WO2000004152A2

WO2000004152A2 - Androgen receptor coactivators

Info

Publication number: WO2000004152A2
Application number: PCT/US1999/016122
Authority: WO
Inventors: Chawnshang Chang
Original assignee: University Of Rochester
Priority date: 1998-07-17
Filing date: 1999-07-16
Publication date: 2000-01-27
Also published as: JP2002520056A; AU2004203835B2; AU2004203835A1; AU771564B2; AU5106999A; EP1098976A2; WO2000004152A3; CA2335101A1

Abstract

Disclosed are androgen receptor-associated proteins, designated ARA24, ARA54, ARA55, and Rb, that have been demonstrated to interact with the androgen receptor to alter levels of androgen receptor-mediated transcriptional activation. Certain of these proteins interact with the androgen receptor in an androgen-dependent manner, whereas certain proteins may induce transcriptional activation in the presence of other ligands, such as E2 or HF. Also disclosed is a method of detecting androgenic or antiandrogenic activity using these proteins in a mammalian two-hybrid transient transfection assay.

Description

ANDROGEN RECEPTOR COACTIVATORS

CROSS-REFERENCE TO RELATED APPLICATIONS Not applicable.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

Not applicable.

BACKGROUND OF THE INVENTION Androgens constitute a class of hormones that control the development and proper function of mammalian male reproductive systems, including the prostate and epididymis. Androgens also affect the physiology of many non-reproductive systems, including muscle, skin, pituitary, lymphocytes, hair growth, and brain. Androgens exert their effect by altering the level of gene expression of specific genes in a process that is mediated by binding of androgen to an androgen receptor. The androgen receptor, which is a member of the steroid receptor super family, plays an important role in male sexual differentiation and in prostate cell proliferation. Binding of androgen by the androgen receptor allows the androgen receptor to interact with androgen responsive element (AREs) , DNA sequences found on genes whose expression is regulated by androgen.

Androgen-mediated regulation of gene expression is a complicated process that may involve multiple co-activators (Adler et al . , Proc. National Acad. Sci . USA 89:6319-6325, 1992) . A fundamental question in the field of steroid hormone biology is how specific androgen-activated transcription can be achieved in vivo when several different receptors recognize the same DNA sequence. For example, the androgen receptor (AR) , the glucocorticoid receptor (GR) , and the progesterone receptor (PR) all recognize the same sequence but activate different transcription activities. Some have speculated that accessory factors may selectively interact with the androgen receptor to determine the specificity of gene activation by the androgen receptor.

Prostate cancer is the most common malignant neoplasm in aging males in the United States. Standard treatment includes the surgical or chemical castration of the patient in combination with the administration of anti-androgens such as 17 β estradiol (E2) or hydroxyflutamide (HF) . However, most prostate cancers treated with androgen ablation and anti-androgens progress from an androgen- dependant to an androgen- independent state, causing a high incidence of relapse within 18 months (Crawford, Br . J. Urology 70: suppl . 1, 1992) . The mechanisms by which prostate cancer cells become resistant to hormonal therapy remain unclear. One hypothesis that has been advanced is that over the course of treatment, a mutation in the AR occurs which alters the receptor's sensitivity to other steroid hormones or anti-androgens , such as E2 and HF, thereby causing the progression from androgen-dependent to androgen-independent prostrate cancer. This hypothesis is supported by transient transfection assays in which it has been shown that anti-androgens may have an agonistic activity that stimulates mutant AR (mAR) -mediated transcription .

Recently, A1B1 was identified as estrogen receptor coactivator that is expressed at higher levels in ovarian cancer cell lines and breast cancer cells than in noncancerous cells (Anzick, et al . Science 277:965-968,

1997) . This result suggests that steroid hormone receptor cofactors may play an important role in the progression of certain diseases, such as hormone responsive tumors.

The identification, isolation, and characterization of genes that encode factors involved in the regulation of gene expression by androgen receptors will facilitate the development of screening assays to evaluate the potential efficacy of drugs in the treatment of prostate cancers.

BRIEF SUMMARY OF THE INVENTION The present invention includes an isolated polynucleotide that encodes a co-activator for human androgen receptor, the polynucleotide comprising a sequence that encodes a polypeptide selected from the group consisting of an ARA54 polypeptide, an ARA55 polypeptide, an ARA24 polypeptide, and an Rb polypeptide.

Another aspect of the present invention is a genetic construct comprising a promoter functional in a prokaryotic or eukaryotic cell operably connected to a polynucleotide that encodes a polypeptide selected from the group consisting of an ARA54 polypeptide, an ARA55 polypeptide, an ARA24 polypeptide and an Rb polypeptide. The present invention provides a method for screening candidate pharmaceutical molecules for the ability to promote or inhibit the interaction of ARs and AREs to modulate androgenic activity comprising the steps of: (a) providing a genetic construct comprising a promoter functional in a eukaryotic cell operably connected to a polynucleotide comprising a sequence that encodes a polypeptide selected from the group consisting of an ARA54 polypeptide, an ARA55 polypeptide, an ARA24 polypeptide, and a retinoblastoma polypeptide; (b) cotransforming a suitable eukaryotic cell with the construct of step a, and a construct comprising at least a portion of an expressible androgen receptor sequence;

(c) culturing the cells in the presence of a candidate pharmaceutical molecule; and

(d) assaying the transcriptional activity induced by the androgen receptor.

It is an object of the present invention to a provide a genetic construct capable of expressing a factor involved in co-activation of the human androgen receptor.

It is an object of the present invention to provide a method for evaluating the ability of candidate pharmaceutical molecules to modulate the effect of androgen receptor coactivators on gene expression.

Other objects, features, and advantages of the present invention will become apparent upon reading the specification and claims.

DETAILED DESCRIPTION OF THE INVENTION Transactivation of genes by the androgen receptor is a complicated system that involves many different coactivators. It is not currently known just how many factors are involved in androgen receptor-mediated regulation of gene expression. The identification and/or characterization of four androgen receptor coactivators is reported herein. Inclusion of one or more of these coactivators in an assay for androgenic and antiandrogenic activity is expected to increase the sensitivity of the assay. Information about these coactivators is valuable in the design of pharmaceutical agents intended to enhance or interfere with normal coactivator function. A preliminary assessment of the efficacy of a potential therapeutic agent can be made by evaluating the effect of the agent on the ability of the coactivator to enhance transactivation by the androgen receptor.

One aspect of the present invention is an isolated polynucleotide that encodes a co-activator for human androgen receptor, the polynucleotide comprising a sequence that encodes a polypeptide selected from the group consisting of an ARA54 polypeptide, an ARA55 polypeptide, an ARA24 polypeptide and an Rb polypeptide. Another aspect of the present invention is a genetic construct comprising a promoter functional in a prokaryotic or eukaryotic cell operably connected to a polynucleotide that encodes a polypeptide selected from the group consisting of an ARA54 polypeptide, an ARA55 polypeptide, an ARA24 polypeptide and an Rb polypeptide.

