EXPERIMENTAL and MOLECULAR MEDICINE, Vol. 32, No. 2, 53-60, June 2000 Dissecting the molecular mechanism of nuclear receptor action: transcription coactivators and corepressors Jae Woon Lee1,2,4, JaeHun Cheong1,2, nucleosomal remodeling histone acetyl transferase (HAT) Young Chul Lee1,2, Soon-Young Na1,3 or deacetylase (HDAC) activities. Thus, these proteins and Soo-Kyung Lee1,3 appear to function by either remodeling chromatin struc- tures and/or acting as adapter molecules between nu- clear receptors and the components of the basal trans- 1 Center for Ligand and Transcription criptional apparatus. Herein, we discuss the recent pro- 2 Hormone Research Center gress in studies of these coactivators and corepressors 3 Department of Biology, Chonnam National University, of nuclear receptors. Kwangju 500-757, Korea 4 Corresponding author: Tel, +82-62-530-0910; Fax, +82-62-530-0772; E-mail, [email protected] The p160 Family Accepted 21 June 2000 A group of related proteins were found to enhance Abbreviations: HRE, hormone response elements; LBD, ligand ligand-induced transactivation function of several nuclear binding domain; AF2, activation function; HAT, histone acetyl trans- receptors, named the p160 family. These proteins are ferase; HDAC, deacetylase; CBP, CREB-binding protein; TRAPs, grouped into three subclasses based on their sequence homology; i.e., SRC-1/NCoA-1 (Hong et al., 1997; Torchia thyroid homone receptor associated proteins; VDR, vitamin D3; ASC-1, activating signal cointegrator-1; RAR, retinoic acid receptor et al., 1997; Voegel et al., 1998), TIF2/GRIP1/NCoA-2 (Hong et al., 1997; Voegel et al., 1998), and p/CIP/ ACTR/AIB1/xSRC-3 (Anzick et al., 1997; Chen et al., 1997; Kim et al., 1998; Torchia et al., 1997). A distinctive Keywords: Nuclear receptor, transcription, coactivators, structural feature of the p160 coactivators is the corepressors presence of multiple LXXLL signature motifs (Heery et al., 1997; Torchia et al., 1997). The AF2 core (helix 12) was recently shown to undergo a major restructuring Introduction upon ligand binding, forming part of a “charged clamp” that accomodates p160 coactivators within a hydrophobic Over 150 members that belong to the nuclear receptor cleft of the receptor LBD, through direct contacts with superfamily have been discovered since glucocorticoid these LXXLL motifs (Darimont et al., 1998; Nolte et al., receptor was first reported in 1985, which primarily regu- 1998). Loss of function studies using antibody microin- late, in a ligand-dependent manner, transcriptional initi- jection technique also suggested that the p160 family ation of the target genes by directly binding to specific proteins are required for nuclear receptor functions in DNA sequences named hormone response elements vivo (Torchia et al., 1997). In addition, these factors can (HRE) (reviewed in Mangelsdorf et al., 1995). The C- also interact with CREB-binding protein (CBP)/p300 via terminus of the ligand binding domain (LBD) of these a separate domain (Kamei et al., 1996; Yao et al., 1996). proteins harbors an essential ligand-dependent trans- A weak intrinsic HAT activity has been reported in SRC- activation function (activation function 2, AF2), whereas 1 and ACTR, suggesting that a function of these factor the N-terminus of many nuclear receptors often includes may involve chromatin remodeling (Chen et al., 1997; AF1. Genetic studies implicated that transcription cofac- Spencer et al., 1997). We have recently shown that tors with no specific DNA-binding activity are essential SRC-1 also mediates transactivation by a series of other components of transcriptional regulation, which ultimate- transcription factors, including AP-1 (Lee et al., 1998), ly led to identify a series of nuclear receptor-interacting NFκB (Na et al., 1998), SRF (Kim et al., 1998a), and coregulatory proteins (reviewed in Horwitz et al., 1996). p53 (Lee et al., 1999). In particular, SRC-1 and p/CIP Thus far, these proteins have been shown to have a few were strong coactivators for p53, whereas AIB1 and characteristic features, as summarized in Figure 1. First, xSRC-3 were repressive (Lee et al., 1999). It is also they bind to target transcription factors in a ligand- noted that the p160 family of proteins has a number of dependent manner. Second, many of them are capable uncharacterized isoforms (Chen et al., 1997; our un- of directly interacting with the basal transcriptional published results). These results suggest a provoking machinery. Third, some of them exhibit enzymatic func- hypothesis that each member of the p160 family or iso- tion intrinsically linked to gene regulation, such as the forms may regulate a specific set of target transcription 54 Exp. Mol. Med. Vol. 32(2), 53-60, 2000 Figure 1. The role of transcriptional cofactors. Three general functions of known receptor cofactors are denoted as I, II, and III (see the text for details). HRE and +1 denote hormone response elements and transcription initiation site, respectively. Nuclear receptors, nucleosomes, cofactor and RNA polymerase II bound to TATA sequences are schematically