Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Dec 1.
Published in final edited form as: Mol Microbiol. 2012 Oct 9;86(5):1063–1072. doi: 10.1111/mmi.12040

Influence of translation on RppH-dependent mRNA degradation in Escherichia coli

Jamie Richards 1, Daniel J Luciano 1, Joel G Belasco 1,*
PMCID: PMC3508308  NIHMSID: NIHMS407907  PMID: 22989003

SUMMARY

In Escherichia coli, the endonuclease RNase E can access internal cleavage sites in mRNA either directly or by a 5′-end-dependent mechanism in which cleavage is facilitated by prior RppH-catalyzed conversion of the 5′-terminal triphosphate to a monophosphate, to which RNase E can bind. The characteristics of transcripts that determine which of these two pathways is primarily responsible for their decay are poorly understood. Here we report the influence of ribosome binding and translocation on each pathway, using yeiP and trxB as model transcripts. Ribosome binding to the translation initiation site impedes degradation by both mechanisms. However, because the effect on the rate of 5′-end-independent decay is greater, poor ribosome binding favors degradation by that pathway. Arresting translation elongation with chloramphenicol quickly inhibits RNase E cleavage downstream of the initiation codon but has little or no immediate effect on cleavage upstream of the ribosome binding site. RNase E binding to a monophosphorylated 5′ end appears to increase the likelihood of cleavage at sites within the 5′ untranslated region. These findings indicate that ribosome binding and translocation can have a major impact on 5′-end-dependent mRNA degradation in E. coli and suggest a possible sequence of events that follow pyrophosphate removal.

Keywords: RNase E, ribosome, mRNA decay, mRNA stability, chloramphenicol

INTRODUCTION

Messenger RNA degradation is an important mechanism for regulating gene expression in all organisms. Within a single cell, the lifetimes of mRNAs can differ from one another by as much as two orders of magnitude (from seconds to one hour in bacteria, from minutes to days in higher eukaryotes), with proportional effects on protein synthesis. These differences in stability are governed by features of RNA sequence and structure, some of which have been empirically defined but many of which are still unknown (Belasco, 2010).

The mechanisms by which transcripts are degraded depends on the cellular enzymes that are available. Escherichia coli appears to lack a 5′ exoribonuclease, and degradation by its 3′ exonucleases is impeded by the stem-loop structures generally present at the 3′ end of mRNAs. Therefore, most E. coli transcripts are thought to be degraded by a mechanism in which repeated endonucleolytic cleavage generates RNA fragments that can then undergo rapid 3′-exonucleolytic degradation because they are no longer protected by a 3′-terminal stem-loop (Belasco, 2010).

The endonuclease that is most important for mRNA degradation in E. coli is RNase E, whose inactivation stabilizes most messages (Ono & Kuwano, 1979, Mudd et al., 1990, Babitzke & Kushner, 1991, Melefors & von Gabain, 1991, Taraseviciene et al., 1991). RNase E cuts RNA within single-stranded regions that are AU-rich (McDowall et al., 1994), a loosely defined specificity that enables it to cleave most mRNAs at several sites. Of great consequence for the substrate specificity of this ribonuclease is its marked preference for RNA substrates that bear a single phosphate group at the 5′ terminus (Mackie, 1998). This property, which results from the ability of RNase E to selectively bind monophosphorylated 5′ ends in a discrete pocket on its surface (Callaghan et al., 2005), enables the enzyme to rapidly fragment the monophosphorylated 3′-terminal cleavage products that are produced when it cleaves mRNA (Spickler et al., 2001).

In E. coli, RNase E can gain access to its initial cleavage site(s) in primary transcripts either directly or by a 5′-end-dependent mechanism (Belasco, 2010). Direct access does not involve interaction with the 5′ terminus, whereas 5′-end-dependent access requires prior conversion of the 5′-terminal triphosphate to a monophosphate by the RNA pyrophosphohydrolase RppH (Celesnik et al., 2007, Deana et al., 2008). No other mechanism for the 5′-end-dependent degradation of primary transcripts by RNase E has been identified. Hundreds of E. coli transcripts are degraded by the 5′-end-dependent pathway, as evidenced by their increased concentration and lifetime in cells lacking RppH activity, but many others appear to be degraded primarily by an RppH-independent mechanism (Deana et al., 2008). A distinguishing characteristic of some members of the latter group of mRNAs is a 5′-terminal stem-loop structure (Emory et al., 1992, Bouvet & Belasco, 1992, Bricker & Belasco, 1999, Baker & Mackie, 2003), which blocks 5′-end-dependent degradation by hindering both pyrophosphate removal by RppH and 5′-monophosphate-stimulated cleavage by RNase E (Mackie, 1998, Deana et al., 2008). Little else is known about the features of E. coli mRNAs that determine whether their degradation is 5′-end-dependent.

