Transcriptional activators enhance polyadenylation of mRNA precursors

Abstract

Polyadenylation of mRNA precursors is frequently coupled to transcription by RNA polymerase II. Although this coupling is known to involve interactions with the C-terminal domain of the RNA polymerase II largest subunit, the possible role of other factors is not known. Here we show that a prototypical transcriptional activator, GAL4-VP16, stimulates transcription-coupled polyadenylation in vitro. In the absence of GAL4-VP16, specifically initiated transcripts accumulated but little polyadenylation was observed, while in its presence polyadenylation was strongly enhanced. We further show that this stimulation requires the transcription elongation-associated PAF complex (PAF1c), as PAF1c depletion blocked GAL4-VP16-stimulated polyadenylation. Furthermore, knockdown of PAF subunits by siRNA resulted in decreased 3' cleavage, and nuclear export, of mRNA in vivo. Finally, we show that GAL4-VP16 interacts directly with PAF1c and recruits it to DNA templates. Our results indicate that a transcription activator can stimulate transcription-coupled 3' processing and does so via interaction with PAF1c.

Figure 1. GAL4-VP16 activates transcription-coupled polyadenylation

(A) Purification of bacterially expressed GAL4-VP16. His-tagged GAL4-VP16 was purified by using talon resin (clontech), and 5 μg resolved by SDS-PAGE. (B) Schematic of the DNA template used for transcription-coupled polyadenylation assays. The template contained tandem repeats of GAL4 binding sites upstream of the adenovirus E4 core promoter region and SVL poly(A) site downstream. The position of the RNA probe used to analyze cleavage levels is also indicated. (C) Transcription-polyadenylation assay with or without GAL4-VP16. After reactions in HeLa NE, RNAs were purified, separated by oligo(dT) selection into poly(A)- and poly(A)+ fraction, and analyzed on 5% denaturing gel. Run-off and polyadenylated products were quantitated with ImageJ and results are shown at the bottom of each lane. (D) RNase protection assay to examine cleavage level. After transcription-polyadenylation was carried out as in (C) without radioactive α-³²P UTP, RNAs were isolated, treated with turbo-DNase (ambion) and subject to RNase protection analysis. Quantitation of cleaved and uncleaved products were done with ImageJ and the results are shown at the bottom of each lane.

Figure 2. GAL4-p53 stimulates transcription-coupled polyadenylation

(A) Purification of bacterially expressed GAL4-p53. His-tagged GAL4-p53 was purified by using talon resin and 5 μg resolved by SDS-PAGE. (B) Transcription-polyadenylation assay with or without GAL4-p53. Transcripts produced in NE were analyzed on 5% denaturing gel and quantitated as in Fig. 1C.

Figure 3. PAF1c depletion diminished VP16-dependent polyadenylation but not transcription

(A) Immunodepletion of PAF complex from NE. Depletion was performed with anti-Cdc73 antibody and was confirmed by Western blot. (B) Transcription-polyadenylation assay with PAF depleted NE in the presence of GAL4-VP16. RNAs were analyzed by RNase protection assays and quantitated with ImageJ as in Fig. 1D. (C) PAF depletion by siRNA knockdown. NE was prepared from 293T cells treated with siRNA targeting Cdc73 and Ctr9. Depletion was confirmed by Western blot. siRNA targeting GFP served as a control. (D) Transcription-polyadenylation assay with PAF depleted NE prepared in (C), and RNAs were analyzed and quantitated as in Fig. 1C. (E) Transcription-polyadenylation was done as in (D) except that α-³²P UTP was omitted from reaction mixtures. RNAs were analyzed by RNase protection assays as in (B). Quantitation was done with ImageJ.

Figure 4. siRNA knockdown of PAF1c inhibits 3′end processing of reporter mRNA

(A) Schematic of reporter plasmid pGL3G5E4 containing GAL4 binding sites upstream of E4 core promoter, Luciferase coding sequences and SVL poly(A) site downstream. (B) RNase protection assay of transcripts isolated from 293T cells treated with siRNA targeting PAF1c subunits. The reporter plasmid was transfected into 293T cells pre-treated with siRNAs targeting PAF1c subunit(s) as indicated. The next day, cells were harvested and total RNAs were obtained using Trizol (Invitrogen), fractionated into nuclear and cytoplasmic fractions and 3′ cleavage efficiency was examined by RNase protection assays. Quantitation was performed with ImageJ. Ethidium bromide staining of 5S rRNA (lower panel) indicates uniform recovery between samples.

Figure 5. VP16 interacts with PAF1c

(A) GST-pulldown assay of PAF1c subunits (Cdc73 and Paf1) from NE. The assay was performed using 5 μg immobilized GST-VP16 together with NE. After extensive washing, bound proteins were eluted by boiling and analyzed by western blotting. For comparison, 2% of input NE is shown in lane 1. (B) Far-western analysis of purified PAF1c. PAF1c purified from 293T cells stably expressing Flag-tagged paf1 was resolved on SDS-PAGE and transferred to a nitrocellulose membrane. After denaturating-renaturation using guanidine hydrochloride, the membrane was probed with either GST or GST-VP16 and subject to western blotting with anti-GST antibodies. (C) GST-pulldown of bacterially expressed His-Paf1. The assay was carried out using GST-VP16 and purified His-tagged Paf1. After extensive washing, bound His-Paf1 were eluted and analyzed by western blotting. 10% of input His-Paf1 is also shown in lane 1.

Figure 6. GAL4-VP16 recruits PAF1c to DNA templates

(A) Immobilized template assay to analyze protein recruitment. Biotinylated DNA templates were immobilized using streptoavidin magnetic beads (Invitrogen) and incubated with NE alone or supplemented with GAL4-VP16 in the presence or absence of NTPs as indicated under transcription-coupled polyadenylation conditions. After extensive washing, proteins bound to the templates were eluted and analyzed by western blotting with the indicated antibodies. Protein bands were detected and quantitated with LI-COR Odyssey. Error bars represent standard deviations from three independent experiments. (B) Immobilized template assay with mutant templates lacking GAL4 binding sites. Assay were done and analyzed as in (A).

Figure 7. Model for GAL4-VP16 induced activation of transcription-coupled polyadenylation

In the absence of GAL4-VP16 (top), certain polyadenylation factors are recruited to the template, such as CPSF by interaction with TFIID. However, under these conditions the factors are not properly integrated into elongation complexes, and subsequent polyadenylation is inefficient. In the presence of GAL4-VP-16 (bottom), PAF1c is recruited to the promoter region, by direct interaction with GAL4-VP16. PAF1c then facilitates proper coordination of poly(A) factors with elongation complexes, which enhances the efficiency of 3′ end formation. Symplekin may play an especially important role in this process (see text). After poly(A) tail formation, PABPC1 is recruited to the poly(A) tail by PAF1c to facilitate mRNA export to the cytoplasm.