The present invention includes a method for screening candidate pharmaceutical molecules for the ability to promote or inhibit the ARs and AREs to result in modulation of androgenic effect comprising the steps of:

(a) providing a genetic construct comprising a promoter functional in a eukaryotic cell operably connected to a polynucleotide comprising a sequence that encodes a polypeptide selected from the group consisting of an ARA54 polypeptide, an ARA55 polypeptide, an ARA24 polypeptide, and a retinoblastoma polypeptide; (b) cotransforming a suitable eukaryotic cell with the construct of step a, and a construct comprising at least a portion of an expressible androgen receptor sequence;

(d) assaying the transcriptional activity induced by the androgen receptor gene .

The human androgen receptor is comprised of a ligand binding domain (LBD) , a DNA binding domain (DBD) , a hinge domain containing nuclear localization signals, and a transactivation domain in the hyper-variable N-terminus. Truncation or deletion of the LBD results in constitutive transactivation by the N- terminal domain.

In certain cases, progression of prostate cancer from androgen dependent- to androgen independent -stage may be caused by a mutation in the LBD that alters the ligand specificity of the mAR (Taplan et al . , New Engl . J. Med . 332:1393-1398 (1995); Gaddipati et al . , Cancer Res . 54:2861-2864 (1994)) . We examined whether differential steroid specificity of wild type (wt) AR and mAR involves the use of different androgen receptor-associated (ARA) proteins or coactivators by these receptors.

As described in the examples, a yeast two-hybrid system with mART887S as bait was used to screen the human prostate cDNA library. The sequences of two clones encoding a putative coactivators (designated ARA54 and ARA55) are shown in SEQ ID NO : 1 and SEQ ID NO : 3 , respectively. The putative amino acid sequences of ARA54 and ARA55 are shown in SEQ ID NO : 2 and SEQ ID NO : 4 , respectively. Also provided are the DNA and amino acid sequences of ARA24 (SEQ ID NO: 5 and SEQ ID NO : 6 , respectively) and Rb (SEQ ID NO : 7 and SEQ ID NO : 8 , respectively) . These coactivators were further characterized as detailed below. It is expected that some minor variations from SEQ ID NOs:l-8 associated with nucleotide additions, deletions, and mutations, whether naturally occurring or introduced in vi tro, will not affect coactivation by the expression product or polypeptide.

Briefly, ARA54 is a 54 kDa protein that interacts with AR in an androgen-dependent manner. Coexpression of ARA54 and AR in a mammalian two-hybrid system demonstrated that reporter gene activity was enhanced in an androgen- dependent manner. ARA54 functions as a coactivator relatively specific for AR-mediated transcription. However, ARA54 may also function as a general coactivator of the transcriptional activity for other steroid receptors through their cognate ligands and response elements. ARA54 was found to enhance the transcriptional activity of AR and PR up to 6 fold and 3-5 fold, respectively. In contrast, ARA54 has only marginal effects (less than 2 fold) on glucocorticoid receptor (GR) and estrogen receptor (ER) in DU145 cells.

Coexpression of ARA54 with known AR coactivators SRC-1 or ARA70 revealed that each of these coactivators may contribute individually to achieve maximal AR-mediated transcriptional activity. Moreover, when ARA54 was expressed simultaneously with SRC-1 or ARA70, the increase in AR-mediated transactivation was additive but not synergistic relative to that observed in the presence of each coactivator alone.

The C-terminal domain of ARA54 (a. a. 361-471 of SEQ ID NO:l) serves as a dominant negative inhibitor of AR- mediated gene expression of target genes. Coexpression of exogenous full-length ARA54 can reduce this squelching effect in a dose-dependent manner.

ARA54 enhanced transactivation of wtAR in the presence of DHT (10-¹⁰ to 10^'8 M) by about 3-5 fold. However, transactivation of wtAR was enhanced only marginally with E2 (10^"9-10-⁷ M) or HF (10-⁷-10-⁵ M) as the ligand. The ability of ARA54 to enhance transactivation by two mutant receptors (mARt877a and mARe708k) that exhibit differential sensitivities to E2 and HF (Yeh et al . , Proc . Na tl . Acad . Sci . USA, in press (1998)) was also examined. The mutant mARt877a, which is found in many prostate tumors (23), was activated by E2 (10-⁹-10-⁷ M) and HF (10^"7-10-⁵ M) , and ARA54 could further enhance E2- or HF-mediated AR transactivation. In contrast, the mutant mARe708k, first identified in a yeast genetic screening (Wang, C.,Ph.D. Thesis of Universi ty of Wisconsin -Madison (1997) ) , exhibited ligand specificity and response to ARE54 comparable to that of wtAR.

It is expected that any polypeptide having substantial homology to ARA54 that still actuates the same biological effect can function as "an ARA54 polypeptide." With the sequence information disclosed herein, one skilled in the art can obtain any ARA54 polypeptide using standard molecular biological techniques. An ARA54 polypeptide is a polypeptide that is capable of enhancing transactivation of AR in an androgen-dependent manner, enhancing E2 or HF transactivation by the mutant receptor mARt877a, and reducing inhibition of AR-mediated gene expression caused by overexpression of the C-terminal domain of ARA54 (a. a. 361-471 of SEQ ID N0:1) . The sequence information presented in this application can be used to identify, clone or sequence allelic variations in the ARA54 genes as well as the counterpart genes from other mammalian species, it is also contemplate that truncations of the native coding region can be made to express smaller polypeptides that will retain the same biological activity.

The polynucleotide sequence of ARA55 (SEQ ID NO : 3 ) exhibits high homology to the C-terminus of mouse hic5 (hydrogen peroxide inducible clone) (Pugh, B., Curr . Opin . Cell Biol . 8:303-311 (1996)), and like hic5, ARA55 expression is induced by TGFb . Cotransfection assays of transcriptional activation, which are described in detail below, revealed that ARA55 is able to bind to both wtAR and mART887S in a ligand-dependent manner to enhance AR transcriptional activities. ARA55 enhanced transcriptional activation by wtAR in the presence of 10^~9M DHT or T, but not 10^~9M E2 or HF . In contrast, ARA55 can enhance transcriptional activation by mART887S in the presence of DHT, testosterone (T) , E2 , or HF. ARA55 did not enhance transcriptional activation of mARe708k in the presence of E2 , but can enhance transcription in the presence of DHT or T. The C-terminal domain of ARA55 (amino acids 251-444 of SEQ ID NO: 3) is sufficient for binding to ARs , but does not enhance transcriptional activation by ARs.

