Logo of narLink to Publisher's site
Nucleic Acids Res. 2007 Apr; 35(8): 2494–2502.
Published online 2007 Mar 29. doi:  10.1093/nar/gkm120
PMCID: PMC1885654

Polyadenylation of a functional mRNA controls gene expression in Escherichia coli


Although usually implicated in the stabilization of mRNAs in eukaryotes, polyadenylation was initially shown to destabilize RNA in bacteria. All the data are consistent with polyadenylation being part of a quality control process targeting folded RNA fragments and non-functional RNA molecules to degradation. We report here an example in Escherichia coli, where polyadenylation directly controls the level of expression of a gene by modulating the stability of a functional transcript. Inactivation of poly(A)polymerase I causes overexpression of glucosamine–6-phosphate synthase (GlmS) and both the accumulation and stabilization of the glmS transcript. Moreover, we show that the glmS mRNA results from the processing of the glmU-glmS cotranscript by RNase E. Interestingly, the glmU-glmS cotranscript and the mRNA fragment encoding GlmU only slightly accumulated in the absence of poly(A)polymerase, suggesting that the endonucleolytically generated glmS mRNA harbouring a 5′ monophosphate and a 3′ stable hairpin is highly susceptible to poly(A)-dependent degradation.


It is well established now that mRNA decay has an integral role in the control of gene expression. mRNAs can be degraded by a number of mechanisms that act independently, in parallel or sequentially. Several cis- and trans-acting factors including higher order RNA structures (1,2), RNA-binding proteins (3–5) and translating ribosomes (6,7) control the sensitivity of an mRNA molecule to degradation. Hairpin structures located at the 3′ end of operons or in intercistronic regions protect mRNA molecules from digestion by 3′–5′ exoribonucleases. To overcome these barriers, bacteria initiate RNA degradation by endoribonucleolytic cleavages mostly carried out by RNase E in E.coli. RNase III and RNase G have also been found to initiate RNA decay in certain cases (8,9). RNA fragments resulting from endonucleolytic cleavages can be either further degraded endonucleolytically or by the 3′–5′ exonucleases, polynucleotide phosphorylase (PNPase), RNase II and RNase R [reviewed in (10)]. Oligo(A) tails can be added to RNA fragments resulting from endo- or exonucleolytic cleavages and promote their exonucleolytic degradation (11).

As opposed to their predominant role in eukaryotes, poly(A) tails act as RNA-destabilizing elements in bacteria (12). A destabilizing function for poly(A) tails has also been subsequently reported in mitochondria of higher plants (13), chloroplasts (14), and the nucleus of yeast and humans [reviewed in (15,16)]. The enzyme responsible for the majority of polyadenylation in E.coli, poly(A)polymerase (PAP I), is encoded by the pcnB gene (17). The consensus idea emerging from numerous studies of pcnB mutations is that the main function of polyadenylation is to facilitate degradation of short folded RNAs such as RNA fragments produced during the processing of longer RNA molecules (17–19) and of a non-functional mutated tRNA (20). This latter example suggested that poly(A)-dependent decay is involved in quality surveillance of bacterial RNA, as previously reported in yeast (21). A special function of polyadenylation is in the control of copy number of ColEI plasmids. The decay of the regulatory RNA I is initiated by RNase E cleavage five nucleotides from the 5′ end, which activates very fast decay via the PAP I-PNPase pathway (22). There are also data indicating that PAP I may control the stability of primary transcripts (23,24). For instance, mutations that inactivate RNase E-dependent decay render mRNA more sensitive to poly(A)-dependent degradation (25–27). In addition, Aiso et al. showed that the pcnB mutation affected degradation of rmf mRNA in stationary phase, but this effect could be indirect as it required de novo mRNA synthesis (24). However, there has been no evidence so far that polyadenylation directly controls protein levels by modulating mRNA stability.

In an attempt to identify genes whose expression is affected by mRNA polyadenylation, we compared proteins of a wild-type and PAP I-deficient strains on 1D and 2D PAGE gels (2D-gel) There were several differences and, in particular, one polypeptide; glucosamine-6-phosphate synthase (GlmS) was significantly more abundant in the mutant strain. In this work, we demonstrate that the overexpression of GlmS is correlated with the accumulation and stabilization of the monocistronic glmS mRNA resulting from the processing of the glmU-glmS cotranscript.


Bacterial strains

The name and origin of the strains used here are listed in Table 1. New strains were constructed by transduction of the ΔpcnB (KanR) and rng::CmR alleles of strains SK7988 and CA244 cafA, respectively, with bacteriophage P1. Cells were grown in a MOPS medium supplemented with 0.2% glucose (w/v) and all amino acids and vitamins (28,29), or in an LB medium at the temperature specified in the text.

Table 1.
E.coli strains used in this work

Functional lifetime of bulk mRNA

Cells were grown in MOPS medium supplemented with glucose and amino acids lacking methionine and cysteine and then treated with rifampicin (500 μg/ml; zero time). At timed intervals, 1 ml aliquots were labelled for 1 min with 25 μCi [35S] methionine (>1000 Ci/mmol) at 37°C. Cells were then treated with excess cold methionine and sodium azide for 5 min and pelleted. Frozen cells were resuspended in 0.1 M Tris–HCl, pH 8, 0.4 M NaCl, 0.1 mM EDTA, 1 mM 2-β-mercaptoethanol, lysed by sonication on ice, and total proteins were separated on 10% SDS-PAGE. Gels were stained with Coomassie blue and dried, and the radioactive pattern was analyzed with a PhosphoImager.