To learn more about the characteristics of E. coli transcripts that govern whether they decay by an RppH-dependent or direct-access mechanism, we have examined the influence of ribosomes on degradation by each of these pathways. Although previous studies have established a clear link between translation and mRNA degradation in E. coli (reviewed in Deana & Belasco, 2005, Dreyfus, 2009), those investigations were conducted before the mechanism of 5′-end-dependent degradation was elucidated, and none compared the impact of ribosome binding on the degradation of the same transcript by each of the two RNase E pathways or distinguished between effects on pyrophosphate removal and RNase E cleavage. Our results indicate that ribosome binding and translation affect both mechanisms of decay but with a differential influence that can affect the relative utilization of the two pathways.

RESULTS

Mutations that affect ribosome binding

As an initial model system for examining the influence of ribosomes on mRNA degradation by the RppH-dependent and direct-access pathways, we chose the monocistronic transcript of the E. coli yeiP gene, which encodes a paralog of the translation elongation factor EF-P. Previously, we have shown that deletion of the rppH gene markedly reduces the percentage of yeiP transcripts that are monophosphorylated while greatly prolonging the lifetime of this message, effects characteristic of degradation by an RppH-dependent mechanism (Deana et al., 2008). That the half-life of yeiP mRNA also increases upon RNase E inactivation (Deana et al., 2008) indicates that this ribonuclease degrades the monophosphorylated decay intermediate that results from pyrophosphate removal by RppH, possibly with some assistance from the low-abundance RNase E paralog RNase G (Lee et al., 2002).

To alter ribosome binding and hence the efficiency of translation initiation, mutations were introduced into the Shine-Dalgarno (SD) element and initiation codon of a plasmid-borne yeiP gene (Figure 1A). Specifically, the complementarity of the SD element to 16S ribosomal RNA was either increased (SDup: AGGA → UAAGGAGG) or decreased (SDdown: AGGA → AGUA), or the canonical AUG initiation codon was changed to CUG. The effect of each mutation on translation was tested by examining the expression of these variants when the 5′ untranslated region (UTR) and first 20 codons were fused in-frame to lacZ. As expected, cellular β-galactosidase levels were higher for the SDup mutant and much lower for the SDdown or CUG mutant than for wild-type (Figure 1B).

Figure 1. Effect of mutations on the translation and decay of yeiP and mini yeiP mRNA.

Figure 1

Open in a new tab

A. Sequence of the yeiP 5′ UTR. The initiation codon and the region encompassing the SD element are underlined. The mutations introduced into these elements (SDup, SDdown, CUG) are shown below the wild-type sequence. Arrows mark sites of RNase E cleavage (W, X, Y, and Z). AUUU tetramers are in boldface.

B. Effect of mutations on yeiP-lacZ expression, as determined by measuring β-galactosidase activity. Bars and error bars represent mean values and standard deviations, respectively. WT, wild-type.

C. Decay of yeiP, yeiP-SDup, yeiP-SDdown, and yeiP-CUG mRNA in wild-type and ΔrppH cells following rifampicin addition, as monitored by Northern blotting. Plasmid-encoded transcripts were examined in host cells that lacked a chromosomal copy of the yeiP gene.

D. Effect of mutations on the half-life of yeiP and mini yeiP mRNA in wild-type and ΔrppH cells. Bars and error bars represent mean values and standard deviations, respectively. Note the difference in scale of the y-axes. For numerical values, see Table S1.

