Volunle 307, number 1, 81-86 FEBS 11244 July 1992 0 1992 Federation of European Biochemical Societies 00145793/92/~5.00 Minireview How do eukaryotie activator proteins stimulate the rate of transcription by RNA polymerase II? Jonathan Ham, Gertrud Steger and Moshe Yaniv Uniti des Wrus Otrcog~nes, DGpurtemetrtdes Biotecbtologies, Institut Pusteur, _35 rue du Dr. Ram, 75724 Park Cedex 15, France Received 18 May t992 A large number of activator proteins have now been identified in higher and lower eukaryotes, which bind to the regulatory rcpjons of protein. encoding genes and increase the rate at which tbcy are transcribed by RNA polymerasc II. The mechanism by which activators function is being intensively studied and some of the targets of transcriptional activation domains have now been identified. These studies have also revealed novel classes of regulatory factors, which were not anticipated by extrapolating from the principles obtaira!! with prokaryotic promoters. Activator: RNA polymerasc 11; Mechauism of activation; General transcription faclor; Coactivalor 1. INTRODUCTION 2. THE TRANSCRlPTION INITIATION COM- PLEX A major goal in the field of eukaryotic gene regula- tion is to understand how the activator proteins that The region of a promoter in which the RNA polym- bind to the upstream promoter elements and enhancers erase 11 transcription initiation complex is assembled of RNA polymerase II-dependent genes stimulate the Aridwhere transcription initiates is known as the core rate at which these genes are transcribed. RNA polym- or minimal promoter (Fig. 1). Usually this contains a erase II is unable to recognize promoters on its own and TATA box sequence [3], which specifies the direction of is assisted by a number of accessory proteins, referred transcription and the site of initiaiion, and which will to as the general transcription factors. A simple hypoth- allow a basal level of transcription in vitro. In addition esis is that transcriptional activators directly interact to the TATA box, or as an alternative, some core pro- with either RNA polymerase 11 or one or more of the moters contain sequences around the initiation site (ini- general transcription factors, and thereby stimulate the tiators) which can also direct the assembly of an initia- rate of formation or activity of the transcription initia- tion complex (see [4]). The general transcription factors tion complex or increase the number of complexes were originally defined as factors necessa.ry for the in formed (see [l] for a discussion of this and alternative vitro transcription of core promoters by RNA polym- ideas). Here, we will discuss the results of a number of erase II, and were thought to be required by all promot- experiments that have attempted to identify the targets ers. General transcription factors wem identified by of transcriptional activation domains. However, it is chromatographically fractionating extracts prepared important to recall that eukaryotic genes are associated from cell lines or tissues and testing the importance of with histones in vivo, which have been found to exclude particular fractions for transcription in ,vitro. Most of the general transcription factors from the promoter the general transcription factors have now been purified both in ,rivo and in vitro and which appear to have a to homogeneity and cDNAs encoding the polypeptides repressive effect on transcription. Therefore activator associated with particular activities have been isolated proteins might also stimulate transcription by altering (see [5,6] and Table I. chromatin structure so as to make the promoter accessi- The transcription initiation complex is assembled in ble to RNA polymerase II and its associated factors. a temporally ordered fashion [21-231. The first event is Evidence in favour of this hypothesis has recently been the binding of TFIID to the TATA box sequence. In reviewed elsewhere [2]. contrast to many activator proteins, TFIID binds slowly to the promoter, but once formed, the TFIID- Correspondemurldrcss: J. Ham, Unite des Virus Oncogencs, D&par- TATA box complex is very stable (see [24] for refs.). The tement des Biotechnologies, Institut Pasteur, 25 rue du Dr. Roux, binding of TFIID is facilitated by another factor, 75724 Paris Ccdcx 15, France. TFIIA, which is unable to bind to the promoter on its Published6y ElsrvierScience P&hirers B, V. 81 Volume 307, number 1 FEBS LETTERS July 1992 activator protalns prelnltletlon complex 3. TFIID AND TFIIB AS TARGETS OF ACTIVA- TION DOMAINS 3.1. The htportuttce oJ TFiiD TFIID has attracted considerable attention as a po- tential target for transcriptional activators, since it ap- pears to play a crucial role in the assembly of the tra.n- scription initiation complex on all promoters tested so far in vitro. In the case of core promoters with a TATA box, TFIID is the only general transcription factor upstream promoter which can specifically recognize the