Total proteins were separated on 10% SDS-PAGE, the proteins were transferred to Hybond C-super (Amersham) by electroblotting. The membrane was incubated with polyclonal anti-GFAT (GFAT = human glucosamine-6P synthase enzyme) antibodies, which cross-react against E.coli GlmS. The bound antibodies were detected by reaction with [125I]-labelled protein A.

RNA extraction and analysis

Templates for the synthesis of the different RNA probes were obtained by PCR amplification using the primers described in Table 2. The upstream primer includes the T7 promoter. RNA were synthesized by T7 RNA polymerase yielding uniformly labelled RNA with [α-32P] UTP (30). 5S rRNA was probed with the 5′-labelled ACTACCATCGGCGCTACGGC oligonucleotide.

Table 2.
PCR primers. Italics correspond to RNA polymerase sequence promoter of the T7 phage

RNA preparation, reverse transcription (using the 5′-CGGTATTCCAGACGACGTAA primer complementary to the glmS mRNA) and Northern blotting on 10% agarose formaldehyde gel were previously described (31,32).

2D gel electrophoresis

Three independent cultures of N3433 and IBPC 903 were grown in LB at 37°C. 50 ml were withdrawn at OD650 = 0.5. Soluble protein extracts were prepared from cells sonicated in water, 20 mg/ml DNase and RNase were added after centrifugation at 10 000 g for 30 min. First-dimension electrophoresis was carried out on a non-linear immobilized pH 3–11 gradient IPG strip (18 cm BioRad), and the second dimension by 12% SDS-PAGE. The six samples were analyzed in parallel. Densitometric quantification of the six blue-colloidal-stained 2D gels was performed using the PDQuest™ 2D Analysis software. The analysis of these gels revealed that expression of two polypeptides was greatly increased in the mutant. These protein spots were excised from gels for trypsin in-gel digestion and MALDI-TOF analysis.


PAP I deficiency modifies gene expression

The presence of 3′ terminal hairpins has been shown to protect mRNAs against exonucleolytic degradation. The fact that the addition of poly(A) tails facilitates exonucleolytic degradation of tightly folded RNA prompted us to investigate whether polyadenylation affects stability of normal cellular mRNAs that possess stable 3′ terminal hairpins. To do this, we examined the functional half-life of mRNAs by comparing the incorporation of radiolabelled methionine into proteins at fixed times after inhibiting RNA synthesis with rifampicin. Newly synthesized proteins were pulse-labelled with [35S] methionine at different times after the addition of rifampicin to exponentially growing cultures of the PAP+ N3433 strain and of its pcnB::kanR derivative IBPC 903. Total proteins were then separated on SDS-polyacrylamide gels that were autoradiographed. Strikingly, the PAP I-deficient strain synthesized a 66-kDa polypeptide for at least 3 min after inhibition of transcription initiation, the protein was hardly visible in PAP I-proficient cells (Figure 1A, right panel). Consistent with this observation, Coomassie blue staining of the gel shows that a 66-kDa polypeptide dramatically accumulates in the absence of PAP I (Figure 1A, left panel). These data suggest that the mRNA coding this 66-kDa polypeptide is either more efficiently translated or more likely that it is more stable and hence more abundant in the absence of polyadenylation. Failure to detect the radioactive polypeptide in the mutant 5 min after transcription inhibition indicates that the corresponding mRNA is not exceptionally stable; its functional half-life is shorter than 2 min even in the absence of polyadenylation.

Figure 1.
Effect of the pcnB mutation on functional mRNA stability and protein synthesis. (A) Cultures of N3433 and IBPC 903 (pcnB) were grown in MOPS medium containing glucose and amino acids lacking methionine and cysteine until mid-log phase. Cells were treated ...

There were a few other differences in the pulse-labelled pattern of wt and pcnB strains. Two polypeptides whose expression is enhanced in the pcnB strain are indicated by asterisks (Figure 1A). It has been reported that overexpression of PAP I induces functional stabilization of rne and pnp mRNAs (33). However co-migration of the labelled proteins with a degradosome preparation did not detect any changes in expression of proteins corresponding to the sizes of these degradosome components.

In this work, we concentrate on the 66-kDa protein, which shows the strongest effect of pcnB-induced overproduction. In order to rule out possible strain or allele-specific effects of PAP I inactivation, we checked whether the 66-kDa polypeptide also accumulates in a strain carrying another pcnB allele and in different genetic backgrounds (Table 1 and Figure 1B). Indeed, the 66-kDa polypeptide accumulates in strains harbouring the pcnB::kanR (IBPC 903 and CA244-PAP) and the pcnB80 point mutation (IBPC 694), but is not detected in their pcnB+ parents. Moreover, this polypeptide is overproduced in the IBPC 903 mutant strain grown either in MOPS medium containing glucose and amino acids (Figure 1A) and in LB medium (Figure 1B). Strain N3433 and its pcnB::kanR derivative grown in LB medium were used in the experiments below.

