Identification of Domains of C-Jun Mediating Androgen Receptor
Total Page:16
File Type:pdf, Size:1020Kb
Oncogene (1998) 16, 2001 ± 2009 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 http://www.stockton-press.co.uk/onc Identi®cation of domains of c-Jun mediating androgen receptor transactivation Scott C Wise1,3,4, Lori A Burmeister1,3, Xiao-feng Zhou1, Athanasios Bubulya1, Jennifer L Ober®eld1, Michael J Birrer2 and Lirim Shemshedini1 1University of Toledo, Department of Biology, Toledo, Ohio 43606 and 2Cell and Cancer Biology Department, Division of Clinical Sciences, National Cancer Institute, 9610 Medical Center Drive, Suite 300 Rockville, Maryland 20850, USA The proto-oncoprotein c-Jun, when complexed with c- conserved DNA-binding domain, required for se- Fos, forms the climeric complex identi®ed as AP-1 which quence-speci®c DNA recognition, and a more varied regulates transcription directly by binding to AP-1- ligand-binding domain giving nuclear receptors ligand- responsive genes. We have previously reported an speci®c and ligand-dependent transcriptional activities indirect mechanism by which c-Jun is able to regulate (reviewed in Mangelsdorf et al., 1995). Recent studies transcription by stimulating androgen receptor transacti- indicate that the transcriptional activity of nuclear vation in the absence of c-Fos or any apparent DNA receptors is dependent on a class of proteins binding. A series of c-Jun mutants were tested in order collectively referred to as coactivators, proteins to characterize the domains of c-Jun responsible for this believed to bridge the nuclear receptors to the RNA eect. The studies reported here indicate that a polymerase II transcription machinery (reviewed in Gill functional bZIP region and a portion of the N-terminal and Tjian, 1992). Biochemical assays and yeast two- activation functions is necessary for c-Jun stimulation of hybrid screens have identi®ed a number of receptor androgen receptor transactivation. Testing c-Jun/v-Jun coactivators, including p140 and p160 (Kurokawa et chimeras, we show that v-Jun is unable to stimulate al., 1995), SRC-1 (Onate et al., 1995), TIF-1 (Le androgen receptor transactivation and the eect is Dourin et al., 1995), TIF-2 (Voegel et al., 1996), RIP- dependent on the c-Jun activation functions. c-Jun 140 (Cavilles et al., 1995), GRIP (Hong et al., 1996), exhibits a bell-shaped activity on androgen receptor- and CBP/p300 (Kamei et al., 1996; Chakravarti et al., mediated transactivation which appears to be distinct 1996). SRC-1, p160, TIF-2, and GRIP-1 are structu- from c-Jun's transactivation ability. A c-Jun mutant rally related proteins which mediate the activity of the de®cient in transactivation is able to stimulate androgen ligand-dependent AF-2 (activation function-2), a receptor activity. These results indicate that c-Jun's conserved motif within the distal C-terminus of many transactivation ability can be separated from c-Jun's nuclear receptors (Durand et al., 1994). ability to stimulate the androgen receptor transactiva- Based on homology within the DNA-binding tion. domains, the AR is most closely related to the receptors for glucocorticoids (GR), progesterone (PR), Keywords: c-Jun; androgen receptor; transactivation and mineralocorticoids (MR). These receptors form a subfamily within the nuclear receptor superfamily that is capable of recognizing a common DNA sequence (Ham et al., 1988; Arriza et al., 1987). While a number Introduction of genes have been shown to be activated by more than one receptor within this family, recent studies have Androgens are steroid hormones essential for proper identi®ed several genes that are activated speci®cally by male development, including the development of male AR (Claessens et al., 1996; Alder et al., 1992) or GR external genitalia and virilization of the pubertal male. (Archer et al., 1994). How gene-speci®c activation by These androgen eects are mediated through the AR or GR occurs is not presently understood, but androgen receptor (AR) (Chang et al., 1988), a several possible mechanisms have been proposed. These member of the nuclear receptor superfamily. Nuclear include promoter-speci®c elements ¯anking and inter- receptors form a class of DNA-binding proteins which acting with the hormone-response elements (HREs) that modulate transcription in a ligand-dependent manner bind nuclear receptors (Scarlett and Robins 1995), (reviewed in Mangelsdorf et al., 1995; Mangelsdorf and dierential chromatin eects on receptor activity Evans, 1995; Beato et al., 1995; Kastner et al., 1995; (Archer et al., 1994), or receptor interaction with Thummel, 1995). While the ligands for nuclear distinct accessory factors (reviewed in Beato et al., receptors are diverse, the receptors themselves are 1995). Nuclear receptors are known to interact with a highly conserved in both structure and function. Most variety of accessory factors. Interestingly, the interac- of the known nuclear receptors have an evolutionary tion with the related coactivators SRC-1 and TIF-2 