Functions of the N- and C-terminal domains of human RAP74 in transcriptional initiation, elongation, and recycling of RNA polymerase II

Abstract

Transcription factor IIF (TFIIF) cooperates with RNA polymerase II (pol II) during multiple stages of the transcription cycle including preinitiation complex assembly, initiation, elongation, and possibly termination and recycling. Human TFIIF appears to be an alpha2beta2 heterotetramer of RNA polymerase II-associating protein 74- and 30-kDa subunits (RAP74 and RAP30). From inspection of its 517-amino-acid (aa) sequence, the RAP74 subunit appears to comprise separate N- and C-terminal domains connected by a flexible loop. In this study, we present functional data that strongly support this model for RAP74 architecture and further show that the N- and C-terminal domains and the central loop of RAP74 have distinct roles during separate phases of the transcription cycle. The N-terminal domain of RAP74 (minimally aa 1 to 172) is sufficient to deliver pol II into a complex formed on the adenovirus major late promoter with the TATA-binding protein, TFIIB, and RAP30. A more complete N-terminal domain fragment (aa 1 to 217) strongly stimulates both accurate initiation and elongation by pol II. The region of RAP74 between aa 172 and 205 and a subregion between aa 170 and 178 are critical for both accurate initiation and elongation, and mutations in these regions have similar effects on initiation and elongation. Based on these observations, RAP74 appears to have similar functions in initiation and elongation. The central region and the C-terminal domain of RAP74 do not contribute strongly to single-round accurate initiation or elongation stimulation but do stimulate multiple-round transcription in an extract system.

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.

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 [α-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 β-γ 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 α-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 α-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%.

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.

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 α-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.

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 αA contained RAP74 and 1 μg of α-amanitin per ml. (B) PhosphorImager quantitation of the data in panel A.

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.

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.

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.