GlmS is upregulated in PAP I mutant

In order to identify the 66-kDa polypeptide overproduced in the mutant, soluble proteins extracted from wild-type and pcnB strains were analyzed by 2D gel electrophoresis. Figure 2A shows that two polypeptides of about 66 kDa, migrating to adjacent positions on the gel, were much more abundant in the mutant than in the wild-type strain. Both spots were excised, and mass spectrometry identified these proteins as glucosamine-6-phosphate synthase (GlmS) (molecular weight 66.9 kDa), one of which may correspond to a modified GlmS form, although no modification of this protein was previously reported in E.coli. Densitometric quantification of the blue-colloidal-stained 2D gels revealed that the two spots of GlmS are 10- and 30-fold more abundant in the pcnB mutant compared to the wild-type strains. Finally, Western blotting with anti-GlmS antibodies confirmed that a 66-kDa polypeptide is produced much more efficiently in the PAP I mutant than in the wild-type strain (Figure 2B).

Figure 2.
The pcnB mutation increases the amount of GlmS. (A) Protein profile of N3433 (wt) and IBPC 903 (pcnB) strains grown in LB medium were analyzed by 2D-PAGE. Representative gels from experiments performed in triplicate were presented. Soluble proteins with ...

GlmS encodes the enzyme that synthesizes glucosamine-6-phosphate (GlcN6P), an essential precursor of lipopolysaccharides and peptidoglycan. The preceding gene on the chromosome, glmU, encodes the bifunctional enzyme glucosamine-1-phosphate acetyl transferase: UDP-GlcNAc synthetase (34,35), which synthesizes UDP-GlcNAc, the first dedicated component for the cell wall components. Previous work has shown that expression of both glmS and glmU is subject to regulation by the availability of amino sugars (35,36). Both genes are believed to be transcribed from two promoters located upstream of the first gene of the operon, glmU, and no promoter has been identified in the intergenic region.

The glmU-glmS primary transcript is processed in the intercistronic region

To investigate whether the overproduction of GlmS in cells lacking PAP I results from an accumulation of its mRNA, we compared the abundance of the glmS transcript in wild-type and PAP I-deficient strains. Total RNA was extracted from cells growing exponentially at 37°C, and amounts of this transcript were monitored on Northern blot. Probing for glmS detected a band of 1.9 kb that was about 16-fold more abundant in the pcnB strain than in the wild-type cells (Figure 3). A 1.9 kb transcript is sufficient to code for a protein of 66 kDa, but not for the glmU-glmS (glmUS) cotranscript. The very low amount of the 3.5-kb mRNA visible in Figure 3 led us to postulate that the expected glmUS cotranscript is very rapidly cleaved into the shorter 1.9-kb glmS mRNA detected in Figure 3.

Figure 3.
The pcnB mutation increases amounts of the glmS mRNA. Total RNA (12 μg) from N4333 (wt) and IBPC 903 (pcnB) strains grown in LB medium at 37°C were analyzed on Northern blot and probed for the glmS mRNA (upper part). Reprobing ...

We therefore investigated whether the endonucleases involved in mRNA maturation and degradation of E.coli mRNAs, namely RNase III (rnc), RNase E (rne) and RNase G (rng), cleave the glmUS cotranscript. Isogenic strains deficient for each of these RNases were used as well as a rne-rng double mutant. The effect of RNase E inactivation was examined by using the thermosensitive rne3071 allele. For comparison, RNAs from all strains were prepared from bacteria grown at 30°C (time 0) and 15 and 30 min after a temperature shift to 44°C required to inactivate thermosensitive RNase E. Total RNAs were then analyzed on Northern blots probed for glmS (Figure 4B). Neither the rnc105, nor the rng::CmR mutations allowed detection of the glmUS primary transcript or an increased level of glmS mRNA. In contrast, RNase E inactivation caused the accumulation of a long transcript with the size of the expected glmUS mRNA (3.5 kb). This result demonstrated that the glmUS cotranscript is cleaved by RNase E. Moreover, inactivation of RNase G in the absence of RNase E increases the level of the dicistronic transcript, suggesting that in the absence of RNase E there is some processing of the long primary transcript by RNase G but that normally rapid degradation by RNase E eliminates any effect of rng mutation. Probing the Northern blot for the glmU mRNA confirms that the full-length 3.5-kb transcript was only present in the rne and rne-rng strains, and that it was more abundant in the double mutant (Figure 4B).

Figure 4.
Fate of the transcripts of the glmU-glmS cotranscript. (A) Structure of the glmU-glmS transcript and of the processed mRNAs. The map of the glmU-glmS genes of E.coli shows the location of the 2 glmU promoters (arrows), of the transcription terminator ...

