Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press The mammalian transcriptional repressor RBP (CBF1) targets TFIID and TFIIA to prevent activated transcription Ivan Olave, Danny Reinberg,1 and Lynne D. Vales2 Department of Biochemistry and 1Howard Hughes Medical Institute, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, Piscataway, New Jersey 08854 USA RBP is a cellular protein that functions as a transcriptional repressor in mammalian cells. RBP has elicited great interest lately because of its established roles in regulating gene expression, in Drosophila and mouse development, and as a component of the Notch signal transduction pathway. This report focuses on the mechanism by which RBP represses transcription and thereby regulates expression of a relatively simple, but natural, promoter. The results show that, irrespective of the close proximity between RBP and other transcription factors bound to the promoter, RBP does not occlude binding by these other transcription factors. Instead, RBP interacts with two transcriptional coactivators: dTAFII110, a subunit of TFIID, and TFIIA to repress transcription. The domain of dTAFII110 targeted by RBP is the same domain that interacts with TFIIA, but is disparate from the domain that interacts with Sp1. Repression can be thwarted when stable transcription preinitiation complexes are formed before RBP addition, suggesting that RBP interaction with TFIIA and TFIID perturbs optimal interactions between these coactivators. Consistent with this, interaction between RBP and TFIIA precludes interaction with dTAFII110. This is the first report of a repressor specifically targeting these two coactivators to subvert activated transcription. [Key Words: RBP; transcriptional repression; TFIIA/TFIID targeting] Received November 17, 1997; revised version accepted April 1, 1998. The role of the cellular RBP protein in gene regulation Barr encoded EBNA2 (Grossman et al. 1994; Henkel et al. has been established fairly recently. Earlier genetic stud- 1994; Laux et al. 1994; Waltzer et al. 1994; Zimber-Strobl ies demonstrated a pivotal role for the Drosophila homo- et al. 1994); EBNA3A,B,C (Robertson et al. 1996); and the log of RBP in development (Nash 1965, 1970; Furukawa mammalian proteins Notch (Fortini and Artavanis-Tsa- et al. 1992; Schweisguth and Posakony 1992; Tun et al. konas 1994; Hsieh et al. 1996)] and KyoT2 (Taniguchi et 1994). Initial studies on mammalian RBP contributed al. 1998). biochemical and genetic characterizations, although The identification of RBP as a transcriptional repressor RBP was thought to be a recombinase at that time in mammalian cells was established during studies of (Hamaguchi et al. 1989, 1992; Matsunami et al. 1989; virus gene expression. The important role of RBP in regu- Kawaichi et al. 1992). Since then, RBP has been recog- lating gene expression in the mature cell is highlighted nized to be a transcription factor that represses mamma- by its sequestration during virus infection. Both adeno- lian gene expression (Dou et al. 1994; Kannabiran et al. virus and Epstein-Barr virus evolved to sequester RBP for 1997; Plaisance et al. 1997; Oswald et al. 1998), but ac- viral advantage. The adenoviral gene encoding the capsid tivates Drosophila gene expression (Brou et al. 1994). protein polypeptide IX (pIX) is expressed only after viral RBP has recently been implicated in the Notch signal DNA replication in infected cells. We found that the pIX transduction pathway in Drosophila, which may bridge promoter contained a repressive element that was spe- the role of RBP in gene expression and development (For- cifically bound by cellular RBP. RBP was shown to re- tini and Artavanis-Tsakonas 1994; Bailey and Posakony press pIX expression before viral DNA replication oc- 1995; Lecourtois and Schweisguth 1995; Hsieh et al. curs; this repression was dependent on the presence of 1996; Eastman et al. 1997). That RBP has a pivotal regu- the RBP-specific DNA-binding site within the pIX pro- latory role in gene expression is highlighted by its inter- moter. Furthermore, purified RBP protein was shown to action with viral and cellular proteins that modulate its repress pIX transcription in a reconstituted transcription activity [Drosophila Hairless (Brou et al. 1994); Epstein– assay performed in vitro (Dou et al. 1994). RBP was shown to be usurped during Epstein-Barr vi- 2Corresponding author. rus infection by the potent transcriptional activator E-MAIL [email protected]; FAX (732) 235-4783. EBNA2. EBNA2 lacks DNA-binding activity, but a com- GENES & DEVELOPMENT 12:1621–1637 © 1998 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/98 $5.00; www.genesdev.org 1621 Downloaded from genesdev.cshlp.org on September 26, 2021 - Published by Cold Spring Harbor Laboratory Press Olave et al. plex formed between RBP and EBNA2 facilitates EBNA2 ous report showing that RBP does not occlude Sp1 or tethering to the DNA via the RBP-specific DNA-binding TBP binding, RBP does not occlude binding of Sp1, activity (Grossman et al. 1994; Henkel et