E. RppH-dependent cleavage within the yeiP 5′ UTR. yeiP mRNA from wild-type (WT) or ΔrppH (Δ) cells that contained or lacked a yeiP gene or yeiP-SDup mRNA from wild-type or ΔrppH cells that lacked a yeiP gene was examined by Northern blot analysis after site-specific DNAzyme cleavage 169 nt from the 5′ end of the 0.65-kb transcript. The full-length transcript (FL) and decay intermediates resulting from RNase E cleavage at sites W, X, Y, and Z are marked. The SDup mutation inhibited cleavage at site Z by altering its sequence (AUUU → GGUU).

F. Decay of the full-length yeiP transcript in wild-type and ΔrppH cells, as monitored by Northern blot analysis after DNAzyme cleavage. The full-length transcript (FL) and decay intermediates W, X, Y, and Z are marked.

Effects of ribosome binding on decay rates and 5′ UTR cleavage

To ascertain the effect of ribosome binding on the degradation of yeiP mRNA by the 5′-end-dependent and 5′-end-independent pathways, the degradation of each of the plasmid-encoded yeiP mRNA variants was monitored in isogenic wild-type and ΔrppH cells lacking a chromosomal copy of the yeiP gene. This was achieved by inhibiting transcription with rifampicin, extracting total cellular RNA at time intervals, and analyzing an equal amount of each RNA sample by Northern blotting (Figure 1C). In the presence of RppH, the wild-type yeiP transcript decayed with a half-life of 1.2 ± 0.1 min. Its half-life increased to 2.9 ± 0.5 min when ribosome binding was optimized (SDup) and fell to 0.4 ± 0.1 min when ribosome binding was greatly impaired (SDdown or CUG). In the absence of RppH, the lifetime of the wild-type transcript was markedly prolonged (half-life of 7.6 ± 0.2 min), as expected for a message ordinarily degraded by a 5′-end-dependent mechanism. The slow 5′-end-independent decay rate of yeiP mRNA in ΔrppH cells was not significantly affected by replacing its already good SD element with an even better one (half-life of 8.4 ± 0.5 min for SDup) but increased 10–15 fold when the SD element or initiation codon was very poor (half-life of 0.8 ± 0.1 or 0.5 ± 0.1 min for the SDdown or CUG mutant, respectively) (Figure 1D). Together, these results indicate that ribosome binding impedes yeiP mRNA degradation by both pathways. Specifically, the observation that the lifetime of the SDup mutant was longer than that of the wild-type transcript in the presence of RppH and increased further in its absence indicates that ribosome binding can hinder 5′-end-dependent degradation, while the diminished longevity of the SDdown and CUG mutants in the absence of RppH shows that ribosome binding can impede 5′-end-independent decay as well.

The 5′-terminal segment of yeiP was examined more closedly by cleaving the 0.65-kb transcript 169 nt from the 5′ end prior to Northern blot analysis. Accomplished by treating total cellular RNA with a site-specific 10–23 DNAzyme (Santoro & Joyce, 1997), this additional step revealed four low-abundance decay intermediates (W, X, Y, and Z) that were slightly shorter than the full-length transcript (Figure 1E). Primer extension analysis showed that these intermediates had been produced by cleavage within the 5′ untranslated region (UTR), either upstream (W, X, and Y) or immediately downstream (Z) of the SD element (Figure S1), and that band X was a closely spaced doublet. Three of these cleavage sites (W, X, and Z) each corresponded to a distinct AUUU tetramer (Figure 1A), a sequence characteristic of several known sites of RNase E cleavage (Ehretsmann et al., 1992). Thermal inactivation of RNase E diminished the abundance of all four intermediates (Figure S2), confirming its role in their origin. Cleavage at these 5′ UTR sites was also strongly RppH-dependent, as evidenced by the much higher concentration of decay intermediates W, X, Y, and Z in wild-type cells than in ΔrppH cells (Figure 1E). Relative to their full-length precursor, these intermediates were at least 13 times more abundant in the presence of RppH, a ratio that exceeded the 6-fold increase in the yeiP decay rate (half-life ratio of 7.6 min/1.2 min). This difference was even greater for yeiP-SDup: a ≥15-fold increase in the relative concentration of the 5′ UTR cleavage products versus a 3-fold increase in the rate of decay (half-life ratio of 8.4 min/2.9 min). Cleavage within the coding region was less sensitive to the presence of RppH than was cleavage within the 5′ UTR (Figure 1E and data not shown).