Surprisingly, thermal inactivation of RNase E leads not only to the accumulation of the uncleaved glmUS mRNA but also to an increased level of the downstream glmS monocistronic mRNA. The detection of the glmS monocistronic transcript implies that residual cleavage still occurs in the intergenic region of the glmUS transcript after the temperature shift that inactivates RNase E. RNase G is not responsible for this cleavage since the 1.9-kb glmS transcript was still detected in the double rne-rng strain. We previously observed residual processing of another RNase E site after thermal inactivation, implying that thermal inactivation of RNase E carrying the rne3071 mutation is not complete (31). Alternatively, the glmS transcript could be due to another unidentified endonuclease, or to the presence of a glmS-specific promoter located near the intergenic region. We note that the monocistronic glmS transcript is somewhat diffuse in the rne and rne-rng strains, which might be consistent with it being due to something other than cleavage at the major RNase E site. The observation that the glmS transcript is less abundant in rnc cells than in wt while no glmUS transcript is detected (Figure 4B lanes 1–3 and 13–15, respectively) suggests that RNase III may indirectly participate to glmS stability.

To localize the processing site that generates the 1.9-kb glmS mRNA, primer extension was performed with an oligonucleotide hybridizing to the beginning of the glmS structural gene. Three 5′ extremities were located, all mapping in the glmU-glmS intergenic region. The most abundant (band a on Figure 4D) is located just downstream of the stop codon of glmU mRNA. The AU-rich sequence surrounding this extremity (AAAAAGUGA⇓UUCU) suggests that it might be generated by RNase E. Two minor 5′ ends located in the intercistronic region were also observed when PAP I was inactive. One (band b) maps 20 nt downstream of the major cleavage (band a) in the sequence CACAU⇓GGG, and the other (band c) maps in a GC-rich sequence that does not conform to the usual RNase E cleavage site sequence (Figure 4D).

Polyadenylation promotes degradation of the glmS transcript

We next investigated whether the abundance of the glmS mRNA in the pcnB strain reflected the stabilization of this molecule. For this purpose, we measured the half-life of the glmS transcript in wild-type and pcnB strains at 37°C. The very low level of the bicistronic transcript detected under these conditions prevents bias due to the production of new glmS molecules during the experiment. The intracellular concentration of the monocistronic glmS mRNA was followed versus time after rifampicin addition (Figure 5A and B). The same Northern blot was probed for 5S rRNA to normalize the data. Figure 5A and B show that the glmS mRNA was more stable in the pcnB strain where its half-life was about 2.95 ± 0.17 min. The very low amount of glmS mRNA in the wild-type strain disappeared very rapidly; it was no longer detected 1.5 min after rifampicin addition. Its half-life was estimated to be 0.72 ± 0.17 min. The relatively short half-life of the glmS transcript in the pcnB strain agrees with the functional half-life measurement showing that GlmS synthesis was already considerably reduced 3 min after inhibition of transcription (Figure 1A). These data clearly demonstrate that PAP I inactivation stabilizes the glmS monocistronic mRNA. However, the ratio of half-lives is only a factor of about four, which is much less than the ratio of the abundance of the glmS transcript levels observed in the pcnB and the wild-type strains (about 16-fold) (Figures 3 and and5A)5A) or of the GlmS protein on 2D gels (Figure 2A). It is clear that the pcnB mutation has some additional effect on the level of the glmS transcript. Whether this involves the RNase E processing of the bicistronic transcript or activation of additional transcription from the glmU promoters or a glmS specific promoter has not been clarified.

Figure 5.
PAP I destabilizes the glmS transcript. (A) Autoradiograph comparing the decay kinetics of glmS mRNA in strains N3433 (wt) and IBPC 903 (pcnB). Times after inhibition of transcription at which glmS (upper part) and 5S (lower part) RNA contents were determined ...

Poly(A)polymerase specifically affects the stability of the glmS mRNA

A typical Rho-independent terminator is located after the glmS gene, and one interpretation of the data described above is that polyadenylation facilitates exonucleolytic degradation of the 3′-terminal G-C rich hairpin of the transcription terminator, which otherwise protects the glmS transcript. Surprisingly, however, the glmUS cotranscript, which should have the same 3′ terminal-protecting motif, is much less abundant than the glmS mRNA in the absence of PAP I. RNAs extracted from wild-type cells and isogenic bacteria deficient for RNase E or PAP I or both RNase E and PAP I were analyzed on Northern blots and probed for glmS and glmU mRNAs (Figure 4B and C). The fact that the amount of the dicistronic mRNA detected in the rne-pcnB strain was slightly more abundant than in rne cells suggested that the long primary transcript was, to some extent, sensitive to poly(A)-assisted exonucleolytic decay. In agreement with this hypothesis, a smear of glmS mRNAs, ranging from 1.9 to 3.5 kb, which presumably results from exonucleolytic degradation of the cotranscript, was detected when PAP I was active but not when it was inactive. However, the monocistronic glmS mRNA accumulated much more strongly than the dicistronic transcript under identical conditions (i.e. after thermal inactivation of RNase E in the pcnB strain). These data reinforce the above observation that thermal inactivation of rne3071 is not complete or that some other mechanism exists to generate the 5′ end of the glmS mRNA. Irrespective of its origin, these data demonstrated that the monocistronic glmS mRNA is much more sensitive to poly(A)-dependent degradation than the glmUS cotranscript. Moreover, accumulation of both the glmS monocistronic and the glmUS dicistronic transcripts in a pcnB strain after RNase E inactivation confirms (see above) that they are both degraded by RNase E. Another unexpected observation was that the glmS mRNA, which accumulates in the pcnB-rne+ strain growing at 30°C (Figure 4B, lane 16) and 37°C (Figure 3), disappeared after the temperature shift to 44°C (Figure 4B, lanes 17–18), suggesting that the RNase E-mediated degradation of glmS mRNA is more efficient at 44°C than at 37°C.

