A Role for the TFIIH XPB DNA Helicase in Promoter Escape by RNA Polymerase II*

TFIIH is an RNA polymerase II transcription factor that performs ATP-dependent functions in both transcription initiation, where it catalyzes formation of the open complex, and in promoter escape, where it suppresses arrest of the early elongation complex at promoter-proximal sites. TFIIH possesses three known ATP-dependent activities: a 3′ → 5′ DNA helicase catalyzed by its XPB subunit, a 5′ → 3′ DNA helicase catalyzed by its XPD subunit, and a carboxyl-terminal domain (CTD) kinase activity catalyzed by its CDK7 subunit. In this report, we exploit TFIIH mutants to investigate the contributions of TFIIH DNA helicase and CTD kinase activities to efficient promoter escape by RNA polymerase II in a minimal transcription system reconstituted with purified polymerase and general initiation factors. Our findings argue that the TFIIH XPB DNA helicase is primarily responsible for preventing premature arrest of early elongation intermediates during exit of polymerase from the promoter.

TFIIH was initially identified by its requirement in transcription initiation by RNA polymerase II (7). Initiation is an ATP-dependent process that requires at minimum the five general initiation factors TFIIB, TFIID, TFIIE, TFIIF, and TFIIH (8,9). Biochemical studies have shown that initiation in this minimal transcription system proceeds through multiple stages beginning with assembly of polymerase and all five general initiation factors into a closed preinitiation complex at the promoter (8,9) and culminating in ATP-dependent formation of the open complex and synthesis of the first phosphodiester bond of nascent transcripts (10 -13). Evidence supporting a role for TFIIH DNA helicase activity in ATP-dependent formation of the open complex was initially suggested by studies indicating that both TFIIH and ATP are dispensible for initiation by RNA polymerase II from artificial promoters containing premelted transcriptional start sites and from promoters on negatively supercoiled DNA templates (14 -19).
In addition to its requirement in transcription initiation, TFIIH is also required for efficient promoter escape by RNA polymerase II (18, 20 -22). Mechanistic studies have shown that a fraction of early RNA polymerase II elongation intermediates are prone to arrest at promoter-proximal sites in the absence of TFIIH or an ATP cofactor (18,(21)(22)(23). Circumstantial evidence that TFIIH DNA helicase activity is responsible for suppressing arrest of early elongation intermediates has come from the observation that promoter escape is blocked by the TFIIH DNA helicase inhibitor ATP␥S, but not by the TFIIH CTD kinase inhibitor H-8 (18).
Although evidence from previous studies suggested that TFIIH DNA helicase activity is required for ATP-dependent formation of the open complex and ATP-dependent promoter escape, a direct test of this hypothesis was not possible until sufficient quantities of purified TFIIH mutants lacking functional XPB or XPD DNA helicase were available. Recently, some of us (F. Tirode and J.-M. Egly) reported the development of methods for reconstitution of TFIIH and TFIIH subassemblies from wild type and mutant subunits (2,4). By investigating the activities of TFIIH mutants, we observed that maximal TFIIH transcriptional activity requires all nine subunits, although the TFIIH subassembly IIH6 lacking CAK is active in ATP-dependent formation of the open complex and supports a reduced level of runoff transcription (4). In addition, by comparing the activities of IIH6 and two IIH6 mutants, IIH6/XPB-K346R and IIH6/XPD-K48R, which contain point mutations in the XPB and XPD ATP binding sites and lack DNA helicase activity (24,25), we obtained evidence supporting the model that the XPB DNA helicase is essential for formation of the open complex and runoff transcription and that the XPD DNA helicase, though not essential, stimulates these reactions (2).
In this report, we exploit recombinant TFIIH mutants lacking functional XPB DNA helicase, XPD DNA helicase, or CAK to investigate the contribution of TFIIH DNA helicase and CTD kinase activities to efficient promoter escape. Our findings argue that the XPB DNA helicase is primarily responsible for TFIIH action in suppression of arrest of early RNA polymerase II elongation complexes during their escape from the promoter.