To exclude the possibility that these previously unresolved cleavage products had distorted the original half-life measurements, we repeated those experiments with DNAzyme-treated RNA samples. The improved electrophoretic resolution afforded by this approach made it possible to monitor the decay of the full-length yeiP transcript selectively (Figures 1F and S3). To within experimental error, the yeiP half-lives determined in this manner were the same as those measured without DNAzyme cleavage, allaying our concern.

Because yeiP encodes a translation factor, we also considered whether its translation might be subject to feedback regulation by its protein product, which could influence its decay. To rule out the possibility that the degradation of the yeiP mutants might have been affected by changes in the cellular concentration of the YeiP protein, we repeated those half-life measurements in the context of an in-frame deletion mutant (mini yeiP) that lacked codons 23–164 and therefore did not encode a functional protein product. Like its full-length counterpart, which contains 190 codons, the mini yeiP transcript was degraded by an RppH-dependent mechanism; in the absence of RppH, its half-life increased substantially. Moreover, the effect of mutations in the SD element and initiation codon on mini yeiP mRNA decay was similar to that observed for the full-length transcript, in both wild-type and ΔrppH cells (Figure 1D). These findings indicate that the degradation of yeiP mRNA is insensitive to YeiP protein levels.

To determine whether the effects of ribosome-binding mutations depend on their sequence context, the same mutations were introduced into an entirely different mRNA, trxB, which encodes thioredoxin reductase (Figure 2A). Like yeiP, the monocistronic trxB transcript is a target of both RppH (Figure 2B) and RNase E (Deana et al., 2008), but unlike yeiP, no decay intermediates resulting from cleavage within the short trxB 5′ UTR were evident (Figures 2C and S1), an observation consistent with the absence there of a sequence element resembling an RNase E cleavage site. The SDup, SDdown, and CUG mutations each had the expected effect on expression of a trxB-lacZ reporter (Figure 2D). Furthermore, in either the presence or absence of RppH, the influence of these mutations on trxB mRNA stability correlated well with their influence on ribosome binding (Figure 2E). Once again, optimizing the SD element impeded RppH-dependent degradation, and a very poor ribosome binding site greatly accelerated RppH-independent decay. Thus, it appears that the ability of ribosomes to inhibit both 5′-end-dependent and 5′-end-independent degradation is pertinent to the decay of a variety of E. coli mRNAs that are targeted by RppH.

Figure 2. Effect of mutations on the translation and decay of trxB mRNA.

Figure 2

Open in a new tab

A. Sequence of the trxB 5′ UTR. The initiation codon and the region encompassing the SD element are underlined. The mutations introduced into these elements are shown below the wild-type sequence.

B. Decay of wild-type trxB mRNA in wild-type and ΔrppH cells following rifampicin addition, as monitored by Northern blotting.

C. Absence of detectable cleavage within the trxB 5′ UTR. Wild-type trxB mRNA from wild-type (WT) or ΔrppH (Δ) cells that contained or lacked a trxB gene was examined by Northern blot analysis after site-specific DNAzyme cleavage 294 nt from the 5′ end of the 1.06-kb transcript. The full-length transcript (FL) and a decay intermediate resulting from cleavage within the coding region (*) are marked.

D. Effect of mutations on trxB-lacZ expression. Bars and error bars represent mean values and standard deviations, respectively. WT, wild-type.

E. Effect of mutations on the half-life of trxB mRNA in wild-type and ΔrppH cells. Plasmid-encoded transcripts were examined in host cells that lacked a chromosomal copy of the trxB gene. Bars and error bars represent mean values and standard deviations, respectively. For numerical values, see Table S1.

Effects of ribosome binding on 5′ phosphorylation state

RppH-dependent degradation is a multistep process in which pyrophosphate removal from a triphosphorylated transcript generates a monophosphorylated intermediate that is then cleaved by RNase E. The relative rates of these sequential events is reflected in the ratio of monophosphorylated to triphosphorylated RNA, such that kPP/kE = monoP/triP, where monoP is the steady-state concentration of the full-length monophosphorylated intermediate, triP is the steady-state concentration of its triphosphorylated precursor, and kPP and kE are the rate constants for formation and breakdown of the monophosphorylated intermediate, respectively (i.e., for pyrophosphate removal and subsequent RNase E cleavage). Therefore, any differential effect of ribosome binding on the rates of these two steps should be manifested as a change in the ratio of monophosphorylated to triphosphorylated RNA.