Degradation of the glmU transcript

Cleavage of the glmUS cotranscript would be expected to generate a glmU-specific transcript that is not detected in the wt strain (Figure 4B, lane 13). However in the rne strain after temperature shift, an additional transcript of about 1.5 kb is observed with the glmU probe (Figure 4B, lanes 11–12 and 4C lanes 2–3). Again, this transcript must be either due to residual RNase E activity in the thermal inactivated cultures, or the result of another endonuclease. As the transcript is only slightly more abundant in the rne-pcnB strain than the rne strain (Figure 4C, lanes 5–6 and 2–3, respectively), it is not dependent on polyadenylation for its degradation, unlike the glmS monocistronic transcript.


We present here evidence that poly(A)polymerase can negatively affect gene expression. Previously, it was thought that the role of the poly(A)-dependent pathway was to increase the turnovers of defective tRNAs (37) and non-processed RNA precursors (38), and to clear the cell of endonucleolytically produced, tightly folded mRNA fragments resistant to further degradation by exonucleases (39,40). Here, we show that overproduction of glucosamine-6P synthase, an essential enzyme in E.coli in the absence of exogenous aminosugars, occurs upon inactivation of PAP I, and that this accumulation is correlated with the accumulation of the glmS mRNA and the reduction in its decay-rate. This led us to conclude that poly(A)-dependent degradation of the functional glmS mRNA determines the yield of GlmS. In agreement with our observation, glmS ORF was recently shown, using macroarrays, to be highly polyadenylated when PAP I is overproduced (41). One cannot exclude, however, that polyadenylation acts indirectly on GlmS expression, e.g. by modulating the stability of a small regulatory RNA or another trans-acting factor.

We also demonstrate that the glmS monocistronic mRNA, stabilized in PAP I-deficient cells, likely results from the endonucleolytic maturation of the glmUS dicistronic primary transcript. Although the dicistronic transcript was only detected in cells lacking RNase E, the fact that the monocistronic glmS transcript is still produced in these bacteria suggests that glmUS could be processed by another enzyme. The 5′ mRNA extremity mapping just downstream of the UGA stop codon suggests that ribosome stalling could be implicated in the mechanism of glmUS RNA processing (42–45).

Interestingly, in spite of the fact that the glmS and glmUS mRNAs have identical 3′ terminal structures, we found that the primary transcript is much less sensitive to the poly(A)-assisted exonucleolytic route of decay than the monocistronic processed molecule. An important question arising is how the poly(A)-assisted machinery of decay recognizes its targets and, in the case described above, how it distinguishes the glmS monocistronic mRNA from glmUS primary transcript and other primary transcripts. The fact that the glmS and glmUS mRNA correspond to the 3′ part of the dicistronic transcript implies that it is the nature of the 5′ part of the RNA that determines its sensitivity to this pathway. One possibility is that, as described earlier for RNA I, the 5′ monophosphorylated terminus generated by RNase E facilitates polyadenylation of the glmS mRNA and thus promotes its degradation (22). Such a model could explain why the RNase E-processed mRNA is more sensitive to the poly(A)-dependent pathway than the primary transcript harbouring a 5′ triphosphate extremity. At the moment, the mechanism by which the 5′ end of the long 1.9-kb glmS mRNA could affect a reaction occurring at its 3′ end is not known. A greater accessibility of the 3′ end has been proposed to explain why the precursor of a mutated tRNATrp is degraded by a poly(A)-dependent pathway (37). One can imagine that refolding of the glmS mRNA consecutive to the processing of the primary transcript improves accessibility of PAP I and/or exoribonucleases at its 3′ end. It is possible, for example, that translation of glmS is coupled to that of the upstream gene glmU and that a reduction of ribosome loading of the glmS message consecutive to the removal of the upstream glmU cistron modifies the folding of the molecule.

It is worth mentioning here that the 5′ monophosphate extremity that may trigger the poly(A)-dependent decay of the glmS mRNA may also promote its degradation by RNase E (46). The fact that the glmS mRNA accumulates in the absence of PAP I at 30°C, 37°C but not at 44°C is consistent with the idea that the efficiency of its degradation by RNase E increases at high temperature. In contrast, this observation indicates that the contribution of the poly(A)-dependent pathway to the decay of this message increases at low temperature. It therefore appears that RNase E exerts a predominant role in mRNA decay around the optimum temperature of growth and that the poly(A)-dependent degradation may substitute for RNase E at lower temperatures. The idea that poly(A)-dependent decay can substitute for RNase E is also supported by earlier data showing that this pathway becomes effective on mRNAs that are no longer degraded by RNase E (25–27).

