RNA polyadenylation occurs not only in eukaryotes but also in bacteria. In prokaryotes, polyadenylated RNA molecules are usually degraded more efficiently than non-modified transcripts. Here we demonstrate that two transcripts, which were shown previously to be substrates for poly(A) polymerase I (PAP I), Escherichia coli lpp messenger RNA and bacteriophage λ oop RNA, are polyadenylated more efficiently in slowly growing bacteria than in rapidly growing bacteria. Intracellular levels of PAP I varied in inverse proportion to bacterial growth rate. Moreover, transcription from a promoter for the pcnB gene (encoding PAP I) was shown to be more efficient under conditions of low bacterial growth rates. We conclude that efficiency of RNA polyadenylation in E. coli is higher in slowly growing bacteria because of more efficient expression of the pcnB gene. This may allow regulation of the stability of certain transcripts (those subjected to PAP I-dependent polyadenylation) in response to various growth conditions. EMBO reports 4, 172-177 (2003) doi:10.1038/sj.embor.embor733
INTRODUCTIONDespite the fact that bacterial poly(A) polymerase was described 40 years ago (August et al., 1962), specific polyadenylation at the 3′ end of RNA was for a long time believed to be unique to eukaryotic messenger RNAs. Studies of the last ten years, initiated by the discovery of the structural gene for poly(A) polymerase I (PAP I) in Escherichia coli (Cao & Sarkar, 1992a), clearly demonstrated that prokaryotic RNAs are also polyadenylated.Two poly(A) polymerases, PAP I and PAP II, were discovered in E. coli. PAP I, which is responsible for over 90% of poly(A) polymerase activity in E. coli cells (O'Hara et al., 1995;Mohanty & Kushner, 1999a), is encoded by the pcnB gene (Cao & Sarkar, 1992a), but the gene encoding PAP II is still unknown (Mohanty & Kushner, 1999b). One could speculate that polynucleotide phosphorylase (PNPase) functions as an exonuclease and also as a poly(A) polymerase, which could account for the activity of PAP II. Various RNA molecules are polyadenylated with different efficiency, and it appears that the susceptibility of certain RNAs to polyadenylation depends on the presence of single-stranded segments at either the 5′ or the 3′ end and monophosphorylation at an unpaired 5′ terminus (Feng & Cohen, 2000;Yehudai-Resheff & Schuster, 2000).There are many reports indicating that bacterial RNA polyadenylation leads to decreased stability of transcripts (O'Hara et al., 1995;Xu & Cohen, 1995;Szalewska-Pal / asz et al., 1998;Blum et al., 1999). Results of other experiments showed either stabilization of specific transcripts, or a lack of effect on the halflives of other mRNAs after enhanced polyadenylation (Mohanty & Kushner, 1999a). However, transcripts stabilized or unaffected by polyadenylation are very rare, and it is therefore generally accepted that this process may regulate gene expression by promoting RNA degradation (for reviews see Sarkar, 1996;Carpousis et al., 1999;Rauhut & Klug, 1999;Regnier & Arraiano, 2000;...