Recognition of a human arrest site is conserved between RNA polymerase II and prokaryotic RNA polymerases

Abstract

DNA sequences that arrest transcription by either eukaryotic RNA polymerase II or Escherichia coli RNA polymerase have been identified previously. Elongation factors SII and GreB are RNA polymerase-binding proteins that enable readthrough of arrest sites by these enzymes, respectively. This functional similarity has led to general models of elongation applicable to both eukaryotic and prokaryotic enzymes. Here we have transcribed with phage and bacterial RNA polymerases, a human DNA sequence previously defined as an arrest site for RNA polymerase II. The phage and bacterial enzymes both respond efficiently to the arrest signal in vitro at limiting levels of nucleoside triphosphates. The E. coli polymerase remains in a template-engaged complex for many hours, can be isolated, and is potentially active. The enzyme displays a relatively slow first-order loss of elongation competence as it dwells at the arrest site. Bacterial RNA polymerase arrested at the human site is reactivated by GreB in the same way that RNA polymerase II arrested at this site is stimulated by SII. Very efficient readthrough can be achieved by phage, bacterial, and eukaryotic RNA polymerases in the absence of elongation factors if 5-Br-UTP is substituted for UTP. These findings provide additional and direct evidence for functional similarity between prokaryotic and eukaryotic transcription elongation and readthrough mechanisms.

Fig. 1. Templates used for in vitro transcription

A, pAdTerm-2 contains the core adenovirus major late promoter (−50 to +10) and a 285-base pair segment of the human histone H3.3 gene. The sequence around arrest site Ia (overlined) and a related site (site II; Ref. 2) is shown. Numbering refers to base positions in the transcript relative to the transcription start (+1). The phosphodiester bond cleaved by RNA polymerase II in arrested Ia complexes is indicated (downward arrowhead). Approximate sizes of the expected transcripts are shown. Bold T represents the positions of the RNA 3′ termini determined previously (37). B, prrn-Ia contains DNA sequence from the P₂ promoter of the E. coli rrnB operon inserted upstream of the human H3.3 sequences in pGEMTerm (see below). Approximate lengths of transcripts initiating from P₂ are indicated on the right. C, pGEMTerm contains the human H3.3 sequences inserted into pGEM2 downstream of an SP6 RNA polymerase promoter. Approximate transcript lengths are indicated on the right.

Fig. 2. In vitro transcription by RNA polymerase II on pAdTerm-2 cut with NdeI (A and B) or PstI (C)

A, time course of elongation at varying CTP concentrations. Transcription was carried out in the presence of ATP, UTP, GTP, and the indicated concentrations of CTP. Samples were stopped at the indicated times and separated by gel electrophoresis. Run-off (RO) and Ia (Ia) RNAs are indicated with arrows. Lane M, reference RNAs of 260, 380, 420, and 540 bases. B, CTP “chase” of arrested complexes. Arrested complexes were assembled by incubation for 1 h at 28 °C with 380 μM (lanes 1–3) or 6 μM (lanes 4–6) CTP and 800 μM each ATP, UTP, and GTP. Samples were stopped (lanes 1 and 4) or incubated for 15 min with an additional 380 μM CTP without (lanes 2 and 5) or with recombinant human SII (0.5 μg; lanes 3 and 6). Lane M, reference RNAs of 260, 380, 420, and 540 bases. C, readthrough in the presence of 5-Br-UTP. Initiated complexes bearing a 14-base RNA were assembled and prepared for electrophoresis (lane 0) or were incubated for the indicated times with 700 μM each ATP, GTP, CTP, and either 700 μM 5-Br-UTP or UTP.

Fig. 3. Effect of CTP concentration on elongation by E. coli RNA polymerase

A, pulse-labeled elongation complexes were assembled on PstI-cut prrn-Ia, washed by immunoprecipitation (lane 1), and incubated for 10 min at 30 °C with ATP, GTP, and UTP (100 μM each), MgCl₂ (7 mM), heparin (50 μg/ml), and the indicated amounts of unlabeled CTP. Samples were challenged with water (lanes 2–6) or an additional 1 mM CTP (lanes 7–11) and incubated for 10 min at 30 °C. A long transcript derived from a promoter located elsewhere on the plas-mid is indicated (p). B, the radioactivity in the run-off and Ia RNAs from lanes 2–6 and 8–11 of panel A were quantitated and percent arrest was calculated (= 100 × (phosphorimager units for Ia RNA divided by the sum of the units in Ia RNA and run-off RNA)).