GlmS catalyzes the conversion of fructose-6-phosphate into glucosamine-6-phosphate, which then undergoes sequential transformations leading to the formation of UDP-N-acetylglucosamine, the major intermediate in the biosynthesis of all amino-sugar-containing macromolecules in the cell (47). Regulation of glmS expression occurs both at the transcriptional level (36) with NagC acting as both an activator and a repressor of the glmUS operon in E.coli (35) and at the posttranscriptional level (present work). Interestingly, the expression of the glmS gene of Bacillus subtilis is also subject to a regulatory mechanism involving the processing and the degradation of its mRNA. However, in these bacteria lacking orthologs of both RNase E and PAP I (48,49), an alternate strategy has emerged. A new class of ribozyme that cleaves the messenger of the glmS mRNA was discovered, which is activated by GlcN6P (50). In spite of some sequence homology in the glmU-glmS intergenic region in bacterial species closely related to E.coli, in-line probing did not reveal any structural changes or self-cleavage activity when the entire E.coli intergenic glmU-glmS region was incubated with either Glc6P or GlcN6P (Elaine Lee and Ron Breaker, personal communication). It therefore appears that different strategies exist that allow sensitive control of glmS mRNA and hence protein levels despite the fact that the machineries of mRNA processing and degradation are very different. Such pathways may present selective advantages that could explain their conservation in Gram-positive and Gram-negative bacteria.


We are grateful to J. Plumbridge for careful critical reading of the manuscript, discussions and her constant interest during this work. We thank M. P. Deutscher and S. D. Kushner for kindly supplying strains, M. A. Badet for antibodies and GlmS purified protein and E. Lee and R. Breaker for communication of unpublished results. Funding to pay the Open Access publication charge was provided by CNRS Departement des Sciences de la Vie.

Conflict of interest statement. None declared.


