Recruitment of β-Catenin by Wild-Type or Mutant Androgen Receptors Correlates with Ligand-Stimulated Growth of Prostate Cancer Cells

0888-8809/04/$15.00/0 Printed in U.S.A. Molecular Endocrinology 18(10):2388–2401 Copyright © 2004 by The Endocrine Society doi: 10.1210/me.2003-0436 Recruitment of ␤-Catenin by Wild-Type or Mutant Androgen Receptors Correlates with LigandStimulated Growth of Prostate Cancer Cells DAVID MASIELLO, SHAO-YONG CHEN, YOUYUAN XU, MANON C. VERHOEVEN, EUNIS CHOI, ANTHONY N. HOLLENBERG, AND STEVEN P. BALK Cancer Biology Program/Hematology-Oncology Division (D.M., S.-Y.C., Y.X., M.C.V., E.C., S.P.B.) and the Thyroid Unit/Endocrinology Division (A.N.H.), Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215; and University of Utrecht (M.C.V.), 3508 GA, Utrecht, The Netherlands Prostate cancers respond to treatments that suppress androgen receptor (AR) function, with bicalutamide, flutamide, and cyproterone acetate (CPA) being AR antagonists in clinical use. As CPA has substantial agonist activity, it was examined to identify AR coactivator/corepressor interactions that may mediate androgen-stimulated prostate cancer growth. The CPA-liganded AR was coactivated by steroid receptor coactivator-1 (SRC-1) but did not mediate N-C terminal interactions or recruit ␤-catenin, indicating a nonagonist conformation. Nonetheless, CPA did not enhance AR interaction with nuclear receptor corepressor, whereas the AR antagonist RU486 (mifepristone) strongly stimulated AR-nuclear receptor corepressor binding. The role of coactivators was further assessed with a T877A AR mutation, found in LNCaP prostate cancer cells, which converts hydroxyflutamide (HF, the active flutamide metabolite) into an agonist that stimulates LNCaP cell growth. The HF and CPA-liganded T877A ARs were coactivated by SRC-1, but only the HF-liganded T877A AR was coactivated by ␤-catenin. L-39, a novel AR antagonist that transcriptionally activates the T877A AR, but still inhibits LNCaP growth, similarly mediated recruitment of SRC-1 and not ␤-catenin. In contrast, ␤-catenin coactivated a bicalutamideresponsive mutant AR (W741C) isolated from a bicalutamide-stimulated LNCaP subline, further implicating ␤-catenin recruitment in AR-stimulated growth. Androgen-stimulated prostate-specific antigen gene expression in LNCaP cells could be modulated by ␤-catenin, and endogenous c-myc expression was repressed by dihydrotestosterone, but not CPA. These results indicate that interactions between AR and ␤-catenin contribute to prostate cell growth in vivo, although specific growth promoting genes positively regulated by AR recruitment of ␤-catenin remain to be identified. (Molecular Endocrinology 18: 2388–2401, 2004) T C-terminal ligand binding domain (LBD) that also has transactivation activity. The transcriptional activity of the LBD, activation function-2, is largely due to a ligand induced shift in the position of helix 12 that generates a binding site for a short hydrophobic helical motif (leucine-X-X-leucine-leucine, LXXLL, where X can be any amino acid) (4–7). The LXXLL motif forms the core of the nuclear receptor box (NR box), which is found in single or multiple copies in many transcriptional coactivator proteins that associate with agonistliganded nuclear receptors (4, 8). Steroid receptor coactivator-1 (SRC-1) and SRC-2 [transcriptional intermediary factor 2 and glucocorticoid receptor (GR)interacting protein 1] are two such well-characterized NR box-containing coactivator proteins that clearly contribute to nuclear receptor functions (9–12). In contrast to other steroid hormone receptors, the AR LBD binds very weakly to the SRC-1 or -2 NR boxes and has minimal transactivation activity when it is expressed independently of the N terminus, whereas the AR N terminus has a very strong autonomous transactivation function (activation function 1) (13, 14). AR binding to SRC-1 and -2 is mediated HE ANDROGEN RECEPTOR (AR) is a steroid hormone receptor member of the larger nuclear receptor family of ligand-activated transcription factors (1, 2). Similarly to other steroid hormone and nuclear receptors, the AR recruits to specific genes a number of proteins and protein complexes that serve to remodel chromatin and stimulate transcription (3). The AR is composed of an N-terminal transactivation domain, a central DNA binding domain (DBD), and a Abbreviations: AR, Androgen receptor; ARE, androgenresponsive element; CDS, charcoal dextran-stripped; CMV, cytomegalovirus; CPA, cyproterone acetate; DHT, dihydrotestosterone; DBD, DNA binding domain; FBS, fetal bovine serum; GR, glucocorticoid receptor; HF, hydroxyflutamide; LBD, ligand binding domain; MMTV, mouse mammary tumor virus; NCoR, nuclear receptor corepressor; NR, nuclear receptor; PI, propidium iodide; PR, progesterone receptor; RLU, relative light unit; siRNA, short interfering RNA; SRC-1, steroid receptor coactivator-1; Tcf, T-cell factor. Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community. 2388 Masiello et al. • AR Recruitment of ␤-Catenin primarily by the AR N terminus and a glutamine-rich domain in the SRC proteins (15–17). Although the AR LBD interacts very weakly with the NR boxes in SRC proteins, it binds strongly to an LXXLL-related motif (FXXLF) found in the AR N terminus (18). This binding makes a major contribution to AR transcriptional activity and mediates what appears to be an intermolecular interaction between the AR N and C termini in the AR homodimer (15, 19–22). Nonetheless, the AR N-C interaction is not absolutely required for transcriptional activity as ligands that do not support the interaction can still stimulate AR when used at relatively high concentrations, and peptides that block the N-C interaction do not necessarily inhibit AR transcriptional activity (23, 24). These results have suggested that the AR N-C interaction, in conjunction with helix 12, may serve to stabilize agonist ligand binding and receptor conformation at physiological agonist concentrations. The AR plays a central role in normal prostate development and in the development and progression of prostate cancer, with the majority of prostate cancer patients responding to therapies that decrease androgen levels (medical or surgical castration) or directly block AR functions (AR antagonists) (25). However, the molecular mechanisms and transcriptional targets mediating AR effects on normal vs. neoplastic prostate growth remain unclear. Bicalutamide, flutamide, and cyproterone acetate (CPA) are the AR antagonists that have been used most extensively for prostate cancer treatment (26). Bicalutamide and hydroxyflutamide (HF, the active metabolite of flutamide) are relatively pure AR antagonists in vivo, although HF has weak agonist activity at high concentrations in transient transfection assays (27–29). Although HF is an antagonist of the wild-type AR, it is an agonist for certain mutant ARs identified in prostat (...truncated)