INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

RNA polymerase II (pol II) interacts with a number of general and regulatory factors to initiate transcription accurately from a promoter (reviewed in references 34 and 56). In the pathway toward initiation, promoter DNA is bent, and DNA may be wrapped around pol II (24, 40). General factors TATA-binding protein (TBP) (or transcription factor IID [TFIID]), TFIIB, TFIIF, and TFIIE cooperate with pol II to strain the DNA helix around the transcriptional start site before ATP-driven helix opening by TFIIH (34, 56). After initiation, pol II releases from the promoter (promoter clearance or promoter escape), elongates the RNA chain, terminates transcription, and recycles. TFIIF, made up of RAP30 (RNA polymerase II-associating protein of 30 kDa) and RAP74 (58 kDa) subunits, may participate in each of these stages of the transcription cycle.

Inspection of its 517-amino-acid (aa) sequence indicates that human RAP74 can be divided into three regions: (i) a highly basic N-terminal domain with significant globular structure (aa 1 to 217); (ii) an overall acidic, highly charged central region lacking in hydrophobic amino acids but rich in E, D, K, R, S, T, G, and P (aa 218 to 398); and (iii) a very basic C-terminal domain (CTD) with globular structure (aa 399 to 517) (2, 15). The N-terminal domain is important for RAP30 binding (54, 55), preinitiation complex assembly (this report), and elongation stimulation (reference 21 and this report). The CTD of RAP74 makes contact with TFIIB (13) and pol II (54) and stimulates the activity of a pol II CTD phosphatase that may have roles in initiation, elongation, termination, and recycling (8).

A pathway for assembly of preinitiation complexes on TATA box-containing promoters has been defined (34, 56). The TBP subunit of TFIID binds to the TATA sequence. Insertion of TBP into the DNA minor groove at TATA induces a 95° bend (23, 25). TFIIB can then enter to form the DB complex, made up of TBP (or TFIID), TFIIB, and promoter DNA. The C-terminal repeats of TFIIB bind DNA upstream and downstream of TATA, stabilizing the DNA bend (26, 33). The N-terminal domain of TFIIB may extend toward the transcriptional start site as a scaffold on which to assemble pol II and TFIIF (34).

To bind efficiently to the promoter, pol II must first bind TFIIF (10, 16, 22). In some cases, the RAP30 subunit has been sufficient to deliver pol II to the promoter (16, 22, 51), but the RAP74 subunit contributes to proper assembly, complex stability, and initiation. For promoters with weak TATA boxes, both RAP30 and RAP74 contribute to template commitment of TFIID, TFIIB, and pol II (49). Furthermore, RAP74 strongly stimulates initiation from supercoiled and premelted templates that are dependent only on TBP, TFIIB, pol II, and TFIIF for accurate transcription (35, 36). In most contexts, therefore, both the RAP30 and RAP74 subunits are important for TFIIF function in complex assembly and initiation.

After fulfilling its role in initiation, TFIIF stimulates the elongation rate of pol II (4, 20, 21, 38, 49). On nonchromatin DNA templates and in the absence of other general factors, TFIIF can accelerate polymerization to up to 1,500 nucleotides per min, close to the estimated physiological rate (20). TFIIF suppresses pausing by pol II (4, 20, 38), but whether this is a cause or effect of rate stimulation is not known. Both the RAP30 and RAP74 subunits of TFIIF are required for elongation stimulation (21, 49), and preliminary mapping studies indicate that the N-terminal domain of RAP74 is most important for elongation (21). Tan et al. (50) identified a class of RAP30 mutants that are impaired for both elongation stimulation and accurate initiation, and these mutants are also defective for binding RAP74, consistent with the requirement of both subunits for elongation. They have also identified classes of RAP30 mutants that are defective only for elongation stimulation or for initiation, not for both. In contrast to their results with RAP30, however, we find that a region within the N-terminal domain of RAP74, that is not essential for RAP30 binding, is nonetheless strongly stimulatory for both initiation and elongation (this report).

