A: Identification of TATA-Binding Activity

For PIC assembly via the sequential assembly pathway or the two-component pol II holoenzyme pathway, TFIID recognition of the promoter is usually the first step to initiate the formation of a transcriptionally competent PIC. As described earlier, TFIID was originally identified as a chromatographic fraction necessary to support site-specific transcription by pol II in vitro (Matsui et al., 1980). The TFIID chromatographic fraction that Roeder and colleagues identified was still very crude and many laboratories sought to isolate the key component(s) in TFIID. Interestingly, TFIID was identified to serve a key function in binding to the TATA box, initially from Drosophila (Parker and Topol, 1984), then mammals (Sawadogo and Roeder, 1985b), and yeast (Buratowski et al., 1988; Cavallini et al., 1988). Once TFIID bound to the TATA box, it was hypothesized to function as a scaffold upon which the PIC could assemble (Workman and Roeder, 1987; Hai et al., 1988; Horikoshi et al, 1988). Eventually a single polypeptide containing the TATA box-binding activity (later termed TBP for TATA box-binding protein) was purified (Horikoshi et al., 1989a) and cloned from yeast (Hahn et al, 1989; Horikoshi et al, 1989b). The cloning of Drosophila and human TBP factors soon followed (Hoey et al, 1990; Kao et al, 1990; Peterson etal., 1990).

The immediate question that followed was whether this single polypeptide TBP is the functional equivalent of TFIID. In an elegant experiment testing for transcriptional activation dependent on the transcriptional activator Spl (specificity protein 1), it was found that Spl could stimulate transcription from TATA-containing promoters only in the presence of partially purified TFIID, but not when TFIID was replaced by recombinant Drosophila or yeast TBP (Pugh and Tjian, 1990). Furthermore, glycerol gradient sedimentation and immunoprecipitation analyses of partially purified TFIID, which supported activator-mediated transcription, indicated that TFIID was a multiprotein complex rather than a single polypeptide (Dynlacht et al, 1991; Pugh and Tjian, 1991). Thus, it was proposed that additional factors, or coactivators, in conjunction with TBP were necessary to potentiate transcriptional activation by Spl. It was subsequently determined that TFIID is a multiprotein complex composed of TBP and TBP-associated factors (TAFs) (Dynlacht et al, 1991; Tanese et al, 1991). Current techniques of immunoaffinity purification using antibodies directed against TBP or an epitope tag linked to the TBP-coding sequence have pulled down TBP and TAFs (Dynlacht et al, 1991; Zhou et al, 1992; Chiang et al, 1993; Poon and Weil, 1993; Sanders et al, 2002; Auty et al, 2004). This has greatly simplified the purification scheme for TFIID and further facilitated the identification and cloning of TAFs. In humans, at least 13 TAFs have been identified (Burley and Roeder, 1996; Hahn, 1998; Albright and Tjian, 2000; Green, 2000; Tora, 2002; see also the Chapter by Thomas and Chiang, 2005). A cell type-specific TAF, TAF4b (formerly named TAFh105) has also been isolated from B cells and found to function as a coactivator for NF-kB (Dikstein et al, 1996b; Matza et al., 2001). The current concept of TFIID is that it functions as: (D a coactivator in mediating interactions between activators and GTFs to enhance PIC assembly, ® a core promoter recognition factor for both TATA-containing and TATA-less promoters, and (3) an enzyme to posttranslationally modify protein factors involved in transcriptional regulation.

B: TFIID as a Coactivator

Characterization of TAFs revealed that many gene-specific activators interact directly with specific TAFs (reviewed by Burley and Roeder, 1996; Verrijzer and Tjian, 1996). For example, the activation domain of Spl was found to interact with Drosophila TAF4 (dTAFnllO; Hoey et al., 1993), whereas the DNA-binding domain of Sp 1 was found to interact with human TAF7 (hTAFn55; Chiang and Roeder, 1995). The finding that distinct domains in an activator contact different TAFs suggests that TFIID may modulate activator function through multiple domain interactions (Chiang and Roeder, 1995). While many other examples of activator-TAF interactions exist, human TAF7 (hTAFn55) is one particular TAF shown to interact with multiple activators (Chiang and Roeder, 1995; Lavigne et al., 1999). These activator-TAF interactions imply that activators may indeed function by recruiting TFIID to the promoter in order to nucleate PIC formation (Burley and Roeder, 1996; Verrijzer and Tjian, 1996; Albright and Tjian 2000; Naar et al., 2001). Consistent with its role as a coactivator, only purified TFIID, but not TBP, supports activator-mediated transcription in partially purified cell-free transcription systems (Dynlacht et al., 1991; Chiang et al., 1993). Thus the in vitro biochemical concept that TFIID was a universal coactivator required for all gene transcription was formed.

Although the universal requirement for TFIID as a coactivator was proposed based on in vitro studies, TFIID in vivo appears to play a more limited role, as subsequent yeast genetic studies have revealed that TAFs are not universally required for activated transcription as previously thought (Moqtaderi et al., 1996; Walker et al., 1996). Rather, TAFs may mediate activator-dependent transcription from only a subset of genes. These results indicate that individual TAFs may be important for mediating distinct activator-dependent transcription in vivo (Walker et al., 1997; Apone et al., 1998; Holstege et al., 1998; Mencia et al., 2002). Subsequent cell-free transcription studies using highly purified and well-defined transcription systems have also shown that TBP, in the absence of TAFs but in conjunction with other GTFs, pol II and general cofactor PC4, may direct activated transcription from TATA-containing promoters in an activator-specific manner with naked (nucleosome-free) DNA templates (Oelgeschlager et al., 1998; Wu and Chiang, 1998; Wu etal., 1998, 1999; Fondell et al., 1999).