To address this point, we used PABLO analysis to measure the percentage of each of the yeiP and trxB ribosome binding mutants that is monophosphorylated in cells containing RppH. This splinted ligation assay makes use of the ability of T4 DNA ligase to covalently join a DNA oligonucleotide to the 5′ end of monophosphorylated RNA, but not triphosphorylated RNA, when the two are juxtaposed by simultaneous base pairing to a bridging DNA oligonucleotide (Celesnik et al., 2007). Following DNAzyme cleavage, the ligation product and its unligated progenitor were separated by gel electrophoresis and quantified by Northern blotting. As a normalization control, the same RNAs were analyzed by PABLO after treatment with tobacco acid pyrophosphatase (TAP) to convert their 5′ ends entirely to ligatable monophosphates. The percentage of 5′ ends that were monophosphorylated was then calculated from the ratio of the ligation yields before and after TAP treatment.

At steady state, 43 ± 2% of wild-type yeiP and 51 ± 5% of wild-type trxB were observed to be monophosphorylated (Figure 3A), indicating that kPP is approximately equal to kE for both transcripts, i.e., that each of the initial steps in the 5′-end-dependent degradation of these mRNAs is partially rate-determining. In the case of trxB, a mutation that improved ribosome binding (SDup) significantly increased the percentage of transcripts that were monophosphorylated, whereas mutations that impaired ribosome binding (SDdown, CUG) had the opposite effect (Figure 3B). Together with its prolonged lifetime, the increased monophosphorylation of the trxB-SDup mutant indicates that ribosome binding impedes the 5′-end-dependent degradation of trxB mRNA primarily by hindering RNase E cleavage (kE) rather than by slowing pyrophosphate removal (kPP). By contrast, the same mutation had little effect on the phosphorylation state of yeiP mRNA despite significantly reducing its decay rate, making it likely that ribosome binding to that transcript impedes both degradative steps.

Figure 3. Effect of mutations on the phosphorylation state of yeiP and trxB mRNA.

Figure 3

Open in a new tab

A. Phosphorylation state of the wild-type yeiP and trxB transcripts, as determined by PABLO analysis of total cellular RNA extracted from wild-type cells. As controls, the transcripts were also examined after extraction from ΔrppH cells, without ligation, and after TAP treatment.

B. Effect of mutations on the phosphorylation state of yeiP and trxB mRNA in wild-type cells. Bars and error bars represent mean values and standard deviations, respectively. WT, wild-type mRNA. For numerical values, see Table S1.

Effect of ribosome movement

Translation is a dynamic process in which the position of bound ribosomes is constantly changing. To gain insight into the consequences of ribosomal movement for 5′-end-dependent degradation, we monitored the decay of yeiP and trxB mRNA after treating E. coli with chloramphenicol to arrest translation elongation. Rifampicin was added just two minutes after chloramphenicol to minimize any possible indirect effects of inhibiting protein synthesis.

At first glance, chloramphenicol seemed to retard the degradation of the yeiP transcript in RppH-containing cells (data not shown). However, closer examination of the 5′-terminal segment of yeiP, achieved by site-specific DNAzyme cleavage to improve electrophoretic resolution, revealed that chloramphenicol-induced translational arrest had no effect on the half-life of full-length yeiP mRNA; instead, it slowed the decay of intermediates X and Z (but not W), allowing them to accumulate to a high concentration (Figure 4). Decay intermediates with 5′ ends within the protein-coding region were also detectable but were less abundant (Figure 4A and data not shown).

Figure 4. Effect of chloramphenicol-induced translational arrest on the decay of yeiP and trxB mRNA.

Figure 4

Open in a new tab

A. Decay of wild-type yeiP and trxB mRNA in the presence of chloramphenicol. Two minutes after chloramphenicol addition, transcription in wild-type (WT) or ΔrppH cells was blocked with rifampicin, and total RNA was extracted at time intervals and analyzed by Northern blotting, with (yeiP) or without (trxB) prior cleavage with a transcript-specific DNAzyme. The full-length transcripts (FL) and decay intermediates resulting from cleavage within the yeiP 5′ UTR (W, X, Y, and Z) are marked.