depicted. Notably, RNA polymerase II and most cofactors exist as steady-state complex of multiple polypeptides, although each of them is represented as a single polypeptide for simplicity. factors in vivo. p/CAF This protein was first discovered on the basis of sequ- CBP/p300 ence homology to the yeast HAT protein Gcn5p (Yang et al., 1996). The N-terminus of p/CAF interacts with CBP was originally isolated on the basis of its associ- CBP and members of the p160 family, whereas inter- ation with CREB in response to cAMP signaling, where- action interface between p/CAF and nuclear receptors as its close homologue, p300, was purified as a cellular was different from that mediating the binding with either binding protein of the adenoviral protein E1A (Chrivia et CBP/p300 or p160s (Blanco et al., 1998; Korzus et al., al., 1993; Eckner et al., 1994). CBP and p300 have 1998). A core p/CAF complex was recently isolated by been implicated in functions of a large number of regu- exploiting an affinity purification approach, which contain- lated transcription factors (reviewed in Goldman et al., ed human homologues of the yeast ADA proteins, TAFs 1997). For nuclear receptors, the interaction with CBP/ or TAF homologs, and p/CAF associated factor 65α which p300 is ligand- and AF2-dependent, although this direct contains histone-like structure (Ogryzko et al., 1998). interaction does not appear to be essential with many These results suggest a possible link between the p/ nuclear receptors (Westin et al., 1998; Li et al., 2000). In CAF complex and the RNA polymerase II core machi- fibroblasts isolated from a p300−/− mouse, however, RA- nery. This p/CAF complex bears resemblance to the dependent transcription was severely impaired, clearly GCN5/SAGA complex in yeast. In particular, other sub- indicating that CBP/p300 are components of hormonal units of the complex facilitate p/CAF to acetylate his- regulation of transcription in vivo (Yao et al., 1998). tones in the context of nucleosomes, although p/CAF Surprisingly, CBP and p300 harbor HAT activity alone is inert (Grant et al., 1997). (Bannister et al., 1996; Ogryzko et al., 1996). In addition, purified p300 was shown to potentiate ligand- induced ER function only on chromatinized template, TRAP/DRIP Complexes strongly indicating that a major function of CBP/p300 could be to modify chromatin structure via histone The thyroid hormone receptor associated proteins (TRAPs), acetylation (Kraus and Kadonaga, 1998). However, it is composed of at least 9 polypeptides, were immuno- notable that CBP/p300 can also acetylate and purified from cells stably transfected with flag-tagged functionally modulate, either in a negative or positive thyroid hormone receptors (Fondell et al., 1996). In re- manner, non-histone proteins, including TFIIEβ (Imhof et constituted in vitro transcription assays utilizing naked al., 1997), p53 (Gu and Roeder, 1997), hematopoietic DNA templates, the TRAP complex potentiated trans- transcription factor GATA-1 (Boyes et al., 1998) and activation function of liganded TR. A highly homologous erythroid Krüppel-like factor (Zhang and Bieker, 1998). DRIP (vitamin D3 receptor (VDR) interacting protein) These results suggest that CBP/p300 may also target complex was also isolated using VDR as the affinity different aspects of gene activation, in addition to their matrix (Rachez et al., 1998), which substantially poten- roles in chromatin remodeling. tiated ligand-dependent transactivation function of VDR Dissecting the molecular mechanism of nuclear receptor action 55 on a chromatinized template in vitro (Rachez et al., tional coactivator, p52, interacted not only with transcrip- 1999). Interestingly, constituents of DRIP complex are tional activators and general transcription factors to almost identical to another newly discovered ARC (acti- enhance activated transcription but also with the essen- vator recruited cofactor) complex, which is essential for tial splicing factor ASF/SF2 both in vitro and in vivo to a number of other transcription factors such as SREBP, modulate ASF/SF2-mediated pre-mRNA splicing (Ge and NFκB and VP16 (Naar et al., 1999; Rachez et al., Wolfe, 1998). These results, along with the notion that 1999). This TRAP/DRIP complex is recruited to the LBD pre-mRNA splicing can take place cotranscriptionally in AF2 core in response to ligand-binding through a single vivo, suggest that, in addition to functioning as a trans- subunit (DRIP205/TRAP220/TRIP2) via a single LXXLL criptional coactivator, ASC-1 may also act as an adaptor motif (Lee et al., 1995; Naar et al., 1999; Rachez et al., to coordinate pre-mRNA splicing and transcriptional 1999). This protein anchors the other components of the activation of class II genes. DRIP/TRAP complex to the receptor, thereby conferring hormone-dependent recruitment of what appears to be a preformed complex. In addition, TRAP/DRIP/ARC also ASC-2 contain part of the “Mediator complex” (Kim et al., 1994), strongly suggesting their direct connection to the RNA Activating signal cointegrator-2 (ASC-2) is another novel polymerase II core machinery.