The invention is not limited to the particular ARA55 polypeptide disclosed in SEQ ID NO:4. It is expected that any ARA55 polypeptide could be used in the practice of the present invention. By "an ARA55 polypeptide" it meant a polypeptide that is capable of enhancing transactivation of wtAR, , the mutant receptor mARt877a, in the presence of DHT, E2, or HF or intact receptor mARe708k in the presence of DHT or T. Such polypeptides include allelic variants and the corresponding genes from other mammalian species as well as truncations.

The AR N-terminal domain comprises a polymorphic poly- glutamine (Q) stretch and a polymorphic poly-glycine (G) stretch that account for variability in the length of human AR cDNA observed. The length of the poly-Q region (normally 11-33 residues in length) is inversely correlated with the risk of prostate cancer, and directly correlated with the SBMA, or Kennedy's disease (La Spada et al . , Nature (London) 352:77-79 (1991)). The incidence of higher grade, distant metastatic, and fatal prostate cancer is higher in men having shorter AR poly-Q stretches. As described in the examples, experiments undertaken to identify potential coactivators that interact with the/ AR poly-Q region led to the isolation of a clone encoding a coactivator, designated ARA24, that interacts with the poly-Q region. The sequences of the ARA24 clone and its putative translation product is shown in SEQ ID NO : 5 and SEQ ID NO: 6.

The ARA24 clone has an ORF that is identical to the published ORF for human Ran, an abundant, ras-like small GTPase (Beddow et al . Proc . Na tl . Acad . Sci . USA 92:3328- 3332, 1995) . Overexpression of ARA24 in the presence of DHT does enhance transcriptional activation by AR over that observed in cells transfected with AR alone. Moreover, expression of antisense ARA24 (ARA24as) does reduce DHT- induced transcriptional activation.

An ARA24 polypeptide is one that interacts with the poly-Q region of an AR. An ARA24 polypeptide is further characterized by its ability to increase transactivation when overexpressed in eukaryotic cells having some endogenous ARA24, but expression of an ARA24 antisense RNA reduces AR receptor transactivation.

Androgen receptor mutations do not account for all cases of androgen-independent tumors, because some androgen- independent tumors retain wild-type AR. A significant percentage of androgen- insensitive tumors have been correlated with reduced expression of retinoblastoma protein (Rb) (Bookstein, et al . , Science 247 :712-715, (1990) ) , expression a truncated Rb protein (Bookstein, et al . Proc . Na tl . Acad . Sci . USA 87:7762-7766 (1990)), or a missing Rb allele (Brooks, et al . Prosta te 26:35-39, (1995) ) . The prostate cancer cell line DU145 has an abnormal short mRNA transcript of Rb exon 21 (Sarkar, et al . Prostate 21:145-152(1992)) and transfecton of the wild- type Rb gene into DU145 cells was shown to repress the malignant phenotype (Bookstein, et al . Proc . Natl . Acad . Sci . USA 87:7762-7766 (1990)).

Rb functions in the control of cell proliferation and differentiation (Weinberg, R.A., Cell 81:323-330 (1995); Kranenburg et al . , FEBS Lett . 367:103-106 (1995)). In resting cells, hypophophorylated Rb prevents inappropriate entry of cells into the cell division cycle. Phosphorylation of Rb by cyclin-dependent kinases relieves Rb-mediated growth suppression, and allows for cell proliferation (Dowdy et al . , Cell 73:499-511 (1993); Chen et al., Cell 58:1193-1198 (1989)). Conversely, dephosphorylation of Rb during Gl progression induces growth arrest or cell differentiation (Chen et al . (1989); Mihara et al . , Science 246:1300-1303 (1989)). In dividing cells, Rb is dephosphorylated during mitotic exit and Gl entry (Ludlow et al . , Mol . Cell . Biol . 13:367-372 (1993)). This dephosphorylation activates Rb for the ensuing Gl phase of the cell cycle, during which Rb exerts it growth suppressive effects.

We investigated the role of Rb in AR transactivation as detailed in the examples . We found that Rb can induce transcriptional activity of wtAR or mARs877t in the presence of DHT, E2 , or HF, and mARe708k in the presence of DHT. We also discovered that Rb and ARA70 transciptional activity act synergistically to enhance transciptional activity of ARs. The sequence of the cloned Rb gene and the deduced amino acid sequence of the ORF are shown in SEQ ID NO: 7 and SEQ ID NO : 8 , respectively. An Rb polypeptide is a polypeptide that is substantially homologous to SEQ ID NO: 8, that interacts with the N-terminal domain of AR, and which acts synergistically with ARA70 in enhancing transactivation by AR. In the examples, various eukaryotic cell types, including yeast, prostate cells having mutant AR and cells lacking AR, were used to evaluate the ability of the putative androgen coactivators to enhance transactivation by AR. It is expected that in the method of the present invention, any eukaryotic cell could be employed in an assay for AR activity. This feature allows the investigator flexibility in designing assays. As described below, cells were transfected using a calcium phosphate technique. It is expected that the method of the present invention could be practiced using any transfection means including, for example, electroporation or particle bombardment.

Changes in the level of transactivation by AR can be assessed by any means, including measuring changes in the level of mRNA for a gene under the control of AR, or by quantitating the amount of a particular protein expressed using an antibody specific for a protein, the expression of which is under the control of AR. Most conveniently, transactivation by AR can be assessed by means of a reporter gene .

As used herein, a reporter gene is a gene under the control of an androgen receptor, the gene encoding a protein susceptible to quantitation by a colormetric or fluorescent assay. In the examples below, a chloramphenicol acetyltransferase or a luciferase gene were used as reporter genes. The gene may. either be resident in a chromosome of the host cell, or may be introduced into the host cell by cotransfection with the coactivator gene. The following nonlimiting examples are intended to be purely illustrative.

