post-translational modifications (e. g., phosphorylation) as a function of their basal state or in response to ligand binding. With ERa, there are a variety of sites on the molecule that are phosphorylated in response to ligand binding. Post-translational modification of mERß or human EFB can be accomplished using the same methods as previously utilized for ERa. Methods of

analyzing phosphorylation include transient or stable expression of the various cDNA constructs in COS-7 cells, or by immunoprecipitation of [32P]-labeled ERP from cells metabolically labeled with [32P]-orthophosphate. Tryptic maps from ligand stimulated or unstimulated cells can be obtained using ERP proteins isolated by immunoprecipitation of the mERß or human ERP molecule using our antibodies (e. g., directed towards products of exon 5B such as the antibody used to obtain Fig. 3) or commercially available antibodies.

For studies performed using in vitro transient transfection, a triple-myc tag or GST tag can also be linked to the carboxyl or amino termini by cloning the appropriate coding sequence into the expression plasmid. The expressed (phosphorylated) protein can then be immunoprecipitated using a very reliable, and commercially available anti-myc antibody (if using the triple-myc tag) or anti-GST antibodies. In addition, exon 5B amino acid residues can be substituted with other residues to prevent phosphorylation. In contrast to either ERa or mERß-1, exon 5B, which is unique to rnERß-3, is located within a region of the molecule that otherwise is extremely hydrophobic. The exon 5B region, however, is unusually hydrophilic and contains a consensus casein kinase II (CKII) phosphorylation site (VLDRSSEDP) that arises as direct consequence of the location of the exon 5-exon 5B-splice junction. Many of the steroid receptors, including the ERa subtype, are phosphorylated on CKII sites. The serines present in the portion of ERP encoded by exon 5B can be substituted with alanine residues (or other uncharged amino acids) or with residues which mimic constitutively phosphorylated molecules (e. g., aspartic acid residues).

Such forms of En can be utilized in screening and isolating drugs which modulate the activity of the various ER (3 isoforms. Alternatively, these mutant forms of ERR or polypeptide fragments containing this region can themselves be tested for agonist or antagonist activity in the ERP signal pathways.

B. Domain Switching. The amino terminus of the ERa contains an autonomous transcriptional activity (AF-1) that is only fully active when"integrated"with the ligand-dependent transcriptional domain (AF-2) present within the ligand binding domain of the ERa molecule. While these domains have yet to be described for the ERR molecule, the high degree of sequence homology at the protein level between ERa and ERP molecules logically suggests that ERP is similarly organized. Many of these domains have been

identified and characterized using portions of the ligand binding domain (LBD) fused to convenient and reliable epitope tags, such as GST and myc. Such constructs can then be utilized to identify, in whole cell lysates or other expression models such as expression libraries, proteins that functionally alter the transcriptional responsiveness of the ER complex. We postulate that specific integrator molecules may be found using the LBD of mERß-3 fused to such convenient epitope tags as probes for protein-protein interactions.

Such proteins can then alter the transcriptional responsiveness of the functional ER complex, (defined as the homo-dimers of ERß3 with ERß3 or hetero-dimers of ERß3 with ERßl, or ERp3 with ERa) portion of the amino terminus fused with such epitope tags as probes for proteins that interact with the ER-complex. These complexes in turn can be used in drug screening assays to identify drugs which modulate ERP isoform activity. Alternatively the complexes themselves may be used to regulate pathways mediated by estrogen receptors.

REFERENCES The following references and all articles, texts and patents referred to above and below, are hereby incorporated by reference in their entirety: Aguilar et al., Mol. Cell. Biochem. 136 (1): 35 (1994).

Angereretal., MethodsEnzvmol., 152: 649 (1987).

Aprikian et al., Cancer 71 (12): 3952 (1993).

Bartlett et al., Br. J. Cancer 73 (12): 1538 (1996).

Berent et al., Biotech 3: 208 (1985).

Berg et al., PNAS 88: 4723 (1991).

Berkowitz and Evans, J. Biol. Chem. 267 (10): 7134 (1992).

Berry et al., EMBO 9: 2811 (1990).

Bonetti et al., Breast Cancer Res. Treat. 38 (3): 289 (1996).

Burch et al., Mol. Cell. Biol. 8 (3): 1123 (1988).

Byers et al., Mol. Endocrinol. 11 (2): 172 (1997).

Cohen et aL, PNAS 69: 2110 (1972).

Collins et al., Steroids 62: 365 (1997).

Cotreau-Bibbo et al., J. Pharm. Sci. 85 (11): 1180 (1996).

CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, F. Ausubel et al., ed. Greene Publishing and Wiley-Interscience, New York (1987,1995).