C: TFIID as a Core Promoter Recognition Factor

In addition to mediating activator-dependent recruitment of GTFs, TFIID has been implicated as a core promoter recognition factor. Besides making contact with the TATA box through TBP binding, TFIID may contact the Initiator (Inr) via Drosophila TAF2 (dTAFnl50) or the downstream promoter element (DPE) via Drosophila TAF6 (dTAFn60) and TAF9 (dTAFn40; Burke and Kadonaga, 1997; Kaufmann et al, 1998; Martinez et al., 1998). Recent evidence has also shown that human TAF6 and TAF9 display sequence-specific binding to the core promoter of human interferon regulatory factor-1 (IRF-1) gene, which bears a functional DPE (Burke and Kadonaga, 1997; Shao et al, 2005). DNase I footprinting experiments showing that TFIID gives extended protection over the TATA box, Inr, and DPE regions are consistent with a role of TFIID functioning as a core promoter recognition factor. These TFIID-DNA interactions may be especially important at promoters lacking a canonical TATA sequence, since TAFs are required for transcription from TATA-less promoters (Pugh and Tjian, 1991; Martinez et al., 1994; Orphanides et al, 1996; Smale, 1997). Furthermore, binding of TAFs to core promoter elements directs promoter selectivity by pol II (Hansen and Tjian, 1995; Verrijzer et al, 1995). Studies in yeast also indicate that the requirement for individual TAFs may depend on core promoter sequences (Kuras et al, 2000; Li et al., 2000, 2002). Thus, a selective requirement of TAFs is determined by both gene-specific activators and the promoter context.

D: Histone-Like Domains in TAFs

The observation that TAFs share structural homology with histones (Hoffmann et al, 1996; Xie et al, 1996; reviewed by Gangloff et al, 2001) suggests TAFs in TFIID may form a histone-like octamer structure. Indeed, initial crystal structures of TAF9/ TAF6 (dTAFn42/dTAFn62) showed that there exists a heterotetramer, resembling the H3/H4 heterotetrameric core of the histone octamer. This finding indicates that TFIID may contain a histone octamer-like substructure (Xie et al., 1996). Further supporting evidence for a histone-like octamer was demonstrated when it was found that yeast TAF9-TAF6-TAF12-TAF4 (yTafl7-yTaf60-yTaf61-yTaf48) may reconstitute an octamer in vitro (Selleck et al, 2001). The immediate question that follows is whether the histone-like octamer of TAFs is able to contact DPE in the context of a nucleosome-like structure. In a study investigating protein-protein interactions among human TAF9, TAF6, TAF4b, and TAF12, which contain sequences related to histones H3, H4, H2A, and H2B, respectively, it was found that these TAFs indeed form an octamer-like complex which enhances both sequence-specific and nonspecific DNA-binding activities of TAF9-TAF6 and TAF4b-TAF12 pairs, respectively (Shao et al, 2005). In contrast to the interaction studies, which suggest that TFIID may contain a histone-like octamer, low-resolution electron microscopy (EM) studies of the TFIID complex do not show an octamer-like structure within TFIID (Andel et al., 1999; Brand et al., 1999; Leurent et al., 2002). These EM studies found TFIID to be a trilobed, horseshoe-shaped structure with TBP sitting in the central cavity while TFIIA and TFIIB bound to opposite lobes of the horseshoe-shaped structure (Andel et al., 1999; Brand et al., 1999; Leurent et al, 2002). Each of the globular domains found in the TFIID EM structure is almost equivalent in size to the histone octamer, but none are large enough to hold all six histone motif-containing TAFs (Brand et al, 1999). Therefore, a compact histone octamer-like structure incorporating all the TAFs with histone motifs is unlikely; instead, the histone folds in TAFs may form interfaces for protein-protein interactions (Brand et al, 1999). Although a histone octamer containing all six histone fold motif-containing TAFs may be unlikely due to spatial constraints, the possibility that fewer TAFs may form a nucleosome-like structure remains to be explored.

E: TFIID as an Enzyme

It is well recognized that within the cell, DNA is wrapped around the histone octamer to form a nucleosome and further condensed into a higher order chromatin structure in order to compact DNA within the nucleus. Formation of this higher order chromatin structure prevents transcription factor access to the promoter region. For transcription to occur on chromatin templates, covalent modification on core histones is necessary to loosen up histone-DNA interactions at the promoter region. Therefore it was an important finding that TAF1 (formerly named TAFn250) possesses histone acetyltransferase (HAT) activity to acetyl ate histones H3 and H4 (Mizzen et al., 1996). Other than HAT activity, TFIID also exhibits multiple enzymatic activities via TAF1, such as kinase activity that phosphorylates the RAP74 subunit of TFIIF (Dikstein et al, 1996a), the (3 subunit of TFIIA (Solow et al, 2001), serine 33 of histone H2B (Maile et al., 2004), and PC4 (Kershnar et al, 1998; Malik et al, 1998), and ubiquitin activating/conjugating activity that targets histone HI (Pham and Sauer, 2000). Interestingly, TAF1-mediated phosphorylation of histone H2B correlates with gene activation (Maile et al., 2004). Thus, the ability of TAF1 to modify distinct histones by acetylation, phosphorylation, or ubiquitination is consistent with the finding that TFIID, rather than TBP, is essential for chromatin transcription (Wu et al, 1999), suggesting that TAFs in TFIID may play a role in modifying chromatin structure during the transcription process. The finding that a subset of TAFs are also integral components of other protein complexes, such as SAGA (Spt, Ada, Gcn5 acetyltransferase), required for nucleosome acetylation and transcriptional stimulation (Grant et al., 1998) further supports the view that TAFs in TFIID are able to modify chromatin structure.

0 0

Post a comment