Rho-dependent termination and RNase E-mediated cleavage: dual pathways for RNA 3' end processing in polycistronic mRNA

Abstract

"Pre-full-length" transcripts are produced at the end of the polycistronic galactose (gal) operon, 5' galE-galT-galK-galM 3', via Rho-dependent transcription termination (RDT) and -independent transcription termination. The 3' end of the full-length galETKM mRNA is acquired by exonucleolytic processing of the 3'-OH ends of the pre-full-length transcripts. However, the gal operon produces an mRNA termed galE whose 3' end forms approximately 120 nucleotides downstream of the galE stop codon, within the subsequent gene, galT, thereby establishing polarity in gene expression. In this study, we investigated the molecular processes that generate the 3' end of galE mRNA. We discovered that the 3' ends of pre-galE mRNA are produced in the middle of galT as a result of the combination of two separate molecular processes-one previously reported as RDT and the other as unreported RNase E-mediated transcript cleavage. The 3' ends of pre-galE mRNA undergo exonucleolytic processing to the 3' end of galE mRNA observed in vivo. A hairpin structure containing an 8 bp stem and a 4-nucleotide loop, located 5-10 nucleotides upstream of the 3' ends of galE mRNA, blocks exoribonuclease digestion and renders transcript stability. These findings demonstrate that RNase E-contrary to its general role in mRNA degradation-produces RNA 3' ends that regulate polarity in gene expression.IMPORTANCEThis study reports the findings of two molecular mechanisms that generate the 3' ends of pre-galE mRNA in the gal operon, viz., Rho-dependent transcription termination and RNase E-mediated cleavage. These 3' ends are subsequently processed to produce stable galE mRNA with a hairpin structure that prevents exoribonuclease degradation. This mechanism establishes gene expression polarity by generating the 3' end of galE mRNA within galT in contrast to the usual mRNA degradation role of RNase E. The study reveals a unique role of RNase E in mRNA processing and stability.

Keywords: RNase E-mediated cleavage; Rho-dependent transcription termination; exoribonuclease digestion; mRNA 3′ end; polarity.

Fig 1

A hairpin structure, galE hairpin, is responsible for the functioning of the 3′ end of galE mRNA as an exo-block. (A) Galactose operon (top) and nucleotide sequence (bottom) of the 3′ end of galE mRNA (green) in the WT gal operon. The galE stop codon is highlighted in red, and the galT initiator codon and SD sequence are highlighted in green and blue (underlined), respectively. The galE hairpin structure is depicted based on base complementarity between positions 1142 and 1161. Numbers indicate the nucleotide residue coordinate of the gal operon, which starts from the transcription initiation site of the galP1 promoter. (B) Schematics of galE-hMM2 and galE-hMM5 mutants. The base changes are in red. The gal mutants were generated in the single-copy plasmid, pGal, where the entire gal operon is cloned and assayed in MG1655 cells from where the entire gal has been removed, MG1655Δgal. (C) 3′ RACE assay of galE mRNA 3′ ends from MG155 cells harboring the plasmid-borne galE-hMM2 and galE-hMM5 mutant operons. The galE mRNA 3′ ends (1166–1172) are missing in lanes 2 and 3. DNA sequencing ladders that serve as length markers are in lanes marked G, A, T, and C.

Fig 2

Exoribonuclease digestion initiates molecular processes that generate the 3′ ends of galE mRNA. (A) Schematics of double hairpin (DH) mutants: DH1200, DH1600, DH1700, and DH1800. The inserted hairpins are in red. (B) 3′ RACE assay of galE mRNA 3′ ends from MG155 cells harboring the plasmid-borne gal DH1600, DH1700, and DH1800 mutant operons. The presence of residual RNA secondary structures that cause slower migration may be the cause of larger bands above the galE mRNA 3′ ends. (C) Relative band intensity of the galE mRNA 3′ ends for the 3′ RACE assay from (B). Error bars represent the mean fold-change ± standard deviation from three independent experiments.