promoter on its elements own. The binding of TFIIB, Pol II, TFIIE and TFlIF I. of RNA-polymeruse preinitiation on depends on the prior formation of a TFIID-TATA box typical promoter. binding the binding complex (see, for example, [X!]). which (TBP). is of to TATA (consensus tain initiator but a box is by in promoter an (INR) at +l The assembly of TFIIB, require [27,2S], may recruited the followed by the RNA-polymcrasc. which is associated with TFIIE and by of direct with TFIIF creates an active preinitiation complex (PIG). In a promoter [61. containing INR scquznccs (consensus YAYTCYYY. where Y=pp A number of early studies either showed that activa- rimidinc), TFII-1 is able to substituie for TFIIA, indicalingaltcrnativc tor proteins could alter the binding of TFIID to the pathways for PIC formation. whereas in a TATA4css promoter, TFII. I is necessary for TFIID binding. The assembly of the PIC might be promoter (USF:[29]; GAL4 derivatives: [30]; ATF: [31]) facilitated by the action of transcriptional trnnsactivator proteins. or implicated TFIID as a target (pseudo rabies virus IE They usually bind as homodimcrs to their binding sites (BS) located protein: [32]). However, all of these experiments were upstream (or downstream) of the TATA-box and might activate Iran- performed with partially purified TFIID and therefore scription by contacting one or more of the &cncral transcription fat- tars. Activator proteins arc composed of distinct domains, a DNA- it could not be excluded that the activators interacted binding and dimerieation domain and a transactivation domain, with other proteins in the TFIID fraction. Subsequent which may be separated by a binge region, Sonic aciivalors do not work has shown that TFIID is in fact a large multipro- bind to specific DNA sequences and recognize the promotrr by inter- tein complex consisting of a TATA box-binding protein acting with otbcr transcription factors already bound to the DNA c,g. (TBP) and a variable number of associated polypeptides VP16 interacts with OCTI and Ela binds to ATF. referred to as TAFs (TBP associated factors; reviewed in [33]j. The TBP has now been cloned from several species (reviewed in [34]) and its structure is shown in Fig. 2. Recombinant TBP can efficiently bind to the own. Once the D-A complex is formed, TFIIB binds to TATA box on its own and will support basal transcrip- the promoter downstream of the TATA box over the tion of a core promoter in vitro when mixed with the initiation site. The D-A-B complex is then recognized other general transcription factors. Is there any evi- by RNA polymerase II in association with TFIIE and dence to suggest that activators directly interact with the TFlIF. This complex is functionally analogous to the TBP? prokaryotic closed complex. Transcription initiation re- quires hydrolysis of ATP and leads to melting of pro- 3.2, The VP16 rxtiwlor moter DNA and open complex formation [25] followed The VP16 protein of herpes simplex virus I contains by mRNA initiation and elongation. In theory, any of a transactivation domain rich in acidic amino acid resi- the steps in the assembly of the preinitiation complex dues, and is one of the best-characterized cukaryotic could be rate limiting on a particular promoter in viva activator proteins (see, for example [35]). In contrast to and might be a target for transcriptional activator pro- many other transcription factors VP1 6 does not bind to teins. Originally the general transcription factors were a specific DN.4 sequence and recognizes the promoter thought to be required by all polymerase II promoters. by interacting with another factor (OCTl) already However, the situation is probably more complex than botrnd to the DNA. Stringer et al. [36] first showed that this. For example, a new factor has been isolated, TFII- both the TFIID activity in HeLa cell nuclear extracts I, which binds to the initiator and can functionally re- and recombinant yeast TBP could speciIically bind to place TFlIA [20]. However, it can only be used by pro- columns containing agarose beads linked to a fusion moters that contain initiator elements. Similarly, whilst between protein A and the transactivation domain of TFIlE is essential for transcription of the adenovirus VP1 6. However, since the TBP is relatively basic and the major late promoter in vitro it is apparently not re- VPlG activation domain is highly acidic, it was neces- quired for the immunoglobulin heavy chain core pro- stir;) to show that hi.nding was not the result of a non- moter, which in turn requires two novel activities re- specific ionic interaction. The same group therefore ferred to as 700 kDa and 90 kDa [26]. tested the yeast TBP on columns prepared from mutant 82 Volume307, number 1 FEBS LETTERS July 1992 VP16 proteins with reduced transcriptional activity [37]. domain but has also been shown to interact directly Cress and Trierenberg [35] had previously shown that with the TBP f40,41].