EXAMPLES Plasmid construction

A human prostate library in pACT2 yeast expression vector (a gift from Dr. S. Elledge) consists of the GAL4 activation domain (GAL4AD, a. a. 768-881) fused with human prostate cDNA. pSG5 wtAR was constructed as described previously (Yeh and Chang, Proc. Natl. Acad. Sci USA 93:5517-5521, 1996) . pGALO-AR (wild-type) was obtained from D. Chen (University of Massachusetts) . pGALO contains the GAL4 DNA binding domain (DBD) . For construction of pAS2-wtAR or -mAR, the C-terminal fragments (aa 595-918) from wtAR, mARt877s (Dr. S.P. Balk, Beth Israel Hospital, Boston, MA) , or mARe708k (H. Shim, Hyogo Medical College, Japan) were inserted in pAS2 yeast- expression vector (Clontech) . Another AR mutant (mARv888m) , derived from androgen insensitive syndrome patient, was constructed as previously described (Mowszowicz, et al . Endocrine 1:203-209, 1993). pGAL4-VP16 was used to construct a fusion of ARA70. pGAL4-VP16 contains the GAL4 DBD linked to the acidic activation domain of VP16. pCMX-Gal-N-RB and pCMX-VP16-AR were constructed by inserting fragments Rb (aa 370-928) and AR (aa 590-918) into pCMX-gal-N and pCMX-VP16, respectively. The sequence of construction junction was verified by sequencing. pYX-ARA24/Ran was constructed by placing the ARA24 gene under the control of the gal-1 promoter of yeast expression plasmid pYX243 (Ingenus) . A cDNA fragment encoding the AR poly-Q stretch and its flanking regions (AR a. a. 11-208) was ligated to a PAS2 yeast expression plasmid for use as bait in the two hybrid assay. AR cDNAs of different poly-Q lengths that span the same AR poly-Q region as our bait plasmid were constructed in pAS2 in the same way, for yeast two-hybrid liquid culture β-gal assay. These AR bait plasmids with poly-Q lengths of 1, 25, 49 were all transformed into yeast Y190 and found to not be autonomously active. pCMV-antisense ARA24/Ran (ARA24as) expression plasmid was constructed by inserting a 334 -bp Bgl II fragment of ARA24/Ran, which spans 5 ' -untranslated region and the translation start codon of ARA24/Ran (nucleotides 1-334 of SEQ ID NO:5), into pCMV vector in the antisense orientation. The MMTV-CAT and MMTV-Luc reporter genes were used for AR transactivation assay. pSG5-AR and pSV-βgal are under the regulation of SV40 promoter and β- globulin gene intron-1 enhancer. p6R-ARQl, p6R-ARQ25, p6R- ARQ49 were kindly provided by Dr. Roger L. Meisfield (Chamberlain, et al . Nucleic Acids Res. 22:3181-3186, 1994) pSG5-GAL4DBD-ARA24 was generated by inserting the coding sequence of Gal4DBD-ARA24 hybrid protein into pSG5 vector. pVP16-ARN-Ql, pVP16-ARN-Q25 , pVP16 -ARN-Q25 , pVP16- ARN-Q35, pVP16-ARN-Q49 were generated by inserting each poly-Q AR N-terminal domain (a. a. 34-555) into pVP16 vector (Clontech) to be expressed as a VP16AD hybrid protein. GALOAR plasmid, which contains GAL4DBD fused to E region of human AR, was a gift from Dr. D. Chen. The pSG5-CAT reporter plasmid (Clontech) contains five GAL4 binding sites upstream of the Elb TATA box, linked to the CAT gene. pSG5-AR and pSG5-ARA70 were constructed as previously described (Yeh and Chang, Proc. Natl. Acad. Sci

USA 93:5517-5521, 1996) . Two mutants of the AR gene (mAR877 derived from prostate cancer, codon 877 mutation Thr to Ala; and mAR708 derived from partial androgen insensitive syndrome (PIAS) , codon 708 mutation Glu to Lys) , were provided by S. Balk (Beth Israel Hospital, Boston) and H. Shi a (Hyogo Medical College, Japan) , respectively.

Clones used in the two-hybrid system to evaluate the role of Rb in AR transactivation were made by ligating an Rb fragment (aa 371-928) to the DBD of GAL4. Similarly, near full-length (aa 36-918) AR (nAR) and AR-LBD (aa 590- 918) fragments ligated to transcriptional activator VP16.

Screening of prostate cDNA library by a yeast two-hybrid system for ARAs associated with the ligand binding domain To identify ARA coactivators interact with the LBD, a pACT2 -prostate cDNA library was cotransformed into Y190 yeast cells with a plasmid of pAS2mAR (mART877S) which contains GAL4DBD(aa 1-147) fused with the C-terminal domain of this mAR. Transformants were selected for growth on SD plates with 3-aminotriazole (25mM) and DHT (lOOnM) lacking histidine, leucine and tryptophan (-3SD plates) . Colonies were also filter-assayed for β-galactosidase activity. Plasmid DNA from positive cDNA clones were found to interact with mtARt877s but not GAL4TR4 was isolated from yeast, amplified in E. coli , and the inserts confirmed by DNA sequencing. To identify clones that interact with the poly-Q region of the N-terminal domain, the AR poly-Q stretch (aa 11-208) was inserted into the pAS2 yeast expression plasmid and cotransformed into Y190 yeast cells with a human brain cDNA library fused to the Gal4 activation domain.

Transformants were selected for growth on SD plates lacking histidine, leucine and tryptophan and supplemented with 3- aminotriazole (40 mM) .

Amplification and characterization of ARA clones Full length DNA sequences comprising two coactivators, designated ARA54 (SEQ ID NO:l)and ARA55 (SEQ ID NO : 3 ) , that were found to interact with mARt877s were isolated by 5 'RACE PCR using Marathon cDNA Amplification Kit (Clontech) according to the manufacturer's protocol. The missing 5 ' coding region of the ARA54 gene was isolated from H1299 cells using the gene-specific antisense primer shown in SEQ ID NO: 9 and following PCR reaction conditions: 94°C for 1 min, 5 cycles of 94°C for 5 sec→72°C for 3 min, 5 cycles of 94°C for 5 sec-70°C for 3 min, then 25 cycles of 94°C for 5 sec-68°C for 3 min. The PCR product was subcloned into pT7-Blue vector (Novagen) and sequenced.

ARA55 was amplified by PCR from the HeLa cell line using an ARA55-specific antisense primer (SEQ ID NO: 10) and the PCR reaction conditions described for isolation of ARA54.

Using the 5'RACE-PCR method, we were able to isolate a 1721 bp DNA fragment (SEQ ID N0:1) from the H1299 cell line with an open reading frame that encodes a novel protein 474 amino acids in length (SEQ ID NO: 2) . The in- vi tro translation product is a polypeptide with an apparent molecular mass of 54±2 kDA, consistent with the calculated molecular weight (53.8 kDa) . The middle portion of ARA54 (a. a. 220-265 of SEQ ID NO:2) contains a cysteine-rich region that may form a zinc finger motif called the RING finger, defined as CX₂CX_9-27CXHX₂CX₂CX_6-17CX₂C (SEQ ID NO: 11), a domain conserved among several human transcriptional factor or proto-oncogeny proteins, including BRCA1 , RING1 , ^" PML and MEL-18 (Miki et al . , Science 266:66-71 (1994); Borden et al . , EMBO J. 14:1532-1541 (1995); Lovering et al., Proc . Na tl . Acad . Sci . USA 90:2112-2116 (1993); Blake et al . , Oncogene 6: 653-657 (1991); Ishida et al, Gene 129:249-255 (1993)). In addition, ARA54 also contains a second cysteine-rich motif which has a B box like structure located at 43 amino acids downstream from the RING finger motif. However, ARA54 differs from members of the RING finger-B-box family in that it lacks a predicted coiled- coil domain immediately C-terminal to the B box domain, which is highly conserved in the RING finger-B-box family. Therefore, ARA54 may represent a new subgroup of this family.