An intriguing feature of the RPB1 subunit of pol II is the CTD which has the consensus sequence YSPTSPS tandemly repeated 52 times in human pol II (reviewed in reference 12). Phosphorylation and dephosphorylation of the CTD by the regulated activities of CTD kinases (14, 29, 31, 45, 46) and phosphatases (7) appear to control progression through the transcription cycle. Pol II enters the preinitiation complex with its CTD in a largely unmodified form designated pol IIa (27, 28). During elongation, pol II is converted to the pol IIo form (3, 27), which is heavily phosphorylated on the SP serines of the YSPTSPS consensus sequence, and so hyperphosphorylation of the CTD is thought to be important to establish and maintain the elongation complex. Removal of phosphates from the CTD may be a signal to terminate transcription and recycle pol II to a promoter.

A recently identified CTD phosphatase that catalyzes the dephosphorylation of pol IIo to pol IIa is stimulated by the C-terminal domain of the RAP74 subunit of TFIIF (8). Interestingly, RAP74-dependent stimulation of CTD phosphatase activity is blocked by addition of TFIIB. The C-terminal domain of RAP74 binds to TFIIB (13) and pol II (54), and so TFIIF, TFIIB, and the CTD phosphatase may be components of a multiprotein complex that binds pol II and regulates pol II recycling. In this report, we demonstrate that the central region and the CTD of RAP74 stimulate multiple-round transcription in an extract system consistent with a role for RAP74 in transcriptional recycling.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Transcription factors and extracts. Recombinant Saccharomyces cerevisiae TBP and human recombinant TFIIB were the kind gifts of Steven Triezenberg and Fan Shen. The clone for production of TFIIB was the kind gift of Danny Reinberg. Recombinant human RAP30, RAP74, and RAP74 mutants were prepared and quantitated as described elsewhere (13, 52, 53, 54). Construction of new mutants is described below. Calf thymus pol II used in electrophoretic mobility shift experiments was prepared by the method of Hodo and Blatti (19) and was primarily in the IIb form, lacking the CTD. Gel mobility shift experiments with calf thymus pol IIa (the kind gift of Richard Burgess) gave similar results (data not shown).

Construction of RAP74 mutants. RAP74(1-217) was constructed by PCR amplification of a plasmid clone encoding RAP74 with primers 5'-CATATGGCGGCCCTAGGCCCT-3' and 5'-CTCGAGAGACATTTCCAGGT-3' and subcloning between the NdeI and XhoI sites (cloning sites are underlined) of pET21a (Novagen). Internal deletion mutants RAP74(306-351), RAP74(276-351), and RAP74(219-351) were constructed by using a Quick Change site-directed mutagenesis kit (Stratagene). All mutants were made by using the same primer for the aa 351 position, 5'-GACATTGACAGCGAGGCCTCCTCAGCCCTCTTCATGGCG-3'. For the second oligonucleotide primers, 5'-GGAGGCCTCGCTGTCAATGTCGCTCTGCTCATCGACACCATTGGG-3', 5'-GGAGGCCTCGCTGTCAATGTCTGACATGTAGTCCACCTCTTGGCC-3', and 5'-GGAGGCCTCGCTGTCAATGTCGGAGGACATTTCCAGGTCGTCTTCAAGGTC-3', respectively, were used. The underlined sequences represent the complementary overlap between the two mutant primers.

Electrophoretic mobility shift assay. The DNA probe for the gel shift assay was the adenovirus major late promoter from positions 53 to +14 relative to the transcriptional start site. The probe was synthesized by PCR, using the primers 5'-³²P-CAGGTGTTCCTGAAGG-3' and 5'-ATGCGGAAGAGAGTGA-3'. The upstream primer was radiolabeled by using [-³²P]ATP and T4 polynucleotide kinase. After amplification, the DNA probe was gel purified by using a Qiaex kit (Qiagen). Mobility shifts were performed as described by Wang and Burton (54), with some modifications. The reaction mixtures (15 µl) contained 20 mM HEPES (pH 7.9), 20 mM Tris (pH 7.9), 50 mM KCl, 2 mM dithiothreitol, 0.5 mg of BSA (bovine serum albumin) per ml, 10% (vol/vol) glycerol, radiolabeled DNA probe, and proteins, incubated at 30°C. S. cerevisiae TBP (0.3 pmol) was combined with the DNA template (approximately 40 fmol) for 15 min. Recombinant human TFIIB (0.3 pmol) was then added, and the mixture was incubated for 15 min. Calf thymus pol II (0.15 pmol) was incubated with human recombinant TFIIF (0.1 pmol) for at least 5 min prior to addition to the DB complex. For reactions involving separate TFIIF subunits, pol II was incubated with RAP30 for 5 min and then mixed with RAP74 or a RAP74 mutant and preincubated for an additional 5 min before addition to DB and further incubation for 15 min. It appeared that prior addition of RAP30 to pol II aided assembly of RAP74 into DBPolF (defined in Results). Reaction mixtures were loaded onto a 4% polyacrylamide gel containing 0.09% bisacrylamide, 2.5% glycerol, and 0.5× TBE (tris-borate-EDTA). Dried gels were analyzed by autoradiography.

