Direct inhibition of RNA polymerase II transcription by RECQL5

Abstract

DNA helicases of the RECQ family are important for maintaining genome integrity, from bacteria to humans. Although progress has been made in understanding the biochemical role of some human RECQ helicases, that of RECQL5 remains elusive. We recently reported that RECQL5 interacts with RNA polymerase II (RNAPII), pointing to a role for the protein in transcription. Here, we show that RECQL5 inhibits both initiation and elongation in transcription assays reconstituted with highly purified general transcription factors and RNAPII. Such inhibition is not observed with the related, much more active RECQL1 helicase or with a version of RECQL5 that has normal helicase activity but is impaired in its ability to interact with RNAPII. Indeed, RECQL5 helicase activity is not required for inhibition. We discuss our findings in light of the fact that RECQ5(-/-) mice have elevated levels of DNA recombination and a higher incidence of cancer.

FIGURE 1.

A central 9-amino acid sequence is essential for RNAPII-RECQL5 interaction in vivo. A, diagram describing the constructs used. B and C, Western blots of FLAG-tagged immunoprecipitations (IP) showing interactions between different forms of FLAG-tagged RECQL5 and RNAPII, using FLAG and RPB1 antibody, respectively.

FIGURE 2.

Activity of the recombinant RECQ proteins used in this study. A, Coomassie-stained SDS-PAGE gel showing recombinant RECQ proteins. M, protein molecular weight standard marker. B, DNA helicase assays showing the relative ATP-dependent helicase activities of the recombinant proteins shown in A.

FIGURE 3.

RECQL5 inhibits reconstituted RNAPII transcription. A, outline of the experiment. PIC, pre-initiation complex. B, 0.113, 0.225, and 0.9 pmol of RECQL5 (lanes 2–4) or RECQL5_ID (lanes 5–7) were added during preinitiation complex formation. Transcripts were resolved by 6% denaturing PAGE and visualized by phosphorimaging. The arrow indicates a 254-nt run-off transcript. C, effect of RECQL5_D157A and RECQL1 as described in B. The experiments in B and C were performed at different times and thus cannot be used for precise quantitative comparisons of the relative effects of RECQL5 and RECQL5D_157A. Please note that the experiments in Figs. 5 and 6 show that RECQL5_D157A does not inhibit transcription more potently than wild type RECQL5.

FIGURE 4.

RECQL5 inhibits an early step of RNAPII transcription. A, An outline of the experiment is shown. RECQL5 was either added during or after PIC assembly. B, Sarkosyl-treated RNAPII transcription reactions were treated with 0.9 pmol of the indicated RECQL5 protein, resolved by 6% denaturing PAGE, and visualized by phosphorimaging. The arrow indicates a 254-nt run-off transcript. Quantification of this transcript, relative to the control (lane 2, set to 1), is indicated below the lanes.

FIGURE 5.

RECQL5 inhibits first phosphodiester bond formation by RNAPII. A, outline of the abortive initiation experiment. B, abortive transcription producing a trinucleotide transcript (indicated by an arrow) resolved by 25% denaturing PAGE. 0.9 pmol of the indicated RECQ protein was added after preinitiation complex formation, together with the nucleotides. The arrow indicates a 3-nt transcript. Quantification of the trinucleotide transcript, relative to the control (lane 2, set to 1), is indicated below the lanes.

FIGURE 6.

RECQL5 inhibits RNAPII elongation. A, an outline of the RNAPII elongation experiment is shown. B, 135-nt RNAPII elongation complexes, stalled by UTP omission, were incubated with all NTPs and the indicated RECQ protein (0.9 pmol) for 2, 5, and 15 min, respectively. The resulting transcripts were resolved by 6% denaturing PAGE and analyzed by phosphorimaging. The arrow indicates the strongest, long transcript obtained. Quantification of this transcript, relative to the control (lane 4, set to 1), is indicated below the lanes. C, as in B, except that ELL/EAF was added, or not added, as indicated, and incubation with all NTPs was for 30 min.