The full-length human ARA55 has an open reading frame that encodes a 444 aa polypeptide (SEQ ID NO: 4) with a predicted molecular weight of 55 kD that ARA55 shares 91% homology with mouse hic5. Human ARA55 has four LIM motifs in the C-terminal region. An LIM motif is a cysteine-rich zinc-binding motif with consensus sequence: CX₂CX₁₆_ ₂₃HX₂CX₂CX₂CX_16-21CX₂(C,H,D) (SEQ ID NO:12) (Sadler, et al . , J. Cell Biol . 119:1573-1587(1992)). Although the function of the LIM motif has not been fully defined, some data suggest that it may play a role in protein-protein interaction (Schmeichel & Beckerle, Cell 79:211-219, 1994). Among all identified SR associated proteins, only ARA55 and thyroid hormone interacting protein 6 (Trip 6) (Lee, et al . Mol . Endocrinol . 9:243-254 (1995)) have LIM motifs. A clone that showed strong interaction with the poly-Q bait was identified and subsequently subjected to sequence analysis. This clone contains 1566 bp insert (SEQ ID NO: 5) with an open reading frame encoding a 216 aa polypeptide (SEQ ID NO: 6) with a calculated molecular weight of 24 kDa . GenBank sequence comparison showed that this clone has the same open reading frame sequence as Ran/TC4 , an abundant ras-like small GTPase involved in nucleocytoplasmic transport that is found in a wide variety of cell types (Beddow et al . , Proc. Natl. Acad. Sci. U.S.A. 92:3328-333.2, 1995) . Accordingly, the factor was designated ARA24/Ran. The cDNA sequence of the ARA24 clone (SEQ ID NO: 5) (GenBank accession number AF052578) is longer than that of the published ORF for human Ran, in that it includes 24 and 891 bp of 5'- and 3 ' -untranslated regions, respectively.

Northern Blotting

The total RNA (25μg) was fractionated on a 1% formaldehyde-MOPS agarose gel, transferred onto a Hybond-N nylon membrane (Amersham) and prehybridized. A probe corresponding to the 900 bp C-terminus of ARA55 or an ARA54 -specific sequence was ³²P-labeled in vi tro using Random Primed DNA Labeling Kit (Boehringer-Mannheim) according to the manufacture's protocol and hybridized overnight. After washing, the blot was exposed and quantified by Molecular Dynamics Phosphorlmager . β-actin was used to monitor the amount of total RNA in each lane . Northern blot analysis indicated the presence of a 2 kb ARA55 transcript in Hela and prostate PC3 cells. The transcript was not detected in other tested cell lines, including HepG2 , H1299, MCF7, CHO, PC12, P19, and DU145 cells. The ARA54 transcript was found in H1299 cells, as well as in prostate cancer cell lines PC3 and LNCaP.

Co-immunoprecipitation of AR and ARAs

Lysates from in-vi tro translated full-length of AR and ARA54 were incubated with or without 10^"8 M DHT in the modified RIPA buffer (50mM Tris-HCL pH 7.4, 150mM NaCl, 5mM EDTA, 0.1% NP40, ImM PMSF, aprotinin, leupeptin, pepstatin, 0.25% Na-deoxycholate, 0.25% gelatin) and rocked at 4°C for

2 hr. The mixture was incubated with rabbit anti-His»tag polyclonal antibodies for another 2 hr and protein A/G PLUS -Agarose (Santa Cruz) were added and incubated at 4°C for additional 2 hr. The conjugated beads were washed 4 times with RIPA buffer, boiled in SDS sample buffer and analyzed by 8% SDS/PAGE and visualized by STORM 840 (Molecular Dynamics) .

ARA54 and AR were found in a complex when immunoprecipitated in the presence of 10^~8 M DHT, but not in the absence of DHT. This result suggests that ARA54 interacts with AR in an androgen-dependent manner.

Interaction between recombinant full length human AR and ARA24/Ran proteins further examined by co- immunoprecipitation, followed by SDS-PAGE and western blotting. Results of the co-immunoprecipitation assay indicate that ARA24/Ran interacts directly with AR. The phosphorylation state of bound guanine nucleotide to the small GTPases does not affect this interaction.

AR pull-down assay using GST-Rb Full-length Rb fused to glutathione-S-transferase

(ST-Rb_1-928) was expressed and purified from E. coli. strain Bl21pLys as described recently (Zarkowska & Mittnacht, J. Biol. Chem. 272:12738-12746, 1997). Approximately 2 μg of His-tag column purified baculovirus AR was mixed with GST- loaded glutathione-Sepharose beads in 1 ml of NET-N (20 mM Tris-HCL(pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% (v/v) Noniodet P-40) and incubated with gentle rocking for 3 hr at 4°C.

Following low-speed centrifugation to pellet the beads, the clarified supernatant was mixed with GST-Rb- loaded glutathione-Sepharose beads in the presence or absence of 10 mM DHT and incubated for an additional 3 hr with gentle rocking at 4°C. The pelleted beads were washed 5 times with NET-N, mixed with SDS-sample buffer, boiled, and the proteins separated by electrophoresis on a 7.5% polyacrylamide gel. A Western blot of the gel was incubated with anti-AR polyclonal antibody NH27 and developed with alkaline phosphatase-conjugated secondary antibodies .

AR was coprecipitated with GST-Rb, but not GST alone, indicating that AR and Rb are associated in a complex together. Transfection Studies

Human prostate cancer DU145 or PC3 cells, or human - lung carcinoma cells NCI H1299 were grown in Dulbecco's minimal essential medium (DMEM) containing penicillin (25U/ml) , streptomycin (25μg/ml) , and 5% fetal calf serum (FCS) . One hour before transfection, the medium was changed to DMEM with 5% charcoal-stripped FCS. Phenol red- free and serum- free media were used on the experiments employing E2 or TGFβ, respectively. A β-galactosidase expression plasmid, pCMV-β-gal, was used as an internal control for transfection efficiency.

Cells were transfected using the calcium phosphate technique (Yeh, et al . Molec . Endocrinol . 8:77-88, 1994). The medium was changed 24 hr posttransfection and the cells treated with either steroid hormones or hydroxyflutamide , and cultured for an additional 24 hr. Cells were harvested and assayed for CAT activity after the cell lysates were normalized by using β-galactosidase as an internal control. Chloramphenicol acetyltransferase (CAT) activity was visualized by Phosphorlmager (Molecular Dynamics) and quantitated by ImageQuant software (Molecular Dynamics) .