Transcription assays. Preparation of immobilized templates was adapted from published methods (1, 30). DNA containing the adenovirus major late promoter was synthesized by PCR, using an upstream 5'-biotinylated primer. The template for amplification was a pBluescript II SK() vector (Stratagene) containing the adenovirus major late promoter (258 to +196) subcloned between the XhoI and HindIII sites of the plasmid. The sequence of the upstream primer was 5'-biotin-CCCTCGAGCGGTGTTCCGCGGTCCTCCTCG-3', and the sequence of the downstream primer was 5'-CGGTGGCGGCCGCTCTAGAACTAGTGGATC-3'. The template extended from positions 263 to +251. Biotinylated DNA was incubated with streptavidin paramagnetic beads (CPG) in 2 M NaCl-1 mM EDTA-10 mM Tris-HCl (pH 7.5) for 15 min at room temperature. Immobilized templates were collected with a magnetic particle separator (CPG), washed four times, and stored at 4°C in phosphate-buffered saline (pH 7.2) containing 1 mg of BSA per ml and 0.03% NaN₃.

Pulse-spin and pulse-chase single-round assays. A 20-µl reaction mixture consisted of 2 µl of paramagnetic beads (about 0.6 µg of DNA) carrying the adenovirus major late promoter and TFIIF-depleted transcription extract (72 µg of total protein) supplemented with recombinant RAP30 and RAP74 or a RAP74 mutant (10 pmol of each), in transcription buffer (12 mM HEPES [pH 7.4], 12% glycerol, 0.12 mM EDTA, 0.12 mM EGTA, 1.2 mM dithiothreitol) containing 60 mM KCl and 12 mM MgCl₂. Preinitiation complexes were formed for 60 min at 30°C; 100 µM ATP, CTP, and GTP and 1 µM [³²P]UTP (10 µCi per reaction) were added to initiate transcription for 1 min. For the pulse-spin protocol, template-associated complexes were diluted with 200 µl of transcription buffer containing 60 mM KCl and 1 mg/ml BSA, isolated by centrifugation, and extracted from beads by boiling in 20 µl of 90% (vol/vol) formamide-1% sodium dodecyl sulfate-10 mM Tris (pH 7.9)-1% mM EDTA-0.01% bromophenol blue-0.01% xylene cyanol. For the pulse-chase protocol, instead of dilution and centrifugation of samples, 1 mM ATP, UTP, GTP, and CTP were added 1 min after addition of nucleoside triphosphates (NTPs), and elongation continued for 10 min. Reactions were stopped by addition of 200 µl of 0.1 M sodium acetate (pH 5.4)-0.5% sodium dodecyl sulfate-2 mM EDTA-100 µg of tRNA per ml, followed by phenol-chloroform extraction and ethanol precipitation. Samples were electrophoresed in a 10% polyacrylamide gel containing 1× TBE and 50% (wt/vol) urea. For quantitation of the gel, the signal in the presence of RAP30 and the absence of added RAP74 was used to estimate background (Fig. 2, lanes 21 and 22). The weak signal obtained in the absence of added RAP74 was attributed to residual RAP74 remaining in the TFIIF-depleted extract.