al. 1994; Laux TFIIA, and TFIID. All four proteins bind the pIX pro- et al. 1994; Waltzer et al. 1994; Zimber-Strobl et al. moter concomitantly. Instead of repressing by occlusion, 1994). Transcriptional activation by EBNA2 is expedited our results show that RBP mediates repression by direct not only by RBP-mediated access to specific promoters, interaction with two coactivators: TFIIA and the but also by EBNA2 masking of the RBP repression do- TAFII110 subunit of TFIID. Moreover, the component of main (Hsieh and Hayward 1995). TFIID targeted by RBP is the same TAF that interacts Transcriptional repression in eukaryotes has been ap- with TFIIA as well as with the activation domains of preciated relatively recently (for review, see Herschbach Sp1, the only activator required for optimal pIX expres- and Johnson 1993; Johnson 1995). Several repressors sion. However, our results also show that the domain of have been identified and their critical role in regulating TAFII110 targeted by RBP is the same as that which in- gene expression has been established. Similar to the teracts with TFIIA, but distinct from that which inter- precedents set by studies of prokaryotic repressors, dif- acts with Sp1. Furthermore, interaction between RBP ferent eukaryotic repressors that target distinct compo- and TFIIA precludes interaction of TAFII110. Therefore, nents and stages of the transcription process have been RBP interaction with TFIID and TFIIA alters optimal documented recently. Some repressors target activated interaction between these two coactivators, not to dis- transcription by competing with activators for overlap- lodge them from the promoter, but instead to subvert ping DNA-binding sites (e.g., Oct-1 and C/EBP; Wu et al. activated transcription. 1997). Other repressors interact with activators to mask activation domains (e.g., MDM2 and p53; Oliner et al. 1993). Activator–repressor interactions can also serve to Results tether the repressor to specific promoters. Then the re- pressor can effectively compete with other activators for Repression depends on the position of the RBP site basal transcription factors (e.g., MDM2 and TFIIE–TBP; in vivo Thut et al. 1997) or compete with a basal transcription factor for interaction with other activators (e.g., pRb and Figure 1a shows the simple adenoviral pIX promoter con- TBP; Weintraub et al. 1995). Some repressors complex taining a single Sp1 site, a single RBP site, and a consen- with basal transcription factors to exclude interaction sus TATA box. Our previous studies showed that this with other factors (Dr1 and TBP; Inostroza et al. 1992), region is sufficient for pIX regulation, that RBP-mediated whereas others can access promoters directly and target repression required the RBP-binding site, and that RBP candidate basal transcription factors to silence gene ex- binding did not dislodge Sp1 or TBP from the pIX pro- pression (e.g., Kruppel and TFIIB–TFIIEb; Sauer et al. moter (Dou et al. 1994). We initiated our studies to iden- 1995). RBP is a transcriptional repressor with specific tify the molecular basis of RBP repression by testing DNA-binding activity, and RBP binding was known to whether the position of the RBP site immediately up- be required for repression, yet the molecular basis of stream of the TATA box was fortuitous or important in RBP-mediated repression was not known. repression. Our initial studies of the adenoviral pIX gene sug- The single RBP site was repositioned at several dis- gested that RBP-mediated repression does not involve tinct sites within the promoter region of the pIX gene. In competition between RBP and other transcription fac- each case, the wild-type RBP site was substituted with tors for promoter access (Dou et al. 1994). The pIX pro- linker sequence and the optimal distance between the moter is relatively simple, containing a single site for the Sp1 site and the TATA consensus was maintained (Fig. cellular transcription factor Sp1 and a consensus TATA 1a). In two cases, the RBP site was repositioned upstream box (Babiss and Vales 1991). The RBP-specific DNA- of the Sp1 site. RBP/20nt/Sp1 contains an RBP site 20 binding site lies between these two positions immedi- nucleotides upstream of the Sp1 site, whereas RBP/Sp1 ately upstream of the TATA box. However, RBP binding contains an RBP site immediately upstream of the Sp1 did not dislodge either Sp1 or TBP from their respective site, similar to its normal position immediately up- sites. With this as a basis, we pursued the molecular stream of the TATA box. In two other cases, the RBP site mechanism by which RBP successfully silences pIX tran- was maintained between Sp1 activator and TATA and scription. placed either immediately downstream of Sp1 (Sp1/RBP) Using a transcription assay reconstituted with recom- or centered between the two sites (RBP center). The re- binant factors and highly purified native factors, we es- positioned RBP site restores RBP-specific DNA-binding tablished conditions for Sp1-activated pIX transcription activity in each case (data not shown).