RESULTS AND DISCUSSION
To investigate the roles of the XPB and XPD DNA helicases and CAK in TFIIH-dependent promoter escape, we compared the abilities of IIH6 and two IIH6 mutants, IIH6/XPB-K346R and IIH6/XPD-K48R, which contain point mutations in the XPB and XPD ATP binding sites and lack DNA helicase activity (24,25), to suppress arrest of early RNA polymerase II elongation intermediates in a minimal transcription system reconstituted with purified polymerase and general initiation factors TBP, TFIIB, TFIIE, and TFIIF. IIH6 and IIH6 mutants were expressed in Sf9 cells coinfected with baculoviruses encoding human TFIIH subunits p34, p44, p52, p62, wild type or mutant XPD, and wild type or mutant XPB (2). Recombinant IIH6 and IIH6 mutants were purified from lysates of Sf9 cells by sequential heparin ultrogel and anti-p44 immunoaffinity chromatography (2,32). Recombinant CAK was purified from lysates of Sf9 cells coinfected with baculoviruses encoding CDK7, cyclin H, and MAT1 (4). The subunit compositions of wild type and mutant IIH6 complexes and CAK were verified by Western blotting, and the relative concentrations of wild type and mutant IIH6 complexes were estimated by quantitative Western blotting ( Fig. 1 and data not shown).
To characterize the transcriptional activities of IIH6 and IIH6 mutants, we began by using a dinucleotide-primed abortive initiation assay to compare their abilities to support transcription initiation from the AdML promoter in the minimal transcription system. In the presence of an ATP cofactor, RNA polymerase II will utilize dinucleotides to prime synthesis of promoter-specific transcripts (34). If only a dinucleotide primer and the next nucleotide encoded by the template are provided as substrates for transcription, polymerase will efficiently synthesize abortively initiated, trinucleotide transcripts (10,35,36). We and others have shown previously that abortive initiation from the AdML promoter can be measured in the presence of [␣-32 P]CTP and either initiating dinucleotide CpA or CpU, which prime transcription at positions Ϫ1 and Ϫ3 relative to the normal AdML transcriptional start site ( Fig. 2A) (35,37,38). In addition, we and others have shown that maximal rates of abortive initiation from the AdML promoter depend strongly on an ATP cofactor and all five general initiation factors (21,22,38).
As shown in Fig. 2A, wild type IIH6 stimulated the rate of abortive initiation above the low background level observed in the absence of TFIIH, whereas equivalent concentrations of the XPB mutant IIH6/XPB-K346R and the XPD mutant IIH6/XPD-K48R did not. CAK, which is composed of CDK7, cyclin H, and MAT1 subunits, detectably stimulated the rate of abortive initiation by both wild type IIH6 and the XPD mutant IIH6/XPD-K48R, but not by the XPB mutant IIH6/XPB-K346R (Fig. 2B). These findings are consistent with the results of Tirode et al. (2), who observed that the XPB mutant IIH6/XPB-K346R did not support detectable open complex formation and runoff transcription in the presence or absence of CAK, whereas the XPD mutant IIH6/XPD-K48R was substantially less active than IIH6, but could support a low level of runoff transcription that was stimulated by CAK.
To investigate the activities of IIH6 and IIH6 mutants in promoter escape, we took advantage of the artificial AdML promoter derivative Ad(Ϫ9/Ϫ1), which contains a premelted region from positions Ϫ9 to Ϫ1 relative to the normal transcriptional start site. The Ad(Ϫ9/Ϫ1) promoter supports transcription initiation by RNA polymerase II in the absence of TFIIH and an ATP cofactor and is therefore a useful model for investigating post-initiation roles of TFIIH and ATP (12, 16 -18, 39). We previously observed that maximal synthesis of 18 nucleotide RNAs terminated at the first G residue of the Ad(Ϫ9/Ϫ1) transcript by incorporation of 3Ј-O-MeG requires TFIIH and ATP and is inhibited by ATP␥S (18). Further elongation of the 18-nucleotide transcript is independent of ATP and TFIIH; thus, RNA polymerase II elongation complexes that have completed synthesis of these transcripts can be considered to have escaped the promoter.