Mammalian Two-Hybrid Assay

The mammalian two-hybrid system employed was essentially the protocol of Clontech (California) , with the following modifications. In order to obtain better expression, the GAL4DBD (a. a. 1-147) was fused to pSG5 under the control of an SV40 promoter, and named pGALO . The hinge and LBD of wtAR were then inserted into pGALO . Similarly, the VP16 activation domain was fused to pCMX under the control of a CMV promoter, and designated pCMX-

VP16 (provided by Dr. R.M. Evan) .

The DHT-dependent interaction between AR and ARA54 was confirmed in prostate DU145 cells using two-hybrid system with CAT reporter gene assay. Transient transfection of either ARA54 or wtAR alone showed negligible transcriptional activity. However, coexpression of AR with ARA54 in the presence of 10^'8 M DHT significantly induced CAT activity.

ARA54 functions as a coactivator relatively specific for AR-mediated transcription. ARA54 induces the transcriptional activity of AR and PR by up to 6 fold and 3-5 fold, respectively. In contrast, ARA54 showed only marginal effects (less than 2 fold) on GR and ER in DU145 cells. These data suggest that ARA54 is less specific to AR as relative to ARA70, which shows higher specificity to AR. However, we can not rule out the possibility that

ARA54 might be more general to other steroid receptors in other cell types under different conditions.

Coexpression of ARA54 with SRC-1 or ARA70 was found to enhance AR transcriptional activity additively rather than synergistically. These results indicate that these cofactors may contribute individually to the proper or maximal AR-mediated transcriptional activity.

Since the C-terminal region of ARA54(a.a. 361-471 of SEQ ID NO: 2) isolated from prostate cDNA library has shown to be sufficient to interact with AR in yeast two-hybrid assays, we further investigated whether it could squelch the effect of ARA54 on AR-activated transcription in H1299 cells, which contain endogenous ARA54. The C-terminal region of ARA54 inhibits AR-mediated transcription by up to 70%; coexpression of exogenous full-length ARA54 reverses this squelching effect in a dose-dependent manner. These results demonstrate that the C-terminal domain of ARA54 can serve as a dominant negative inhibitor, and that ARA54 is required for the proper or maximal AR transactivation in human H1299 cells.

Examination of the effect of ARA54 on the transcriptional activities of wtAR and mtARs in the presence of DHT, E2 and HF revealed differential ligand specificity. Translational activation of wtAR occurred in the presence of DHT (lO^-10 to 10^"8 M) ; coexpression of ARA54 enhanced transactivation by another 3-5 fold. However, wtAR responded only marginally to E2 (10^_9-10-⁷ M) or HF (10^"7-10^"5 M) in the presence or absence of ARA54. As expected, the positive control, ARA70, is able to enhance^' the AR transcriptional activity in the presence of 10^~9 - 10^"7 M E2 and 10^~7 - 10^"5 HF, that matches well with previous reports (Yeh, PNAS, Miyamoto, PNAS) .

The AR mutants Art877a, which is found in many prostate tumors (23) , and Are708k, found in a yeast genetic screening (24) and a patient with partial androgen insensitivity, exhibited differential specificity for lignands. In the absence of ARA54 , Art877a responded to E2 (10-⁹-10^"7 M) and HF (10^"7-10-⁵ M) , and ARA54 could further enhance E2- or HF-mediated AR transactivation. These results suggested that mtARs might also require cofactors for the proper or maximal DHT- , E2-, or HF-mediated AR transcriptional activity. The DHT response of mARe708k was only a slightly less sensitive than that of wtAR or mARt877s, whereas E2 and HF exhibited no agonistic activity toward ARe708k. Together, these results imply that the change of residue 708 on AR might be critical for the interaction of the antiandrogen-ARe708k-ARA54 complex, and that both AR structure and coactivators may play a role in determining ligand specificity.

CAT activity in DU145 cells cotransfected with a plasmid encoding the hormone binding domain of wtAR fused to the GAL4 DBD (GALOAR) and a plasmid encoding full-length ARA55 fused to the activation domain of VP16 (VP16-ARA55) was significantly induced by the cotransfection of VP16- ARA55 and GALOAR in the presence of 10 nM DHT, but not induced by E2 or HF . Combination of GAL0 empty vector and VP16-ARA55 did not show any CAT activity. Combination of GALOAR and VP16 vector showed negligible CAT activity. These results indicate that ARA55 interacts with AR in an androgen-dependent manner .

Transient transfection assays were conducted to investigate the role of ARA55 in the transactivation activity of AR. DU145 cells were cotransfected with MMTV- CAT reporter, increasing amounts of ARA55 and wtAR under eukaryotic promoter control. Ligand- free AR has minimal MMTV-CAT reporter activity in the presence or absence of ARA55. ARA55 alone also has only minimal reporter activity Addition of 10 nM DHT resulted in 4.3 fold increase of AR transcriptional activity and ARA55 further increased this induction by 5.3 fold (from 4.3 fold to 22.8 fold) in a dose-dependent manner. The induced activity reached a plateau at the ratio of AR:ARA55 to 1:4.5. Similar results were obtained using PC3 cells with DU145 cells, or using a CAT reporter gene under the control of a 2.8 kb promoter region of a PSA gene. The C-terminus of ARA55 (ARA55_251.444) (a. a. 251-444 of SEQ ID NO : 4 ) did not enhance CAT activity. Cotransfection of PC3 cells, which contain endogenous ARA55, with ARA55_251-444, AR and MMTV-CAT reporter in the presence of 10 nM DHT demonstrated dramatically reduced AR transcriptional activity relative to cells transfected with AR and MMTV-CAT alone. These results demonstrate that ARA55 is required for the proper or maximal AR transcriptional activity in PC3 cells., and that the C- terminus of ARA55 can serve as a dominant negative inhibitor.

The effect of ARA55 on mARt877s and mARe708k in the presence of DHT and its antagonists, E2 , and HF . The mARt877s receptor is found in LNCaP cells and/or advanced prostate cancers and has a point mutation at codon 877 (Thr to Ser) (Gaddipati et al . , Cancer Res . 54:2861-2864 (1994); Veldscholte et al . , Biochem . Biophys . Commun . 173:534-540 (1990) ) . The mARe708k receptor, has a point mutation at codon 708 (Glu to Lys) , was isolated by a yeast genetic screening and exhibits reduced sensitivity to HF and E2 relative to wtAR(Wang, C. , PhD thesis of Universi ty of Wisconsin -Madison (1997) ) . The transcriptional activities of wtAR, mARt877s, and mARe708k are induced by DHT (10^"11 to 10"⁸ M) . ARA55 enhanced the transactivation of all three receptors by 4-8 fold. In the presence of E2 or HF, wtAR responded marginally only at higher concentrations (10^~7 M for E2 and 10^~5 M for HF) . Cotransfection of wtAR with ARA55 at a 1:4.5 ratio, however, increases AR transcriptional activity at 10^~8-10^~7 M for E2 or 10^~6 to 10^~~5 M for HF. Compared to wtAR, the LNCaP mAR responded much better to E2 and HF and ARA55 significantly enhanced its transcriptional activity. ARA55 may be needed for the proper or maximal DHT- , E2-, or HF-mediated AR transcriptional activity.