Elongation stimulation assay. Stimulation of pol II elongation was determined by adding TFIIF subunits to transcriptionally engaged pol II molecules that were washed free of associated elongation factors. A 20-µl reaction mixture consisted of 2 µl of paramagnetic beads carrying the adenovirus major late promoter, transcription extract derived from HeLa cell nuclei (108 µg of total protein), and transcription buffer containing 60 mM KCl and 12 mM MgCl₂. Samples were incubated for 60 min at 30°C to form preinitiation complexes. Transcription was initiated by addition of 100 µM ATP, GTP, CTP, and 1 µM (10 µCi) [-³²P]UTP (2-µl volume). After 1 min, elongation complexes were diluted in 200 µl of transcription buffer containing 500 mM KCl and 1 mg of BSA per ml and isolated by centrifugation. The purpose of this treatment was to remove accessory factors from the elongation complex. The 500 mM KCl wash appeared to be effective because elongation rate stimulation was highly dependent on addition of both RAP30 and RAP74 (Fig. 3). Complexes were diluted with transcription buffer containing 60 mM KCl and 1 mg of BSA per ml and isolated by centrifugation two times. Complexes were then resuspended in 20 µl of transcription buffer containing 60 mM KCl and 12 mM MgCl₂; recombinant RAP30 and RAP74 or a RAP74 mutant was added (10 pmol each), and the mixture was incubated for 5 min; 100 µM ATP, GTP, CTP, and UTP were added in 2 µl, and transcripts were elongated for 2 min. By using this protocol but allowing 5 min to elongate RNA chains, most of the observed pol II transcripts were extended to close to the +251 runoff position (data not shown). Transcripts were isolated by phenol extraction and ethanol precipitation and electrophoresed in a 6% polyacrylamide gel, as described above. Accurate transcription was quantitated for all of the transcripts from +122 to +251. Elongation stimulation was determined by subtracting the average value for samples containing RAP30 but no RAP74 as background (Fig. 3A, lanes 14 and 15) and expressed as a percentage of the highest signal obtained for RAP74 (lane 3).

Multiple-round and Sarkosyl block assays. Transcription was initiated from an adenovirus major late promoter template digested with SmaI to produce a 217-base runoff transcript. The source of transcription factors was TFIIF-depleted extract (72 µg of total protein) supplemented with recombinant RAP30 and RAP74 (10 pmol of each, except as noted). For all reactions, preinitiation complexes were formed for 60 min at 30°C. For the multiple-round assay, 600 µM ATP, CTP, and GTP and 25 µM [³²P]UTP (10 µCi per reaction) were added, and transcription continued for the indicated times (Fig. 5 and 6). For the 10-min cold-multiple-round assay, 600 µM ATP, CTP, and GTP and 25 µM UTP were added and incubated for 10 min, and then [-³²P]UTP (10 µCi per reaction) was added and transcription continued for 60 min (Fig. 5). For the single-round Sarkosyl block assay (18), 600 µM ATP and CTP and 25 µM [³²P]UTP (10 µCi per reaction) were added and incubated for 1 min; 600 µM GTP and 0.25% (wt/vol) Sarkosyl were added and transcription continued for 59 min (Fig. 5), 79 min (Fig. 6B and C), or 99 min (Fig. 6D). Since 0.05% Sarkosyl was previously shown to be sufficient to block new initiation by pol II (18), Sarkosyl was added in fivefold excess over the amount necessary to constrain transcription to a single round. As a control, Sarkosyl was added to reactions before NTPs, causing initiation to be completely blocked, indicating that the level of detergent added in experiments was sufficient to block reinitiation (data not shown).

