TFIIF facilitates dissociation of RNA polymerase II from noncoding RNAs that lack a repression domain

Abstract

Noncoding RNAs (ncRNAs) have recently been found to regulate multiple steps in mammalian mRNA transcription. Mouse B2 RNA and human Alu RNA bind RNA polymerase II (Pol II) and repress mRNA transcription, using regions of the ncRNAs referred to as repression domains. Two other ncRNAs, mouse B1 RNA and human small cytoplasmic Alu (scAlu) RNA, bind Pol II with high affinity but lack repression domains and hence do not inhibit transcription. To better understand the interplay between ncRNAs that bind Pol II and their functions in transcription, we studied how Pol II binding and transcriptional repression are controlled by general transcription factors. We found that TFIIF associates with B1 RNA/Pol II and scAlu RNA/Pol II complexes and decreases their kinetic stability. Both subunits of TFIIF are required for this activity. Importantly, fusing a repression domain to B1 RNA stabilizes its interaction with Pol II in the presence of TFIIF. These results suggest a new role for TFIIF in regulating the interaction of ncRNAs with Pol II; specifically, it destabilizes interactions with ncRNAs that are not transcriptional repressors. These studies also identify a new function for ncRNA repression domains: they stabilize interactions of ncRNAs with Pol II in the presence of TFIIF.

FIG. 1.

B1 RNA and B2 RNA likely bind overlapping sites on Pol II. (A) B1 RNA and B2 RNA bind Pol II competitively. Purified human Pol II (2 nM) was incubated with ³²P-labeled B2 RNA (5 nM) and unlabeled B1 RNA (at the molar ratios indicated), and complexes were resolved by EMSA. The relative fraction of B2 RNA bound in the absence of unlabeled B1 RNA was set to 1 (lane 1). Two experiments were performed, and representative data are shown. (B) B1 RNA prebound to Pol II blocks the association of B2 RNA. ³²P-labeled B2 RNA and unlabeled B1 RNA were added to Pol II in the order indicated. RNAs added first were given 10 min to bind Pol II prior to the addition of the second RNA, after which the incubation was continued for an additional 15 min. Complexes were resolved by EMSA. Four experiments were performed, and representative data are shown. (C) B1 RNA/Pol II complexes are kinetically stable. ³²P-labeled B1 RNA/Pol II complexes were challenged with a 100-fold molar excess of unlabeled B1 RNA, and EMSA was used to measure the amount of ³²P-labeled B1 RNA/Pol II complex remaining over time. Two experiments were performed, and representative data are shown. (D) Shown are the data from C quantified and plotted as the fraction of B1 RNA bound to Pol II versus time and fit to a first-order exponential decay equation. The rate constant for the dissociation of the B1 RNA/Pol II complex is 1.5 × 10⁻⁴ ± 0.3 × 10⁻⁴ s⁻¹, averaged from two independent experiments, with the error representing the range of the measurements.

FIG. 2.

B1 RNA cannot block transcriptional repression by B2 RNA in vitro. (A) B2 RNA represses transcription when added to reaction mixtures after B1 RNA is prebound to Pol II. ncRNAs (5 nM) were added to reaction mixtures as indicated. The 390-nt G-less transcript is shown. Two experiments were performed, and representative data are shown. NTPs, nucleotide triphosphates. (B) Of the proteins present in the minimal transcription system, only Pol II binds B1 RNA. ³²P-labeled B1 RNA (0.5 nM) and ³²P-labeled B1 RNA/Pol II (2 nM) complexes were incubated with human TBP, TFIIF, or TFIIB, as indicated, at the concentrations used for in vitro transcription. Bound and free RNAs were resolved by EMSA. Three experiments were performed, and representative data are shown. GTFs, general transcription factors.

FIG. 3.

TFIIF destabilizes the B1 RNA/Pol II complex. (A) Dissociation of the B1 RNA/Pol II complex was monitored over 45 min in the presence of TBP, TFIIB, TFIIF, or all three factors. Assays were performed as described in the legend of Fig. 1C. Three experiments were performed, and representative data are shown. (B) The rate constant for the dissociation of the B1 RNA/Pol II/TFIIF complex is 1.9 × 10⁻³ ± 1.0 × 10⁻³ s⁻¹, averaged from three independent experiments, with the errors representing the standard deviations. Shown are representative data in which the fractions of B1 RNA bound to Pol II/TFIIF were quantified and fit to a single exponential decay equation. (C) TFIIF has little effect on the kinetic stability of the B2 RNA/Pol II complex. Shown are representative data in which the fractions of B2 RNA bound to Pol II (−TFIIF) or Pol II/TFIIF (+TFIIF) were quantified and fit to a single exponential decay equation. The rate constants for dissociation are 2.9 × 10⁻⁵ ± 0.1 × 10⁻⁵ s⁻¹ in the absence of TFIIF and 4.5 × 10⁻⁵ ± 1.9 × 10⁻⁵ s⁻¹ in the presence of TFIIF; each is the average of data from two experiments, with the error representing the range in the measurements. (D) B1 RNA inhibits transcription on a bubble template in the absence of TFIIF. ncRNAs (5 nM) were added to reaction mixtures containing Pol II in the presence and absence of TFIIF as indicated; the 40-nt transcript is shown. Two experiments were performed, and representative data are shown.

FIG. 4.

Dissociation of B1 RNA/Pol II by TFIIF requires both subunits of TFIIF. (A) Both RAP30 and RAP74 are required to facilitate the dissociation of B1 RNA from Pol II. Dissociation assays were performed as described in the legend to Fig. 1C. Two experiments were performed, and representative data are shown. (B) Plot of the EMSA data shown in panel A.

FIG. 5.

TFIIF facilitates the dissociation of scAlu RNA from Pol II. (A) TFIIF supershifts and destabilizes the scAlu RNA/Pol II complex. Dissociation assays were performed as described in the legend to Fig. 1C. Two experiments were performed, and representative data are shown. (B) The relative fraction of ncRNA bound was quantified, averaged between two experiments, and plotted against time. The error bars represent the ranges in the measurements.

FIG. 6.

The addition of a repression domain from Alu RNA onto B1 RNA stabilizes its interaction with Pol II in the presence of TFIIF. Dissociation assays with Pol II or Pol II/TFIIF were performed as described in the legend to Fig. 1C using the following four ncRNAs: B1 RNA, B1-A RNA, B1-L RNA, and B1 RNA plus the isolated A region. The relative fraction of ncRNA bound was quantified, averaged between two experiments, and plotted against time. The error bars represent the ranges in the measurements.