Fig. 4. Time course of elongation by E. coli RNA polymerase at 1μ

M CTP. Arrested elongation complexes were assembled with 1 μM CTP (lane 1) as shown in Fig. 3, lane 3, and incubated for the indicated additional times at 30 °C. After 1 (lane 5) or 2 h (lane 7) a portion of the sample was adjusted to 1 mM each ATP, UTP, GTP, and CTP, and incubation continued for 10 min. RNA identities including pause site II are indicated on the left.

Fig. 5. Effect of GreB on arrested elongation complexes

Initiated complexes were assembled with E. coli RNA polymerase as described under “Experimental Procedures.” Arrested elongation complexes were formed by incubation with 100 μM CTP and 800 μM each ATP, GTP, and UTP and washed by immunoprecipitation. Complexes were incubated with 7 mM MgCl₂, and either buffer (lane 1) or 2 μg GreB (lanes 2–6). Sarkosyl (0.25%; lane 3), heparin (10 μg/ml; lane 4), 1 mM each of all four nucleotides (lane 5), or 1 mM each ATP, UTP, GTP, and 3′-deoxy-CTP (lane 6) were also added, and incubation was continued for 5 min at 30 °C. The major GreB cleavage product is indicated (*). Lane M, reference RNAs of 260, 380, 420, and 540 bases

Fig. 6. Rate of arrest by E. coli RNA polymerase at site Ia

A, arrested complexes were assembled (lane 1) as shown in Fig. 3, lane 3, washed by immunoprecipitation, and treated with 7 mM MgCl₂ and GreB (1.3 μg) for 5 min at 30 °C (lane 2). These cleaved complexes were washed again and incubated with 0.5 M NaCl, 7 mM MgCl₂ and 1 mM each ATP, UTP, and GTP (lane 4) or all four NTPs (lane 3) for 10 min at 30 °C. An aliquot of washed, cleaved complexes were adjusted to 0.5 M NaCl, 7 mM MgCl₂ and 1 mM each ATP, UTP, and GTP. At the indicated times (lanes 5–10), samples were made of 1 mM CTP at 30 °C and incubated for 10 min. Run-off (RO), transcripts Ia and II, and the major cleavage product (*) are indicated. B, comparative first-order plot showing loss of factor-independent elongation ability for E. coli RNA polymerase and rat liver RNA polymerase II at site Ia at the indicated temperatures. The natural logarithm of percent readthrough (=100 × (phosphorimager units in run-off RNA divided by sum of units in run-off RNA and Ia RNA)) was plotted versus time before CTP was added. The average of three independent trials (including the experiment shown in Fig. 6A) were plotted for E. coli RNA polymerase. Standard deviation was calculated for each time point and is represented by error bars. Points for which no error bars are visible had standard deviations that were too small to permit error bars to be drawn. Measurements made in similar experiments on RNA polymerase II at 15 or 28 °C were replotted from Ref. without error bars.

Fig. 7. Readthrough of site Ia by E. coli RNA polymerase in the presence of 5-Br-UTP

Pulse-labeled elongation complexes were assembled on PstI cut prrn-Ia, washed by immunoprecipitation (0 min), and provided with 7 mM MgCl₂, 50 μg/ml heparin, 1 μM CTP, 120 μM each ATP and GTP and either 120 μM UTP or 5-Br-UTP as shown, and incubated for the indicated times at 30 °C.

Fig. 8. Transcription through site Ia by phage SP6 RNA polymerase

A, pGEMTerm DNA cut with BstEII (lane 1), PstI (lane 2), or NsiI (lanes 3–5) were transcribed in vitro with SP6 RNA polymerase in the presence of 2.4 mM each ATP, UTP, and GTP, and the indicated concentration of CTP for 2 h (lanes 1–5) or 30 min (lanes 6–9). In lane 9, UTP was replaced with 2 mM 5-Br-UTP. After incubation, samples were made of 100 μg/ml in heparin (lanes 7 and 8) and 1 mM in CTP (lane 8) and incubated for an additional 15 min. The addition of heparin results in a new slower migrating band, the identity of which is unknown. Lane M, reference RNAs of 260, 380, 420, and 540 bases. B, reactions containing 2.4 mM each ATP and GTP, 10 μM CTP, and either 1 mM (lane 1), 750 μM (lane 2), 500 μM (lane 3), 250 μM (lane 4), or no UTP (lane 5) were incubated for 30 min at 37 °C. The indicated concentrations of 5-Br-UTP were added to each to bring the total uridine nucleotide concentration to 1 mM. Percent readthrough (100 × (phosphorimager units in the run-off RNA)/(units in run-off RNA + transcript Ia + transcript II)), was plotted versus Br-UTP concentration.