B. Effect of translational arrest on the decay rate of the full-length yeiP and trxB transcripts in wild-type and ΔrppH cells. The intensities of the FL bands in panel A were plotted semilogarithmically as a function of time along with the corresponding data from cells that had not been treated with chloramphenicol (Figures 1F and 2B). Best-fit lines were calculated by linear regression. Cm, chloramphenicol. For half-lives, see Table S2.

Unlike the full-length yeiP transcript, full-length trxB mRNA was stabilized 3.5-fold by chloramphenicol (Figure 4). Although trxB decay intermediates resulting from cleavage within the coding region were faintly visible in the presence of chloramphenicol, no cleavage in the 5′ UTR was evident (data not shown).

Chloramphenicol had a similar differential effect on mRNA decay in cells lacking RppH, where it hindered the degradation of trxB mRNA by a factor of 2 but left the lifetime of full-length yeiP unchanged (Figure 4). Once again, decay intermediates resulting from 5′ UTR cleavage were evident for yeiP but not for trxB. These yeiP intermediates accumulated more slowly than in wild-type cells, consistent with the slow decay of their full-length precursor when RppH is absent. Thus, in either the presence or absence of RppH, arresting translation elongation appears to hinder RNase E cleavage within the coding region but not within the 5′ UTR upstream of the ribosome binding site.

DISCUSSION

These studies of the influence of translation on bacterial mRNA decay show that ribosome binding can hinder early events in degradation by both the 5′-end-dependent and 5′-end-independent pathway, irrespective of the potential for RNase E to cleave well upstream of the SD element. In principle, this protective effect could be a consequence of increasing the occupancy of the ribosome binding site, the efficiency of translation initiation, or both. By contrast, chloramphenicol-induced arrest of translation elongation impedes RNase E cleavage within the coding region but has no immediate effect on cleavage upstream of the ribosome binding site.

Although diminishing the affinity of mRNA for ribosomes can expedite degradation by both pathways, ribosome binding that is especially poor appears to favor 5′-end-independent decay due to its greater effect on that pathway. Thus, whereas pyrophosphate removal by RppH accelerates the decay of wild-type yeiP and trxB mRNA by a factor of 3–6, it has little if any effect on the decay rate of yeiP and trxB mutants with a weak SD element or a poor initiation codon, owing to the vulnerability of the mutant transcripts to rapid degradation by the RppH-independent pathway. The influence of weak ribosome binding on the rate of RppH-independent yeiP and trxB degradation is consistent with previous reports of its effect on the decay of transcripts bearing a 5′-terminal stem-loop and circular mRNA (Arnold et al., 1998, Mackie, 2000, Baker & Mackie, 2003). Conversely, a mutation that optimizes ribosome binding impedes degradation of the yeiP and trxB transcripts by the RppH-dependent pathway but has little effect on their RppH-independent decay, which is already quite slow.

Decay intermediates produced by RNase E cleavage within the yeiP 5′ UTR are much more abundant in wild-type cells than in cells lacking RppH, increasing ≥13-fold relative to the concentration of their full-length precursor for yeiP and ≥15-fold for yeiP-SDup. By comparison, the overall rate at which that precursor is cleaved to generate all possible decay intermediates increases only 6-fold for yeiP and 3-fold for yeiP-SDup. These differences imply that pyrophosphate removal by RppH makes it more likely that the 5′ UTR rather than another segment will subsequently be cut, especially when ribosome binding is very strong, as the decay rates of all of the resulting 3′-terminal cleavage products are expected to be RppH-independent. Evidently, RNase E binding to a monophosphorylated 5′ end disproportionately enhances cleavage at sites within the 5′ UTR, presumably due to their greater proximity and lesser ribosomal protection.