The effect of ARA55 on transcriptional activation by GR, PR, and ER was tested in DU145 cells. ARA55 is relatively specific to AR, although it may also enhance GR and PR to a lesser degree, and has only a marginal effect on ER. ARA70 shows much higher specificity to AR than ARA55, relative to the other tested steroid receptors. Although ARA55 enhances AR-mediated transcription to a greater degree than GR- , PR-, or ER-mediated transcription, it appears to be less specific than ARA70.

Because the amino acid sequence of ARA55 has very high homology to mouse hic5, and early studies hic5 suggested this mouse gene expression can be induced by the negative TGFβ(Shibanuma et al . , J^". Biol . Chem . 269:26767-26774 (1994)), we were interested to see whether ARA55 could serve as a bridge between TGFβ and AR steroid hormone system. Northern blot analysis indicated that TGFβ treatment (5 ng/ml) could induce ARA55 mRNA by 2-fold in PC3 cells. In the same cells, TGFβ treatment increased AR transcriptional activity by 70%. This induction is weak relative to the affect achieved upon transfection of PC3 cells with exogenous ARA55 (70% vs. 4 fold) . This may be related to the differences in the ratios of AR and ARA55. The best ratio of AR:ARA55 for maximal AR transcriptional activity is 1:4.5. Whether other mechanisms may also be involve in this TGFβ-induced AR transcriptional activity will be an interesting question to investigate. The unexpected discovery that TGFβ may increase AR transcriptional activity via induction of ARA55 in prostate may represent the first evidence to link a negative regulatory protein function in a positive manner, by inducing the transcriptional activity of AR, the major promoter for the prostate tumor growth.

The ability of ARA55 to induce transcriptional activity of both wtAR and mARt877s in the presence of DHT, E2 , and HF suggests an important role for ARA55 in the progression of prostate cancer and the development of resistance to hormonal therapy. Evaluation of molecules that interfere with the function of ARA55 may aid in the identification of potential chemotherapeutic pharmaceuticals.

Human small lung carcinoma H1299 cell line, which has no endogenous AR protein, were transfected with AR and ARA24/Ran. Because ARA24/Ran is one of the most abundant and ubiquitously expressed proteins in various cells, both sense and antisense ARA24/Ran mammalian expression plasmids were tested. Overexpression of sense ARA24/Ran did not significantly enhance the AR transactivation, a result that is not surprising, in view of the abundance of endogenous ARA24/RAN. However, expression of antisense ARA24/Ran (ARA24as) markedly decreased DHT- induced CAT activity in a dose dependent manner. Furthermore, increasing the DHT concentration from 0.1 nM to 10 nM DHT resulted in strong induction of AR transactivation and decreased the inhibitory effect of ARA24as effect, indicating that increased DHT concentration can antagonize the negative effect of ARA24as.

The affinity between ARA24/Ran and AR is inversely related to the length of AR poly-Q stretch. AR transactivation decreases with increasing AR poly-Q length. Reciprocal two-hybrid assays with exchanged fusion partners, Gal4DBD-ARA24/Ran and VP16AD-ARNS (a. a. 34-555 with poly-Q lengths of 1, 25, 35, 49 residues) were conducted using mammalian CHO cells. These results consistently show that the affinity between ARA24/Ran and AR poly-Q region is inversely correlated with AR poly-Q length in both yeast and mammalian CHO cells.

The regulation of AR transactivation by ARA24/Ran correlates with their affinity. These results suggest that ARA24/Ran could achieve differential transactivation of AR, with ARs having different poly-Q length could existing in a single cell or cell system. ARA24as was again used in the ARE-Luc transfection assays to address the role of AR poly- Q length in the regulation of AR by ARA24/Ran. ARs of poly-Q lengths 1, 25, and 49 residues, and increasing amounts (1, 2, and 4 μg) of ARA24as expression vectors were co-transfected with equal amounts of reporter plasmid (pMMTV-Luc) in CHO cells. Although the basal reporter activity is slightly affected by increasing amounts of antisense ARA24/Ran, ARA24as showed a more significant decrease of AR transactivation. As AR poly-Q length increased, the ARA24as effect on AR transactivation decreased. These results suggest that the affinity of

ARA24/Ran for AR and the effect of decreasing ARA24/Ran on AR transactivation faded over the expansion of AR poly-Q length.

Coexpression of Rb and AR expression plasmids in DU145 cells using the mammalian two-hybrid system resulted in a 3 fold increase in CAT activity by cotransfection of near full length AR (nAR, amino acids 36-918) and Rb. Cells cotransfected with nAR and PR-LBD or Rb and ARA70 did not show increased CAT activity. Surprisingly, addition of 10 nM DHT made very little difference in the interaction between Rb and nAR. The inability of Rb to interact with AR-LBD suggest that interaction site of AR is located in N- terminal domain (aa 36 to 590) . Together, our data suggest the interaction between Rb and AR is unique in the following ways: first, the interaction is androgen- independent and binding is specific but relatively weak as compared to other AR associated protein, such as ARA70 (3 fold vs. 12 fold induced CAT activity in mammalian two- hybrid assay, data not shown) . Second, unlike most identified steroid receptor associated proteins that bind to C-terminal domain of steroid receptor, Rb binds to N- terminal domain of AR. Third, no interaction occurred between Rb and ARA70, two AR associated proteins in DU145 cells.

DU145 cells containing mutated Rb (Singh et al . , Na ture 374: 562-565 (1995)) were cultured with charcoal- stripped FCS in the presence or absence of 1 nM DHT. No AR transcriptional activity was observed in DU145 cells transiently transfected with wild type AR and Rb at the ratio of 1:3 in the absence of DHT. When However, AR transcriptional activity could be induced 5-fold when wild type AR was expressed in the presence of 1 nM DHT.

