The rpsO mRNA, encoding ribosomal protein S15, is only partly stabilized when the three ribonucleases implicated in its degradation-RNase E, polynucleotide phosphorylase, and RNase II-are inactivated. In the strain deficient for RNase E and 3'-to-5' exoribonucleases, degradation of this mRNA is correlated with the appearance of posttranscriptionally elongated molecules. We report that these elongated mRNAs harbor poly(A) tails, most of which are fused downstream of the 3'-terminal hairpin at the site where transcription terminates. Poly(A) tails are shorter in strains containing 3'-to-5' exoribonucleases. Inactivation of poly(A) polymerase I (pcnB) prevents polyadenylylation and stabilizes the rpsO mRNA if RNase E is inactive. In contrast polyadenylylation does not significantly modify the stability of rpsO mRNA undergoing RNase E-mediated degradation.

The rpsO monocistronic messenger, encoding ribosomal protein S15, is destabilized upon polyadenylation occurring at the hairpin structure of the transcription terminator t1. We report that mRNA fragments differing from the monocistronic transcript by their 3′ termini are also polyadenylated in the absence of polynucleotide phosphorylase and RNase II. Some of these 3′ extremities result from endonucleolytic cleavages by RNase E and RNase III and from exonucleolytic degradation. Most of these mRNA fragments are destabilized upon polyadenylation with the exception of the RNA species generated by RNase III. RNase E appears to reduce the amount of poly(A) added at the transcription terminator t1.

No abstract

SummaryIn Escherichia coli , RNA degradation is mediated by endonucleolytic processes, frequently mediated by RNase E, and also by a poly(A)-dependent mechanism. The dominant pathway of decay of the rpsO transcripts is initiated by an RNase E cleavage occurring at a preferential site named M 2 . We demonstrate that mutations which prevent this cleavage slow down degradation by RNase E. All these mutations reduce the single-stranded character of nucleotides surrounding the cleavage site. Moreover, we identify two other cleavage sites which probably account for the slow RNase E-mediated degradation of the mutated mRNAs. Failure to stabilize the rpsO transcript by appending a 5 ¢ ¢ ¢ ¢ hairpin indicates that RNase E is not recruited by the 5 ¢ ¢ ¢ ¢ end of mRNA. The fact that nucleotide substitutions which prevent cleavage at M 2 facilitate the poly(A)-dependent degradation of the rpsO transcripts suggest an interplay between the two mechanisms of decay. In the discussion, we speculate that a structural feature located in the vicinity of M 2 could be an internal degradosome entry site promoting both RNase E cleavages and poly(A)-dependent degradation of the rpsO mRNA. We also discuss the role of poly(A)-dependent decay in mRNA metabolism.