G-less cassette pol II trap. To characterize multiple-round transcription in the extract system, a G-less cassette trap for pol II was used (48). The template was plasmid pML(C₂AT)19, the kind gift of Michele Sawadogo, containing the adenovirus major late promoter fused to a G-less cassette at position +11 (42, 43). Preinitiation complexes were formed for 1 h. Reactions contained TFIIF-depleted extract supplemented with 10 pmol of recombinant RAP30 and RAP74; 600 µM ATP, GTP, and CTP, 1 mM 3'-O-methyl-GTP, and 25 µM [³²P]UTP (10 µCi per reaction) were added to reactions as indicated in Fig. 7. At t = +1 min, 0.05% Sarkosyl was added to some reactions to estimate single-round transcription. In the presence of ATP, CTP, UTP, and 3'-O-methyl-GTP, a transcript of 390 bases was synthesized. Under this condition, pol II stalled after insertion of 3'-O-methyl-GMP into the RNA chain at position +390. The template was digested with PvuII to allow simultaneous detection of stalled transcription at +390 and runoff transcription at position +602. In control experiments, 0.05% Sarkosyl was found to be sufficient to constrain transcription to a single round (data not shown); 0.05% sarkosyl was used in this experiment because, unexpectedly, the early elongation complex formed from the G-less cassette template was much more sensitive to disruption with Sarkosyl than that initiated from the wild-type promoter (data not shown). The promoters in these two plasmids are identical from positions 256 to +10, and we do not know the explanation for the observed difference in Sarkosyl sensitivity. For chase reactions, 1 mM GTP and UTP were added and elongation continued for 10 or 60 min, as indicated in Fig. 7.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The N-terminal domain of RAP74 supports preinitiation complex assembly. A primary function of TFIIF in accurate initiation is to deliver pol II to the promoter; therefore, we used an electrophoresis mobility shift assay to determine which regions of RAP74 were required to bring pol II into a stable complex with adenovirus major late promoter DNA, TBP, TFIIB, and RAP30 (DBPolF complex) (Fig. 1). By itself, TBP did not efficiently induce a shift of the promoter fragment (lane 2). Upon addition of TFIIB, however, a DB complex consisting of TBP, TFIIB, and promoter DNA was observed (lane 3). When pol II was added to DB, a weak DBPol shift was seen (lane 4). RAP30 alone did not stimulate pol II binding to DB (lane 6), nor did RAP74 (data not shown). The TFIIF complex and separately added RAP30 and RAP74 subunits, however, supported assembly of DBPolF (lanes 5 and 7). RAP74(1-172) was minimally required to support assembly (lane 10). A number of RAP74 mutants that have been shown to be defective for RAP30 binding (54, 55) failed to support formation of DBPolF (lanes 11 to 15). The different mobilities of complexes containing RAP74 deletion mutants may be attributable to differences in the charge or the degree of DNA bending or flexibility in the complex. Because RAP74(1-517), RAP74(1-296), RAP74(1-205), and RAP74(1-172) are predicted to carry charges of 0, 4, +6, and +7; however, it is difficult to account for all observed mobility differences solely on the basis of charge. For instance, RAP74(1-172) is predicted to be more basic than RAP74(1-205) and yet supported a DBPolF complex with a higher mobility. Furthermore, RAP74(1-205) and (1-172) have different transcriptional activities (see below).

View larger version (101K): [in this window] [in a new window]   FIG. 1.   Electrophoretic mobility shift assay to analyze the requirement for RAP74 to form DBPolF. The probe was the adenovirus major late promoter between positions 53 and +14. D, recombinant yeast TBP (0.3 pmol); B, recombinant human TFIIB (0.3 pmol); Pol, calf thymus pol II (0.15 pmol); F, recombinant human TFIIF complex or RAP30 and RAP74 or a RAP74 mutant, added separately (0.1 pmol). DBPolF* indicates the different mobilities of complexes containing different RAP74 mutants.

The N-terminal domain of RAP74 supports accurate initiation. To determine which regions of RAP74 are most important for initiation, RAP74 mutants were tested for accurate initiation and runoff transcription from the adenovirus major late promoter (Fig. 2). RNA within the early elongation complex was visualized directly on gels in the pulse-spin protocol or extended to the runoff position in the pulse-chase protocol. Both protocols limited transcription to a single round. In the pulse-spin protocol, pol II remained close to the promoter, blocking a second round of transcription. In the pulse-chase protocol, a 1,000-fold excess of unlabeled UTP was added during the chase; so, although reinitiation could occur, it was not detected.