Cotransfection of Rb with AR can further enhance the AR transcriptional activity from 5-fold to 21-fold in the presence of 1 nM DHT. As a control, cotransfection of ARA70, the first identified AR coactivator, can further enhance in DU145 cells transcriptional activity from 5-fold to 36-fold. In DU145 cells transfected with Rb, ARA70, and AR, the induction of AR transcriptional activity was synergistically increased from 5-fold to 64-fold. Upon transfection of wild type AR without Rb or ARA70, only marginal induction (less than 2-fold) was detected in the presence of 10 nM E2 or 1 μM HF. In contrast, cotransfection of the wild type AR with Rb or ARA70 can enhance the AR transcriptional activity to 12 -fold (E2) or 3-4 fold (HF) , and cotransfection of Rb and ARA₇₀ with AR can further enhance the AR transcriptional activity to 36- fold (E2 or 12-fold (HF) . We then extended these findings to two different AR mutants: mARt877s from a prostate cancer patient and mARe708k from a partial-androgen- insensitive patient. Similar inductions were obtained when wild type AR was replaced by mARt877s. In contrast, while similar induction was also detected in the presence of 1 nM DHT when we replace wild type AR with mARe708k, there was almost no induction by cotransfection of meAR708k with Rb and/or ARA70 in the presence of 10 nM E2 or 1 μM HF. These results indicated that Rb and ARA70 can synergistically induce the transcriptional activity of wild type AR and mAR877 in the presence of 1 nM DHT, 10 nM E2 or 1 μM HF . However, Rb and ARA70 synergistically induce the transcriptional activity of mAR708 only in the presence of 1 nM DHT, but not 10 nM E2 or 1 μM HF . The fact that Rb and ARA70 can induce transcriptional activity of both wild type AR and mutated AR that occur in many prostate tumors may also argue strongly the importance of Rb and ARA70 in normal prostate as well as prostate tumor. Also, the differential induction of DHT vs. E2/HF may suggest the position of 708 in AR may play vital role for the recognition of androgen vs anti- androgens to AR.

We also examined the effect of Rb and ARA70 on the transcriptional activity of other steroid receptors through their cognate DNA response elements [MMTV-CAT for AR, glucocorticoid receptor (GR) , and progesterone receptor (PR); ERE-CAT for estrogen receptor (ER) ] . Although Rb and ARA70 can synergistically induce AR transcriptional activity up to 64 -fold, Rb and ARA70 can only have marginal induction on the transcriptional activity of GR, PR, and ER in DU145 cells. These results suggest that Rb and ARA70 are more specific coactivators for AR in prostate DU145 cells. However, it cannot be ruled out that possibly the assay conditions in prostate DU145 cells are particularly favorable for Rb and ARA70 to function as coactivators for AR only, and Rb and ARA70 may function as stronger coactivators for ER, PR, and GR in other cells or conditions. Failure of Rb to induce transactivation by mutant AR888, which is unable to bind androgen, suggests that while interaction between Rb and AR is androgen- independent, the AR-Rb (and AR-ARA70) complexes require a ligand for the transactivation activity.

The activity of Rb in cell cycle control is related essentially to its ability to bind to several proteins, thus modulating their activity. To date, many cellular proteins have been reported which bind to Rb (Weinberg, R.A., Cell 81:323-330 (1995)) . These include a number of transcription factors, a putative regulator of ras, a nuclear structural protein, a protein phosphatase, and several protein kinases. Whether all of these proteins actually complex, and are regulated by Rb, in cells remains to be seen.

Much attention has been given to the functional interaction between Rb and transcription factors. To date, several of these factors have been shown to form complexes with Rb in cells. Such complex formation and subsequent function studies have revealed that the modulating activity of Rb can take the form of repression of transcription as with E2F (Weintraub et al . , Na ture 375:812-815 (1995)), or activation as with NF-IL6 (Chen et al . , Proc . Natl . Acad . Sci . USA 93:465-469 (1996)) and the hBrm/BRGl complex (Singh et al . , (1995)) . Here, we show that Rb can bind to AR and induce the AR transcriptional activity. To our knowledge, this is the first demonstration of a negative growth regulatory protein functioning in a positive manner, by initiating transcription via a signal transduction mechanism involving binding to a nuclear receptor. When place in the context of regulating the cell cycle and differentiation, these data suggest a previously undescribed function for Rb which underscores the importance of this protein in regulating transcription by direct binding to transcription factor, but this protein can also regulate transcription by stimulating at least one type of signal transduction mechanism.

A relationship between Rb expression and response to endocrine therapy of human breast tumor has been suggested (Anderson et al . , J^". Pathology 180:65-70 (1996)). Other studies indicate that Rb gene alterations can occur in all grades and stages of prostate cancer, in localized as well as metastatic disease (Brooks et al . , Prostate 26:35-39 (1995) ) . How Rb function may be linked to androgen- dependent status in prostate tumor progression remains unclear. One possible explanation is that Rb alteration may be a necessary event in prostate carcinogenesis for a subset of prostatic neoplasms, which may be also true for the AR expression in prostate tumors . All publications cited in this application are incorporated by reference.

The present invention is not limited to the exemplified embodiment, but is intended to encompass all such modifications and variations as come within the scope of the following claims .

Claims

CLAIMS WE CLAIM:

1. An isolated polynucleotide comprising a sequence selected from the group consisting of SEQ ID NO : 1 and SEQ ID NO: 3.

2. A genetic construct comprising a promoter capable of causing expression of a protein coding region in a cell, the promoter operably connected to a protein coding region encoding the expression of a polypeptide from coding regions of ARA54 or ARA55.

3. The genetic construct of claim 2 wherein the polypeptide encoded by the protein coding sequence comprises a sequence selected from the group consisting of SEQ ID NO : 2 and SEQ ID NO : 4.

4. A eukaryotic host cell comprising the genetic construct of claim 2.

5. A method for testing the androgenic or antiandrogenic effect of a chemical compound comprising the steps of : (a) transfecting a host cell with at least one genetic construct capable of producing in the host cell a polypeptide selected from the group consisting of ARA54, ARA55, ARA24, and Rb, the host cell also producing human androgen receptor protein; (b) exposing the cell to the chemical compound; and (c) measuring the level of transcriptional activity caused by the androgen receptor.

6. The method of claim 5 wherein the host cell is a prostate cell.

7. The method of claim 5, wherein the cell is a eukaryotic cell that lacks native endogenous androgen receptor, the cell having also an introduced genetic construct producing androgen receptor protein.

8. The method of claim 5, wherein the genetic construct comprises a DNA sequence selected from the group consisting of SEQ ID NO:l, SEQ ID NO : 3 , SEQ ID NO : 5 , and SEQ ID NO: 7.

9. The method of claim 5, wherein the cell is transfected with a genetic construct comprising a reporter gene expressible in the cell, the expression of said reporter gene being susceptible to detection and quantitation.

10. The method of claim 9, wherein the reporter gene is selected from the group consisting of a chloramphenicol acetyltransferase gene and a luciferase gene.

11. A method for testing the androgenic or antiandrogenic effect of a chemical compound comprising the steps of :

(a) transfecting a host cell with at least one genetic construct capable of producing in the host cell human androgen receptor protein and a polypeptide selected from the group consisting of ARA54 , ARA55, ARA24, and Rb;

(b) exposing the cell to the chemical compound; and

(c) measuring the interaction between AR and an AR co-activator.

12. A method as claimed in claim 11 wherein the coactivator is selected from the group consisting of ARA5 , ARA55, ARA24 and Rb .