In principle, ribosome binding could inhibit 5′-end-dependent degradation by hindering pyrophosphate removal by RppH or subsequent cleavage by RNase E. The correlation between ribosome binding affinity and the percentage of trxB messages that are monophosphorylated at steady state implies that improved ribosome binding slows trxB-SDup degradation mainly by hindering RNase E cleavage rather than by impeding pyrophosphate removal. By contrast, the ability of the SDup mutation to inhibit the degradation of yeiP mRNA without altering its phosphorylation state indicates that tighter ribosome binding impedes both of the initial steps in yeiP decay. That better ribosome binding can hinder cleavage near the SD element or within the coding region is not unexpected; however, its apparent inhibitory effect on pyrophosphate removal from the 5′ terminus of yeiP mRNA and on RNase E cleavage at yeiP sites W and X is more surprising in view of their distance from the ribosome binding site (see Figures 1A and 2A). One possible explanation is that the affinity of ribosomal protein S1 for AU-rich RNA (Boni et al., 1991, Komarova et al., 2002) may enable it to bind the 5′-terminal segment of the yeiP 5′ UTR and protect it from both RppH and RNase E. Consistent with this hypothesis, β-galactosidase synthesis from a yeiP-lacZ reporter is greatly reduced by deletion of the AU-rich segment comprising nucleotides #5–19 of the yeiP 5′ UTR, including sites W and X (data not shown).

5′-end-dependent mRNA degradation can also be impeded by blocking translation elongation, but only if there are no RNase E cleavage sites upstream of the ribosome binding site. Thus trxB mRNA, whose 5′ UTR appears to lack such sites, is stabilized by chloramphenicol, whereas the full-length yeiP transcript, which has a number of RNase E cleavage sites in its 5′ UTR, is not. Even in the presence of chloramphenicol, the yeiP 5′ UTR undergoes facile, 5′-monophosphate-assisted RNase E cleavage at those sites to generate three prominent decay intermediates. One (W), which is only five nucleotides shorter than the full-length transcript, is short-lived, whereas two others, with 5′ ends several nucleotides upstream (X) or immediately downstream (Z) of the SD element, decay slowly in chloramphenicol-treated cells even though both are monophosphorylated and one (X) contains RNase E cleavage sites in its 5′ UTR. We propose that intermediates X and Z are each generated from a distinct pool of translationally arrested yeiP transcripts, some of which have a ribosome stalled so close to the translation initiation site as to block access to sites Y and Z and others of which have a ribosome stalled far enough downstream to expose site Z without leaving room for another ribosome to bind.

The ability of chloramphenicol to prolong the lifetimes of trxB mRNA and yeiP intermediates X and Z but not of full-length yeiP or intermediate W suggests that arresting translation elongation hinders RNase E cleavage in or near the coding region but does not interfere with cleavage upstream of the ribosome binding site. This regioselective influence of chloramphenicol on ribonuclease access also enables it to impede the 5′-end-independent degradation of trxB mRNA but not of full-length yeiP mRNA in cells lacking RppH. The distinct effects of chloramphenicol treatment on the degradation of yeiP and trxB mRNA and its regioselective impact on the various yeiP decay intermediates indicate that its influence on these messages immediately after its addition is a direct consequence of mRNA protection by static ribosomes and not an indirect result of RNase E titration by excess RNA (Lopez et al., 1998).

The protective effect of bound ribosomes, especially when they are static, suggests a possible sequence of events during the normal course of 5′-end-dependent mRNA degradation in E. coli (Figure 5). Following pyrophosphate removal by RppH, we propose that RNase E engages the monophosphorylated 5′ terminus and searches nearby for a cleavage site that is not occluded by a bound ribosome, with a preference for those that do not require it to reach around intervening ribosomes (Dreyfus, 2009). When the ribosome binding site is occupied, these constraints favor cleavage upstream of the SD element (provided that a good RNase E cleavage site is present there), generating another monophosphorylated intermediate. Once the initiating ribosome moves forward from the start codon and begins translation, the vacated RNA segment behind it becomes accessible to RNase E until the next ribosome binds. Cleavage there severs the ribosome binding site and prevents further translation initiation. A fraction of a minute thereafter, the ribosomes already translating the message reach the termination codon and dissociate, leaving the RNA entirely unprotected and vulnerable to swift degradation by RNase E.

Figure 5. Model of 5′-end-dependent mRNA degradation in E. coli and the influence of ribosome translocation.