To compare the abilities of IIH6 and IIH6 mutants to support efficient promoter escape, RNA polymerase II preinitiation complexes were assembled at the Ad(Ϫ9/Ϫ1) promoter in the minimal transcription system in the presence of either IIH6, IIH6/XPB-K346R, or IIH6/XPD-K48R. Transcription was carried out in the presence of ATP or ATP␥S and the initiating dinucleotide CpU, UTP, [␣- 32  As shown in Fig. 3B, in the presence of ATP␥S, the majority of RNA polymerase II elongation intermediates suffered arrest before completing synthesis of the 18 nucleotide, 3Ј-O-MeGterminated transcript; similar levels of the 18-nucleotide transcript were synthesized whether reactions contained IIH6, IIH6/XPB-K346R, or IIH6/XPD-K48R. Substitution of ATP for ATP␥S increased accumulation of the 18-nucleotide transcript ϳ7-fold in reactions containing IIH6 and ϳ5-fold in reactions containing the XPD mutant IIH6/XPD-K48R. In contrast, substitution of ATP for ATP␥S had no significant effect on accumulation of the 18 nucleotide transcript in reactions containing the XPB mutant IIH6/XPB-K346R, arguing that the XPB DNA helicase makes a significantly greater contribution than the XPD DNA helicase to TFIIH function in ATP-dependent promoter escape.
As shown previously (2) and in Fig. 2B, the presence of the CAK subunits increases TFIIH activity in abortive initiation and in synthesis of runoff transcripts. To investigate the contribution of CAK to TFIIH-dependent promoter escape, IIH6 and IIH6 mutants were assayed in the presence and absence of CAK, and reaction products were analyzed without prior gel filtration. As shown in Fig. 3C, CAK had no detectable effect on the levels of 18 nucleotide transcripts synthesized in the presence of either wild type IIH6, IIH6/XPB-K346R, or IIH6/XPD-K48R, arguing that CAK does not contribute significantly to TFIIH-dependent promoter escape. Because these reactions were not gel-filtered, a large number of abortive transcripts can be observed. Nonetheless, the relative amounts of 18 nucleotide transcript synthesized in the presence of wild type IIH6, IIH6/ XPB-K346R, and IIH6/XPD-K48R are comparable with those seen when reaction products were gel filtered prior to polyacrylamide gel electrophoresis (Fig. 3B, lanes 1, 3, and 5).
In summary, in this report we have taken advantage of recombinant TFIIH mutants to investigate the contributions of TFIIH DNA helicase and CTD kinase activities to efficient promoter escape by RNA polymerase II in a minimal transcription system reconstituted with purified polymerase and general initiation factors. By comparing the activities of the TFIIH subassembly IIH6 and two IIH6 mutants, IIH6/XPB-K346R and IIH6/XPD-K48R, which contain point mutations in the XPB and XPD ATP binding sites and lack DNA helicase activity (24,25), we have obtained evidence supporting the model that the XPB DNA helicase is primarily responsible for TFIIH action in ATP-dependent promoter escape. We observe (i) that the IIH6 point mutant IIH6/XPB-K346R, which contains wild type XPD DNA helicase but lacks functional XPB DNA helicase, is inactive in promoter escape and (ii) that the IIH6 point mutant IIH6/XPD-K48R, which contains wild type XPB DNA helicase but lacks functional XPD DNA helicase, supports promoter escape but less actively than wild type IIH6. Together with the recent findings of Tirode et al. (2), who presented evidence supporting the model (i) that the XPB DNA helicase is essential for ATP-dependent formation of the open complex and (ii) that the XPD DNA helicase stimulates this reaction, our results indicate that the relative contributions of the XPB and XPD DNA helicases to promoter escape closely parallel their contributions to open complex formation and suggest that TFIIH performs similar roles during both open complex formation and promoter escape.