View larger version (48K): [in this window] [in a new window]   FIG. 2.   Regions of RAP74 required for accurate initiation. (A) Short and runoff RNAs accurately initiated from the adenovirus major late promoter. All lanes contained TFIIF-depleted extract (DE; 72 µg of total protein), recombinant human RAP30 (10 pmol), and RAP74 or a RAP74 mutant (10 pmol), preincubated with immobilized template for 60 min. Transcripts were initiated with all four NTPs and radiolabeled with [-32P]UTP (ACGU*) for 1 min. For the pulse-chase protocol, samples were chased by addition of 1 mM each NTP for 10 min (+). For the pulse-spin protocol, initiated complexes were diluted with transcription buffer and centrifuged briefly to isolate short, template-associated RNAs (). The approximate sizes of short RNAs can be estimated by comparison to 5'-phosphorylated 16- and 18-nucleotide (nt) DNA markers. In lane 1, AMPPCP (a - nonhydrolyzable ATP analog) and 2',3'-dideoxy-ATP (ddA) were substituted for ATP. In lane 2, AMPPCP was substituted for ATP. In lane 3, 1 µg of -amanitin per ml was included in the reaction. The gel band indicated with an asterisk is not a pol II transcript because it is synthesized in the presence of 1 µg of -amanitin per ml (data not shown). (B) PhosphorImager quantitation of the data shown in panel A combined with data from two other experiments, reported as average ± standard deviation. Short transcripts generated in the presence of RAP74 are expressed as 100%.

View larger version (46K): [in this window] [in a new window]   FIG. 3.   The region of RAP74 between aa 172 and 217 is critical for both accurate initiation and elongation stimulation. (A) Elongation stimulation assay. 32P-labeled short RNAs were accurately initiated from the adenovirus major late promoter immobilized on beads. Complexes were washed with 0.5 M KCl to remove associated elongation factors from pol II. RAP30 (10 pmol) and RAP74 or a RAP74 mutant (10 pmol) were added and incubated for 5 min. All four NTPs (100 µM each) were added, and RNA chains were allowed to elongate for 2 min. NE, nuclear extract. (B) PhosphorImager quantitation of the data shown in panel A for transcripts between +137 and +251 in length. (C) Comparison of elongation stimulation (B), pulse-spin, and pulse-chase data (from Fig. 2). Values for RAP74(1-296) are reported as 100% of signal.

The N-terminal domain of RAP74 stimulates elongation by pol II. Because preinitiation complex assembly (Fig. 1) and single-round initiation (Fig. 2) were supported by the N-terminal domain of RAP74, we wondered whether the same or distinct regions might stimulate elongation by pol II (Fig. 3). Consistent with previous reports (21, 49), both RAP30 and RAP74 were required to stimulate elongation (Fig. 3A, lanes 1 and 14 to 17). RAP74(1-217) stimulated elongation to about the same extent as RAP74(1-296) and RAP74. RAP74(1-205), however, stimulated elongation at a reduced level, and RAP74(1-172) was inactive or nearly so. The region between aa 1 and 205, therefore, appeared to be essential for elongation stimulation, and the region between aa 205 and 217 was strongly stimulatory. RAP74(74-517) and RAP74(87-517), from which N-terminal sequences were deleted, did not stimulate pol II elongation (reference 21 and data not shown).

View larger version (39K): [in this window] [in a new window]   FIG. 4.   A conserved region of RAP74 between amino acids 170 and 178 is important for both initiation and elongation. (A) Triple-alanine substitutions 170A3, 173A3, and 176A3, constructed in RAP74(1-217), are indicated. Related sequences from human (hRAP74), Xenopus (xRAP74), Drosophila (dF5a), and yeast (ySsu71/Tfgl) cells are shown. (B) RAP74(1-217)170A3, -173A3, and -176A3 were compared with RAP74(1-517) and RAP74(1-217) in pulse-chase initiation (top panel) and in elongation stimulation (lower panel) assays. RAP74 samples were reconstituted with RAP30 in vitro prior to assay. Pulse-chase initiation reactions contained TFIIF-depleted extract with the indicated TFIIF or TFIIF mutant (10 pmol of TFIIF complex) combined with an adenovirus major late promoter template digested with SmaI at position +217. The protocol for the initiation assay was as in Fig. 2 except that the DNA template was not immobilized and the elongation time was 30 min. Elongation stimulation reactions contained salt-washed elongation complexes supplemented with the indicated TFIIF or TFIIF mutant protein (20 pmol) (as in Fig. 3). Lane 1 contains reconstituted TFIIF and -amanitin 1 µg/ml. Lane 14 contains no added TFIIF. Lanes 2 and 3 contain 10 pmol and lanes 15 and 16 contain 20 pmol of RAP30 but no added RAP74. All other lanes are as indicated above both panels. (C) PhosphorImager quantitation of the data shown in panel B. Elongation stimulation was calculated for the +192 to +251 transcripts.