Delmas et al., J. Clin. Oncol. 15: 955 (1997).

Engelhardt et aL, J. Am. Coll. Cardiol. 27 (1): 146 (1996).

Evans et al., Bone 17 (4S): 1815 (1995).

FormsanetaL, Cell 81 (4): 541 (1995).

Gaido et al., Toxicol. Appl. Pharmacol. 143 (1): 205 (1997).

Gallo etal., Semin. Oncol. 24: S 1 (1997).

Garcia et al., Mol. Endocrinol. 6 (12): 2071 (1992) Garnier et aL, J. Mol. Bio. 120: 97 (1978).

Goethuys et al., Am. J. Clin. Oncol. 20 (1): 40 (1997).

Goldstein et al., Neuroscience 71 (1): 243 (1996).

Graham et al., Virology 52: 456 (1973).

Grasser et al., J. Cell Biochem. 65: 159 (1997).

Green et aL, Cold Spring Harbor Svmposia on Quantitative Biology 51 (2): 751 (1986).

Guldenaar et al., Brain Res. 700 (1-2): 107 (1995).

Habenich et al., J. Steroid Biochem. Mol. Biol. 44: 557 (1993).

Hafner et al., J. Steroid Biochem. Mol. Biol. 58 (4): 385 (1996).

Harlow and Lane, ANTIBODIES: A LABORATORY MANUAL, Cold Spring Harbor Pubs., N. Y.

(1988).

Hopp and Woods, PNAS 78: 3824 (1981).

Hughes et al., Nat. Med. 2 (10): 1132 (1996).

Ichinose et al., Gene 188: 95 (1997).

Jackson et al., Mol. Endocrinol. 11: 693 (1997).

Kangas, Acta Oncol. 31 (2) : 143 (1992).

Kohler and Milstein, Eur. J. Immunol. 6: 511 (1976).

Komm et al., Science 241 : 81 (1988).

Kraus et al., Mol. Cell. Biol. 15 (4): 1847 (1995).

Kuiper et al., Endocrinologv 138 (3): 863 (1997).

Kuiper and Gustafsson, FEBS 410: 87 (1997).

Kuiper et al., PNAS 93: 5925-30 (1996).

Maas et al., Cancer Lett. 97 (1): 107 (1995).

Maniatis et al., MOLECULAR CLONING, A LABORATORY MANUAL, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1982).

Maruyama et al., Biochemical Biophvsical Research Communications 246 (1): 142 (1998) Metzger et al., J. Biol. Chem. 270 (16): 9535 (1995).

Morgan, U. S. Patent No. 4,732,904 (1988).

Mosselman et al., FEBS Letters 392: 49 (1996).

Muhn et al., Ann. N. Y. Acad. Sci. 761: 311 (1995).

Nawaz et al., Gene Expr. 2 (1): 39 (1992).

Obourn et al., Biochemistrv 32 (24): 6229 (1993).

Onate et al., Science 270 (5240): 1354 (1995).

Palkowitz et al., J. Chem. Med. 40 (10): 1407 (1997).

PCR PROTOCOLS: A GUIDE TO METHODS AND APPLICATIONS, Innis, M, Gelfand, D., Sninsky, J. and White, T., eds., (Academic Press, San Diego 1990).

Queen et al., U. S. Patent No. 5,585,089.

Raval, J. Pharmacol. Toxicol. Methods 32 (2): 125 (1994).

Riechmann et aL, Nature 332: 323 (1988).

Safarians et al., Cancer Res. 56 (15): 3560 (1996).

Sambrook et al., MOLECULAR CLONING: A LABORATORY MANUAL (2nd ed.), Vols. 1-3, Cold Spring Harbor Laboratory, (1989).

Sasson et al., J. Steroid Biochem. 29 (5): 491 (1988).

Scopes, PROTEIN PURIFICATION: PRINCIPLES AND PRACTICE, Springer-Verlag: New York (1982).

Somjen et al., J. Steroid Biochem. Mol. Biol. 59: 389 (1996).

Southern, J. Mol. Biol. 98: 503 (1975).

Southern et al., J. Mol. Anal. Genet. 1: 327 (1982).

Shughrue et al., Steroids 61 (12): 678 (1996).

Tora et al., Cell 59: 447 (1989).

Tremblay et al., Mol. Endocrin. 11 (3): 353 (1997).

Wang et al., J. Biol. Chem. 270 (40): 23,322 (1992).

Walfish et al., PNAS 94 (8): 3697 (1997).

Wigler et al., PNAS 76: 1373 (1979).

Wotiz et al., U. S. Patent No. 5,312,752 (1994).