Fig 3

RNase E-mediated endonucleolytic cleavage and RNase II-mediated exonucleolytic processing are the source of the 3′ ends of pre-galE. (A) Northern blot with the E probe of the gal mRNA from GW20Δgal (temperature-sensitive RNase E mutant) cells. Northern blot of E DNA probe (500 bp) was used for the assay that was generated by polymerase chain reaction (PCR) amplification with primers corresponding to the galE region (from +27 to +527 in gal coordinates), subsequently radiolabeled with 32P as previously described. Shown at the bottom is 16S rRNA, which was used as a loading control. (B) Relative band intensity of galE mRNA for the northern blot assay from (A). Error bars represent the mean fold-change ± standard deviation from three independent experiments. (C) 3′ RACE assay of the 3′ ends of galE mRNA from GW20Δgal (temperature-sensitive RNase E mutant) cells. Cells were cultured at both permissive (30°C) and nonpermissive (44°C) temperatures for analysis. (D) Relative band intensity of galE mRNA 3′ ends for the 3′ RACE assay from (C). Error bars represent the mean fold-change ± standard deviation from three independent experiments. (E) Nucleotide sequence of the upper (red) and lower (green) 3′ ends of galE mRNA based on the 3′ RACE assay from (C). (F) 3′ RACE assay of galE mRNA 3′ ends from ΔgalΔrnb cells, where RNaseII is deleted from the chromosome harboring the plasmid-borne gal DH1200 mutant operon. The red dot shows the shift in the 3′ ends in the absence of RNase II. (G) 5′ RACE assay of gal mRNA to identify RNase E cleavage from GW20Δgal cells. The lanes show the 5′ ends of RNA at the permissive (30°C, lanes 1 and 3) and the nonpermissive (44°C, lanes 2 and 4) temperatures between 1701 and 1971. The red arrow represents the 5′ ends that are absent at the nonpermissive temperature.

Fig 4

Rho and RNase E affect galE mRNA. (A) Northern blot with the E probe of gal mRNA from MG1655 cells harboring the plasmid-borne galT start° mutant operon with or without the Rho inhibitor bicyclomycin (BCM). To determine whether RDT produces galE mRNA, we treated MG1655 cells (OD600 of 0.6) in LB medium with 25 µg/mL BCM for 10 min. (B) Northern blot with the E probe of gal mRNA from GW20Δgal cells harboring the plasmid-borne galT start° mutant operon. (C) 3′ RACE assay of galE mRNA 3′ ends from GW20Δgal cells with or without the Rho inhibitor BCM. Inactivation of RNase E eliminates the “lower bands,” as it processes RNA. Conversely, inhibiting Rho disrupts transcription termination, leading to the accumulation of unprocessed or terminated RNA, which corresponds to the loss of “upper bands.” This distinction provides a compelling basis for differentiating the two 3' ends (see Fig. 3E). (D) Relative band intensity of galE mRNA 3′ ends for the 3′ RACE assay from (C). Error bars represent the mean fold-change ± standard deviation from three independent experiments.

Fig 5

Transcription model. (A) Model representing a transcription–translation complex in which the leading ribosome’s 30S remains linked to the paused RNA polymerase prolonged by the galE hairpin following the cessation of galE translation termination. In a paused state, NusG-S10 might help maintain the position of RNAP on the DNA, ensuring that transcription resumes correctly without disassociating or misprocessing the transcript. (B) Model of transcription termination in which the RNA polymerase is stalled downstream of the galE stop codon, and Rho is brought to it by the NusG protein. (C) RNase E-mediated cleavage model, in which RNase E cleaves the transcript’s RNA free from the ribosome and subjects it to 3′ to 5′ exonucleolytic digestion, which is blocked by the galE hairpin to produce the 3′ end of galE mRNA.