Function of the Bacillus subtilis transcription elongation factor NusG in hairpin-dependent RNA polymerase pausing in the trp leader

Abstract

NusA and NusG are transcription elongation factors that bind to RNA polymerase (RNAP) after sigma subunit release. Escherichia coli NusA (NusA(Ec)) stimulates intrinsic termination and RNAP(Ec) pausing, whereas NusG(Ec) promotes Rho-dependent termination and pause escape. Both Nus factors also participate in the formation of RNAP(Ec) antitermination complexes. We showed that Bacillus subtilis NusA (NusA(Bs)) stimulates intrinsic termination and RNAP(Bs) pausing at U107 and U144 in the trpEDCFBA operon leader. Pausing at U107 and U144 participates in the transcription attenuation and translational control mechanisms, respectively, presumably by providing additional time for trp RNA-binding attenuation protein (TRAP) to bind to the nascent trp leader transcript. Here, we show that NusG(Bs) causes modest pause stimulation at U107 and dramatic pause stimulation at U144. NusA(Bs) and NusG(Bs) act synergistically to increase the U107 and U144 pause half-lives. NusG(Bs)-stimulated pausing at U144 requires RNAP(Bs), whereas NusA(Bs) stimulates pausing of RNAP(Bs) and RNAP(Ec) at the U144 and E. coli his pause sites. Although NusG(Ec) does not stimulate pausing at U144, it competes with NusG(Bs)-stimulated pausing, suggesting that both proteins bind to the same surface of RNAP(Bs). Inactivation of nusG results in the loss of RNAP pausing at U144 in vivo and elevated trp operon expression, whereas plasmid-encoded NusG complements the mutant defects. Overexpression of nusG reduces trp operon expression to a larger extent than overexpression of nusA.

Fig. 1.

Models of B. subtilis trp operon regulation. (Upper) Transcription attenuation model. During transcription RNAP pauses at U107. Under limiting tryptophan conditions, TRAP does not bind to the RNA. Once RNAP resumes transcription, the AT forms, resulting in transcription readthrough. Under excess tryptophan conditions, TRAP binds to the 5′SL and the (G/U)AG repeats. Bound TRAP releases paused RNAP and prevents AT formation. Thus, formation of the T causes transcription to terminate at G140 or U141. Because termination is never 100% efficient, a fraction of RNAP molecules will not terminate despite the presence of bound TRAP. (Lower) trpE translational control model. During transcription of readthrough transcripts, RNAP pauses at U144. Under limiting tryptophan conditions, TRAP does not bind to the RNA. Once RNAP resumes transcription, the RNA adopts a structure such that the trpE SD sequence is available for ribosome binding. Under excess tryptophan conditions, TRAP binds to the paused transcript. Once RNAP resumes transcription, the trpE SD-sequestering hairpin forms and inhibits translation of trpE. The same structure functions as the terminator and U144 pause hairpins. The 5′SL is shown only in the Upper Left drawing.

Fig. 2.

NusA_Bs- and NusG_Bs-stimulated pausing at the trp U107, trp U144, and his pause sites, and competition of NusG_Bs with NusG_Ec at the trp U144 pause site. Single-round in vitro transcription reactions were performed with the indicated Nus factors and stopped at the times shown above each lane. Chase reactions (Ch) were extended for an additional 10 min at 37°C in the presence of each NTP (500 μM). Pause half-lives (T_1/2) from these representative gels are shown. (A) Transcription of the trp leader by RNAP_Bs was performed at 23°C in the presence of 10 μM ATP (limiting) and 150 μM concentrations of the other three NTPs. Positions of the U107 paused, U144 paused, and runoff (RO) transcripts are shown. The 167-nt band is a pausing artifact caused by the close proximity of this base to the end of the DNA template. Because position 143 is an A residue, the 142-nt terminated transcript is an artifact of limiting ATP. (B) Transcription of the trp leader by RNAP_Bs was performed at 23°C in the presence of a 150 μM concentration of each NTP (nonlimiting). (C) Transcription of the his leader by RNAP_Ec was performed at 37°C in the presence of 10 μM GTP (limiting) and a 150 μM concentration of the other three NTPs. Positions of the his pause and runoff (RO) transcripts are shown. (D) Transcription of the trp leader by RNAP_Bs was performed at 23°C with a 150 μM concentration of each NTP and 0.25 μM NusG_Bs ± 1 μM NusG_Ec. Pause efficiencies (Eff) are shown.

Fig. 3.

Effect of NusG_Bs and NusA_Bs on footprints of paused transcription bubbles in vitro. Single-round transcription was performed at 23°C with RNAP_Bs and the indicated Nus factors with limiting ATP. The DNA template was 5′ end-labeled on the ntDNA strand. Reactions were treated with KMnO₄ at the times shown above each lane. Chase reactions (Ch) were extended for an additional 10 min at 37°C in the presence of a 500 μM concentration of each NTP before KMnO₄ treatment. Positions of the trp pause sites (T107 and T144) and other T residues are marked. KMnO₄-treated and piperidine-cleaved fragments are 1 base shorter than the corresponding base in the DNA sequence. The relative band intensity of the T residues (133–146) for several lanes is shown below the gel.

Fig. 4.

Effect of NusG_Bs and NusA_Bs on footprints of U144 paused transcription bubbles in vivo. WT (+) or nusG knockout (−) strains contained or did not contain (−) a xylose-inducible gene on a plasmid as shown. Cells were treated with KMnO₄ at the times shown above each lane after rifampicin addition to block reinitiation of RNAP. Primer extension reactions were performed on plasmid DNA purified from KMnO₄-treated cells. The positions of T135, T144 (pause site), and T156 are marked.