appears to be receptor-speci®c. While SRC-1 stimulates the transcriptional activity of GR and PR (Onate et al., Correspondence: L Shemshedini 1995), TIF-2 enhances transactivation by AR and PR, 3These authors contributed equally to this work but not GR (Voegel et al., 1996). An unrelated putative 4Present address: Cell and Cancer Biology Department, Division of receptor coactivator, ARA70, was recently identi®ed Clinical Sciences, National Cancer Institute, 9610 Medical Center Drive, Suite 300 Rockville, Maryland 20850 that mediates only AR-induced transcription (Yeh and Received 4 August 1997; revised 3 November 1997; accepted 4 Chang, 1996) and thus may be partially responsible for November 1997 androgen-speci®c responses. c-Jun domains mediating androgen receptor activity SC Wise et al 2002 We have previously published that another mediator of AR activity is the proto-oncoprotein c-Jun, which acts on AR without any apparent DNA binding (Bubulya et al., 1996; Shemshedini et al., 1991). However, unlike SRC-1, TIF-2, or ARA70, all of which have only positive eects on nuclear receptors, c- Jun has been shown in several studies to negatively interact with many nuclear receptors, except AR (reviewed in Pfahl, 1993). Our results indicate that the c-Jun positive eect on AR is independent of cell- speci®c and/or promoter-speci®c factors (Shemshedini et al., 1991; J Ober®eld and L Shemshedini, unpub- lished results), making it distinct from the positive interaction between c-Jun and GR that has been seen only in several T cell lines (Maroder et al., 1993). Furthermore, we have recently demonstrated that AR is stimulated by both exogenous and endogenous c- Jun, that c-Jun can relieve AR self-squelching, and that the c-Jun eect is primary (Bubulya et al., 1996), all Figure 1 Schematic structure of c-Jun mutants. The various consistent with c-Jun serving a coactivator role in AR- leucine zipper (LZ) mutants include the C-terminal truncation mediated transactivation. mutant, JunD287 ± 331, which lacks a LZ, and the point mutants, c-Jun has been extensively characterized as a M8, M9 and M14. M8 and M9 each have single leucine residues component of AP-1, a transcription factor that directly mutated, whereas M14 has two leucine residues mutated. c-Jun Ala63/73 has alanines substituted for serines 63 and 73. The activates transcription of genes containing DNA various transactivation mutants include the N-terminal truncation sequences known as TREs (TPA-responsive elements) mutant, JunD1 ± 245, which lacks all known activation functions (reviewed in Angel and Karin, 1991). The AP-1 family of (AFs), and the internal deletion mutants, JunD194 ± 223 and 265 transcription factors belong to the bZIP protein family JunD146 ± 221. The basic region mutant JunA?D In265, has a one amino-acid substitution of glutamic acid for alanine at (reviewed in Angel and Karin, 1991) and contains Jun- residue 265 and after a three amino-acid insertion of glutamines and Fos-related proteins. c-Jun binds to TREs either as (Brown et al., 1996) a homodimer or heterodimer with either other Jun proteins or Fos proteins. Studies have identi®ed domains within c-Jun mediating dimerization, DNA binding, transactivation, and cellular transformation (reviewed in These proteins were expressed from expression Angel and Karin, 1991). More recent studies examining plasmids driven by either the rous sarcoma virus the transactivation properties of c-Jun have reported (RSV) promoter (M8, M9, M14, JunD194 ± 223, and that this proto-oncogene activates AP-1-responsive JunD146 ± 221) or cytomegalovirus (CMV) promoter promoters by interacting with coactivator proteins. It (JunD287 ± 331, JunD1 ± 245, JunA?D265In265, and c- has been proposed that two of these proteins, CBP Jun Ala63/73). These plamids were cotransfected in (Bannister et al., 1995) and P300 (Lee et al., 1996), highly Cos-1 cells with a pSG5-derived hAR (Bubulya et al., related coactivators important for both AP-1 and 1996) and a reporter construct containing the nuclear receptor activity, may mediate the mutual Escherichia coli chloramphenicol acetyltransferase transcriptional interference between AP-1 and nuclear (CAT) gene under the control of the hAR-inducible receptors (Kamei et al., 1996). mouse mammary tumor virus (MMTV) promoter While this ®nding with CBP/P300 may provide a (Shemshedini et al., 1991). potential molecular mechanism by which AP-1 and nuclear receptors negatively interact, it is unlikely to The bZIP region of c-Jun is required, but not sucient, provide a better understanding of the c-Jun positive for mediating hAR-induced transactivation interaction with AR. As a way to better understand this c-Jun eect, we have begun studies to identify The bZIP region of c-Jun, consisting of the basic functional domains within c-Jun important for hAR region and leucine zipper, is required for dimerization activity. In this paper we describe results indicating and DNA binding (reviewed in Angel and Karin, that multiple regions of c-Jun are involved in mediating 1991).