1. Belasco JG, Nillson G, von Gabain A, Cohen SN. The stability of E.coli gene transcripts is dependent on determinants localized to specific mRNA segments. Cell. 1986;46:245–251. [PubMed]
2. Bouvet P, Belasco J. RNA degradation in E.coli by endonuclease RNase E: control by 5′-terminal base pairing. Nature. 1992;360:488–491. [PubMed]
3. Folichon M, Arluison V, Pellegrini O, Huntzinger E, Regnier P, Hajnsdorf E. The poly(A) binding protein Hfq protects RNA from RNase E and exoribonucleolytic degradation. Nucleic Acids Res. 2003;31:7302–7310. [PMC free article] [PubMed]
4. Feng Y, Huang H, Liao J, Cohen SN. Escherichia coli poly(A) binding proteins that interact with components of degradosomes or impede RNA decay mediated by polynucleotide phosphorylase and RNase E. J. Biol. Chem. 2001;276:31651–31656. [PubMed]
5. Iost I, Dreyfus M. mRNAs can be stabilized by DEAD-box proteins. Nature. 1994;372:193–196. [PubMed]
6. Braun F, Le Derout J, Régnier P. Ribosomes inhibit an RNase E cleavage which induces the decay of the rpsO mRNA of Escherichia coli. EMBO J. 1998;17:4790–4797. [PMC free article] [PubMed]
7. Hambraeus G, Karhumaa K, Rutberg B. A 5′ stem-loop and ribosome binding but not translation are important for the stability of Bacillus subtilis aprE leader mRNA. Microbiology. 2002;148:1795–1803. [PubMed]
8. Bardwell JCA, Régnier P, Chen SM, Nakamura Y, Grunberg-Manago M, Court DL. Autoregulation of RNase III operon by mRNA processing. EMBO J. 1989;8:3401–3407. [PMC free article] [PubMed]
9. Wachi M, Kaga N, Umitsuki G, Clark DP, Nagai K. A novel RNase G mutant that is defective in degradation of adhE mRNA but proficient in the processing of 16S rRNA precursor. Biochem. Biophys. Res. Com. 2001;289:1301–1306. [PubMed]
10. Deutscher MP. Degradation of RNA in bacteria: comparison of mRNA and stable RNA. Nucleic Acids Res. 2006;34:659–666. [PMC free article] [PubMed]
11. Haugel-Nielsen J, Hajnsdorf E, Regnier P. The rpsO mRNA of Escherichia coli is polyadenylated at multiple sites resulting from endonucleolytic processing and exonucleolytic degradation. EMBO J. 1996;15:3144–3152. [PMC free article] [PubMed]
12. Dreyfus M, Régnier P. The poly(A) tail of mRNAs: bodyguard in eukaryotes, scavenger in bacteria. Cell. 2002;111:611–613. [PubMed]
13. Gagliardi D, Stepien PP, Temperley RJ, Lightowlers RN, Chrzanowska-Lightowlers ZM. Messenger RNA stability in mitochondria: different means to an end. Trends Genet. 2004;20:260–267. [PubMed]
14. Komine Y, Kwong L, Anguera MC, Schuster G, Stern DB. Polyadenylation of three classes of chloroplast RNA in Chlamydomonas reinhadtii. RNA. 2000;6:598–607. [PMC free article] [PubMed]
15. Anderson JT. RNA turnover: unexpected consequences of being tailed. Curr. Biol. 2005;15:R635–638. [PubMed]
16. West S, Gromak N, Norbury CJ, Proudfoot NJ. Adenylation and exosome-mediated degradation of cotranscriptionally cleaved pre-messenger RNA in human cells. Mol. Cell. 2006;21:437–443. [PubMed]
17. Cao G-J, Sarkar N. Poly(A) RNA in Escherichia coli: nucleotide sequence at the junction of the lpp transcript and the polyadenylate moiety. Proc. Natl. Acad. Sci. USA. 1992;89:7546–7550. [PMC free article] [PubMed]
18. He L, Söderbom F, Wagner EGH, Binnie U, Binns N, Masters M. PcnB is required for the rapid degradation of RNAI, the antisense RNA that controls the copy number of ColE1-related plasmids. Mol. Microbiol. 1993;9:1131–1142. [PubMed]
19. Xu F, Lin-Chao S, Cohen SN. The Escherichia coli pcnB gene promotes adenylylation of antisense RNAI of ColE1-type plasmids in vivo and degradation of RNAI decay intermediates. Proc. Natl. Acad. Sci. USA. 1993;90:6756–6760. [PMC free article] [PubMed]
20. Deutscher MP. Degradation of stable RNA in bacteria. J. Biol. Chem. 2003;278:45041–45044. [PubMed]
21. Kadaba S, Krueger A, Trice T, Krecic AM, Hinnebusch AG, Anderson J. Nuclear surveillance and degradation of hypomodified initiator tRNAMet in S. cerevisiae. Genes Dev. 2004;18:1227–1240. [PMC free article] [PubMed]
22. Xu F, Cohen SN. RNA degradation in Escherichia coli regulated by 3′ adenylation and 5′ phosphorylation. Nature. 1995;374:180–183. [PubMed]
23. Mohanty BK, Kushner SR. Analysis of the function of Escherichia coli poly(A) polymerase I in RNA metabolism. Mol. Microbiol. 1999;34:1094–1108. [PubMed]
24. Aiso T, Yoshida H, Wada A, Ohki R. Modulation of mRNA stability participates in stationary-phase-specific expression of ribosome modulation factor. J. Bacteriol. 2005;187:1951–1958. [PMC free article] [PubMed]
25. Hajnsdorf E, Braun F, Haugel-Nielsen J, Regnier P. Polyadenylylation destabilizes the rpsO mRNA of Escherichia coli. Proc. Natl. Acad. Sci. USA. 1995;92:3973–3977. [PMC free article] [PubMed]
26. O'Hara EB, Chekanova JA, Ingle CA, Kushner ZR, Peters E, Kushner SR. Polyadenylation helps regulate mRNA decay in Escherichia coli. Proc. Natl. Acad. Sci. USA. 1995;92:1807–1811. [PMC free article] [PubMed]
27. Marujo PE, Braun F, Haugel-Nielsen J, Le Derout J, Arraiano CM, Regnier P. Inactivation of the decay pathway initiated at an internal site by RNase E promotes poly(A)-dependent degradation of the rpsO mRNA in Escherichia coli. Mol. Microbiol. 2003;50:1283–1294. [PubMed]
28. Neidhardt FC, Bloch PL, Pedersen S, Reeh S. Chemical measurement of steady-state levels of ten aminoacyltransfer ribonucleic acid synthetases in Escherichia coli. J. Bacteriol. 1977;129:378–387. [PMC free article] [PubMed]
29. Neidhardt FC, Bloch PL, Smith DF. Culture medium for enterobacteria. J. Bacteriol. 1974;119:736–747. [PMC free article] [PubMed]
30. Hajnsdorf E, Régnier P. Host factor HFq of Escherichia coli stimulates elongation of poly(A) tails by poly(A)polymerase I. Proc. Natl. Acad. Sci. USA. 2000;97:1501–1505. [PMC free article] [PubMed]