The central region and CTD of RAP74 stimulate multiple-round transcription. RAP74 mutants were next tested for the ability to support multiple-round transcription in vitro (Fig. 5). In one protocol, NTPs were added with [³²P]UTP radiolabel, and transcription was allowed to continue for 60 min. In a modified protocol, unlabeled NTPs were added for 10 min before addition of radiolabel (10-min cold-multiple round), and transcription continued for 60 min. For comparison, we performed a single-round protocol in which Sarkosyl (0.25%) was added 1 min after addition of NTPs to block new initiation (18). The final specific activity of radiolabel was the same in all three procedures, and so the intensities of transcription signals can be compared directly. For the Sarkosyl block assay (Fig. 5A), the observed transcriptional activities of RAP74 mutants were very similar to those determined in the pulse-spin initiation assay (Fig. 2).

View larger version (48K): [in this window] [in a new window]   FIG. 5.   The CTD of RAP74 stimulates multiple-round transcription. (A) Single-round and multiple-round transcription assays. All samples contained TFIIF-depleted extract (DE; 72 µg of protein), recombinant human RAP30 (10 pmol), and RAP74 or a RAP74 mutant (10 pmol). In the single-round protocol (s), reinitiation was blocked by addition of the anionic detergent sarkosyl. In the multiple-round protocol (m), transcription was allowed to proceed for 60 min. In the 10-min cold-multiple-round protocol, (c), transcription with all four NTPs was allowed to proceed for 10 min before addition of radiolabel and incubation for 60 min. Reactions labeled A contained RAP74 and 1 µg of -amanitin per ml. (B) PhosphorImager quantitation of the data in panel A.

View larger version (32K): [in this window] [in a new window]   FIG. 6.   The central region and the CTD of RAP74 cooperate to stimulate multiple-round transcription. (A) New initiation occurred throughout the 90-min course of the reaction. Assays were done by the multiple-round protocol (Fig. 5) and stopped at the indicated times (filled squares), or instead of stopping the reactions, sarkosyl was added to block new initiation and transcription continued for an additional 30 min to complete any previously initiated chains (open squares). (B and C) Both the central region and the CTD of RAP74 contributed to multiple-round transcription. Multiple-round transcription was determined by the protocol used for Fig. 5 except that reactions were stopped at the indicated times. Single-round transcription was estimated by using a Sarkosyl block procedure with a 79-min elongation. Cycles of transcription were estimated as transcription in the absence (time indicated) or presence (79-min elongation) of Sarkosyl. (D and E) TFIIF containing RAP74(1-217) (abbreviated F217) and F172 competed with TFIIF (F) for transcription complex formation and inhibited multiple-round transcription. The reaction protocol was the same as used for panels B and C except that the reaction time after NTP addition was 100 min. TFIIF or mutants were added to the reaction at 60 min or +10 min, as indicated. (D) Reactions in columns 1 to 6 contained 2 pmol of TFIIF; reactions in columns 2 and 5 also contained 2, 4, 10, or 20 pmol F217 (data points were combined within the error bar because the effect was essentially maximal with 2 pmol of F217); reactions in columns 3 and 6 contained 2 or 20 pmol of F172. (E) The reaction in column 1 contained 2 pmol of TFIIF; reactions in columns 2 to 8 contained 2 pmol of F217; reactions in columns 3, 4, and 5 contained 2, 4, or 10 pmol of TFIIF added at +10 min; reactions in columns 6, 7, and 8 contained 10 pmol of RAP30, RAP74, or F217 added at +10 min.