31. Hajnsdorf E, Carpousis AJ, Régnier P. Nucleolytic inactivation and degradation of the RNase III processed pnp message encoding polynucleotide phosphorylase of Escherichia coli. J. Mol. Biol. 1994;239:439–454. [PubMed]
32. Hajnsdorf E, Steier O, Coscoy L, Teysset L, Régnier P. Roles of RNase E, RNase II and PNPase in the degradation of the rpsO transcripts of Escherichia coli: stabilizing function of RNase II and evidence for efficient degradation in an ams-rnb-pnp mutant. EMBO J. 1994;13:3368–3377. [PMC free article] [PubMed]
33. Mohanty BK, Kushner SR. Polyadenylation of Escherichia coli transcripts plays an integral role in regulating intracellular levels of polynucleotide phosphorylase and RNase E. Mol. Microbiol. 2002;45:1315–1324. [PubMed]
34. Mengin-Lecreulx D, van Heijenoort J. Copurification of glucosamine-1-phosphate acetyltransferase and N-acetylglucosamine-1-phosphate uridyltransferase activities of Escherichia coli: characterization of the glmU gene product as a bifunctional enzyme catalyzing two subsequent steps in the pathway for UDP-N-acetylglucosamine synthesis. J. Bacteriol. 1994;176:5788–5795. [PMC free article] [PubMed]
35. Plumbridge J. Co-ordinated regulation of amino sugar biosynthesis and degradation: the NagC repressor acts as both an activator and a repressor for the transcription of the glmUS operon and requires two separated NagC binding sites. EMBO J. 1995;14:3958–3965. [PMC free article] [PubMed]
36. Plumbridge JA, Cochet O, Souza JM, Altamirano MM, Calcagno ML, Badet B. Coordinated regulation of amino sugar-synthesizing and -degrading enzymes in Escherichia coli K-12. J. Bacteriol. 1993;175:4951–4956. [PMC free article] [PubMed]
37. Li Z, Reimers S, Pandit S, Deutscher MP. RNA quality control: degradation of defective transfer RNA. EMBO J. 2002;21:1132–1138. [PMC free article] [PubMed]
38. Li Z, Deutscher MP. RNase E plays an essential role in the maturation of Escherichia coli tRNA precursors. RNA. 2002;8:97–109. [PMC free article] [PubMed]
39. Kushner SR. mRNA decay in Escherichia coli comes of age. J. Bacteriol. 2002;184:4658–4665. [PMC free article] [PubMed]
40. Hajnsdorf E, Régnier P. E.coli rpsO mRNA decay: RNase E processing at the beginning of the coding sequence stimulates poly(A)-dependent degradation of the mRNA. J. Mol. Biol. 1999;286:1033–1043. [PubMed]
41. Mohanty BK, Kushner SR. The majority of Escherichia coli mRNAs undergo post-transcriptional modification in exponentially growing cells. Nucleic Acids Res. 2006;34:5695–5704. [PMC free article] [PubMed]
42. Pedersen K, Zavialov AV, Pavlov MY, Elf J, Gerdes K, Ehrenberg M. The bacterial toxin relE displays codon-specific cleavage of mRNAs in the ribosomal A site. Cell. 2003;112:131–140. [PubMed]
43. Christensen SK, Gerdes K. RelE toxins from Bacteria and Archaea cleave mRNAs on translating ribosomes, which are rescued by tmRNA. Mol. Microbiol. 2003;48:1389–1400. [PubMed]
44. Sunohara T, Jojima K, Tagami H, Inada T, Aiba H. Ribosome stalling during translation elongation induces cleavage of mRNA being translated in. Escherichia coli. J. Biol. Chem. 2004;279:15368–15375. [PubMed]
45. Yamanishi H, Yonesaki T. RNA cleavage linked with ribosomal action. Genetics. 2005;171:419–425. [PMC free article] [PubMed]
46. Mackie GA. Ribonuclease E is a 5′-end-dependent endonuclease. Nature. 1998;395:720–723. [PubMed]
47. Badet B, Vermoote P, Haumont PY, Lederer F, LeGoffic F. Glucosamine synthetase from Escherichia coli: purification, properties, and glutamine-utilizing site location. Biochemistry. 1987;26:1940–1948. [PubMed]
48. Raynal LC, Krisch HM, Carpousis AJ. The Bacillus subtilis nucleotidyltransferase is a tRNA CCA-adding enzyme. J. Bacteriol. 1998;180:6276–6282. [PMC free article] [PubMed]
49. Condon C. RNA processing and degradation in Bacillus subtilis. Microbiol. Mol. Biol. Rev. 2003;67:157–174. [PMC free article] [PubMed]
50. Winkler WC, Nahvi A, Roth A, Collins JA, Breaker RR. Control of gene expression by a natural metabolite-responsive ribozyme. Nature. 2004;428:281–286. [PubMed]
51. Le Derout J, Régnier P, Hajnsdorf E. Both temperature and medium composition regulate RNase E processing efficiency of the rpsO mRNA coding for ribosomal protein S15 of Escherichia coli. J. Mol. Biol. 2002;319:341–349. [PubMed]
52. Goldblum K, Apirion D. Inactivation of the ribonucleic acid-processing enzyme RNase E blocks cell division. J. Bacteriol. 1981;146:128–132. [PMC free article] [PubMed]
53. Reuven NB, Zhou Z, Deutscher MP. Functional overlap of tRNA nucleotidyltransferase, poly(A)polymearse I and polynucleotide phosphorylase. J. Biol. Chem. 1997;272:33255–33259. [PubMed]
54. Régnier P, Hajnsdorf E. Decay of mRNA encoding ribosomal protein S15 of Escherichia coli is initiated by an RNase E-dependent endonucleolytic cleavage that removes the 3′ stabilizing stem and loop structure. J. Mol. Biol. 1991;217:283–292. [PubMed]
55. Bardey V, Vallet C, Robas N, Charpentier B, Thouvenot B, Mougin A, Hajnsdorf E, Regnier P, Springer M, et al. Characterization of the molecular mechanisms involved in the differential production of erythrose-4-phosphate dehydrogenase, 3-phosphoglycerate kinase and class II fructose-1,6-bisphosphate aldolase in Escherichia coli. Mol. Microbiol. 2005;57:1265–1287. [PubMed]
56. Abe H, Abo T, Aiba H. Regulation of intrinsic terminator by translation in Escherichia coli: transcription termination at a distance downstream. Genes to Cells. 1999;4:87–97. [PubMed]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene
    Gene records that cite the current articles. Citations in Gene are added manually by NCBI or imported from outside public resources.
  • GEO Profiles
    GEO Profiles
    Gene Expression Omnibus (GEO) Profiles of molecular abundance data. The current articles are references on the Gene record associated with the GEO profile.
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...