New initiation resulted from previously unused pol II molecules initiating from previously unused promoters. In the extract system, new initiation events could involve reuse of promoters, reuse of pol II, or initiation from many promoters from which pol II is enabled to initiate at various times. We used a G-less cassette template as a pol II trap to discriminate between these possibilities (48). With the G-less cassette, an adenovirus major late promoter transcript can be synthesized with only ATP, CTP, and UTP, omitting GTP (42). Pol II stalls at the first position at which GMP must be incorporated. If promoters are reused, multiple pol II molecules pile up at the end of the G-less cassette, resulting in a ladder of transcripts each one shorter by about 30 nucleotides (48). The number of rungs on the ladder corresponds to the number of pol II molecules that have stalled on the template. If pol II must be reused for new initiation at later reaction times, trapping pol II at the end of the G-less cassette is expected to inhibit new initiation.

View larger version (50K): [in this window] [in a new window]   FIG. 7.   Multiple-round transcription in an extract system can be described by a kinetic limitation model. The template contains the adenovirus major late promoter (AdMLP) fused to a G-less cassette that extends to position +389. The template was digested with PvuII at position +602. Transcripts formed in the presence of the chain terminator 3'-O-methyl-GTP (mG) and in the absence of GTP stalled at the end of the G-less cassette. When GTP was included in the reaction, pol II continued transcription to the +602 runoff position. The template and protocols are summarized at the top. Expected results for the promoter limitation, pol II limitation, and kinetic limitation models are shown on the left and discussed in the text. Experimental data are shown on the right.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

From primary sequence (2, 15), RAP74 is proposed to have highly basic N- and C-terminal domains separated by a highly charged, overall acidic, and flexible central region that is rich in charged amino acids, E, D, K, and R, and also S, T, P, and G. In this report, we show that these primary sequence features correspond to distinct N- and C-terminal functional domains (Fig. 8). The N-terminal domain is sufficient to support preinitiation complex assembly, single-round initiation, and elongation by pol II. The central region and CTD of RAP74 have a small stimulatory effect on initiation, but their most important function is in multiple-round transcription. Deletion of just the central region or just the CTD of RAP74 creates a protein with partial function in multiple-round transcription.

View larger version (33K): [in this window] [in a new window]   FIG. 8.   N- and C-terminal domains of RAP74, which were originally proposed from sequence analysis (2, 15), correspond to distinct functional domains. The N-terminal domain has most of the functions required for single-round initiation and elongation. The central region and CTD of RAP74 function in multiple-round transcription. Regions shaded dark are very important for activity. IIB, TFIIB; CTDP, CTD phosphatase; PIC, preinitiation complex.

The N-terminal domain of RAP74 stimulates initiation and elongation. The RAP74 N-terminal domain extending from aa 1 to 217 supports most RAP74 functions for preinitiation complex assembly (Fig. 1), accurate initiation (Fig. 2), and elongation stimulation (Fig. 3 and 4). In this work, we have mapped a critical region for these functions between aa 136 and 217. RAP74(1-172) binds tightly to the RAP30 subunit, but RAP74(1-136) does not (54, 55). Consistent with its capacity to bind RAP30, RAP74(1-172) is minimally sufficient to support pol II assembly into DBPolF (Fig. 1). Fulfilling this role in assembly, however, is not sufficient to support TFIIF function in initiation or elongation, which minimally requires RAP74(1-205) (Fig. 2). RAP74(1-217) is significantly more active than RAP74(1-205) in both accurate initiation and elongation stimulation (Fig. 2 and 3).

The central region and CTD of RAP74 stimulate multiple-round transcription. When cycles of transcription were plotted against time for multiple-round transcription reactions, the resulting curve showed a high initial rate of RNA synthesis followed by a lower rate (Fig. 6B and C). The initial burst of synthesis persisted for 10 to 15 min, and then the lower rate dominated the reaction. This lower rate represents multiple-round transcription and is likely to represent a recycling mechanism for pol II or an essential pol II transcription factor. RAP74 mutants from which C-terminal and/or central regions were removed were found to be defective for this low rate of transcription at later reaction times.