Figure 5

Open in a new tab

RppH (hatchet) removes pyrophosphate from the 5′ end of a triphosphorylated primary transcript. Now deprotected, the resulting monophosphorylated decay intermediate is rapidly cut by RNase E (scissors) either in the 5′ UTR (thin line) or coding region (cylinder), depending on the location of suitable cleavage sites and the occupancy of the ribosome binding site. If RNase E bound to the monophosphorylated 5′ end and searching for a cleavage site has difficulty bypassing an intervening ribosome, an occupied ribosome binding site would favor cleavage in the 5′ UTR, while an unoccupied ribosome binding site would facilitate cleavage in the coding region. By preventing further translation initiation, severing the ribosome binding site would soon lead to a complete loss of ribosomal protection and thereby trigger rapid degradation.

Previous studies have shown that 5′-terminal base pairing prevents RNA degradation via the 5′-end-dependent pathway by inhibiting both pyrophosphate removal and monophosphate-assisted RNase E cleavage (Mackie, 1998, Deana et al., 2008). Such transcripts must instead be degraded by a 5′-end-independent mechanism. Here we have shown that the relative utilization of these two pathways can also be influenced by ribosome binding, the contribution of the 5′-end-dependent mechanism being greatly diminished when the binding affinity of ribosomes is very weak. Further investigations can be expected to reveal additional features that govern the susceptibility of mRNA to RppH-mediated degradation.

EXPERIMENTAL PROCEDURES

Strains and plasmids

Measurements of the lifetime and phosphorylation state of mRNA were performed in E. coli K-12 strain BW25113 and derivatives bearing an in-frame deletion of the yeiP, trxB, or rppH coding region (Baba et al., 2006). Combinations of these deletions were generated by P1-mediated transduction. The role of RNase E in cleaving the yeiP 5′ UTR was examined by using E. coli strain N3431 (temperature-sensitive rne-3071 allele) and its isogenic counterpart N3433 (wild-type rne allele) (Goldblum & Apirion, 1981).

Plasmid pYeiP1 was constructed by inserting the entire E. coli yeiP gene (including 128 bp upstream and 82 bp downstream of the coding region) between the EcoRI and PstI sites of plasmid pBR322fd, a derivative of pBR322 in which a 0.35-kb HindIII fragment of pLBU1 containing the bacteriophage fd terminator (Gentz et al., 1981) had been inserted at the HindIII site so as to block transcription of the bla gene from promoters upstream of its natural promoter. Plasmid pYeiP1mini was constructed from pYeiP1 by removing 420 bp from the coding region of yeiP so as to generate an in-frame fusion of codon 23 to codon 164. Plasmid pTrxB1 was constructed by inserting the entire E. coli trxB gene (including 124 bp upstream and 122 bp downstream of the coding region) between the EcoRI and PstI sites of plasmid pBR322fd. Plasmids pYeiP-LacZ and pTrxB-LacZ were derivatives of pPM30 (Meacock & Cohen, 1980) in which the promoter, 5′ UTR, and first 20 codons of yeiP or trxB was fused in-frame to codons 9–1024 of lacZ. The SDup, SDdown and CUG variants of each plasmid were identical to the parental plasmid except for modifications to the SD element or initiation codon.

Measurement of mRNA lifetimes

To measure rates of mRNA decay, E. coli cells were grown to mid-log phase at 37°C in defined MOPS medium containing 0.2% (w/v) glucose, transcription was inhibited with rifampicin (0.2 mg ml−1), and total cellular RNA was extracted at time intervals (Celesnik et al., 2008). To inhibit translation elongation, chloramphenicol (100 μg ml−1) was added 2 min before rifampicin addition. Equal amounts of each RNA sample were then subjected to gel electrophoresis on 6% or 4.5% polyacrylamide containing 8 M urea or 1.2% agarose containing 2.4% formaldehyde. RNA was transferred to a Hybond-XL membrane (GE Healthcare) by electroblotting for polyacrylamide gels or by overnight capillary transfer for agarose gels. The yeiP, mini yeiP and trxB transcripts were detected by probing the blots with complementary 5′-end-labeled oligodeoxynucleotides (Table S3). Radioactive bands were visualized with a Storm 820 PhosphorImager (Molecular Dynamics), and the band intensities were quantified by using ImageQuant software (Molecular Dynamics). RNA half-lives were calculated by linear regression analysis.

DNAzyme cleavage of mRNA

Total RNA from E. coli (10μg) was combined in a final volume of 36 μl with 400 pmoles of a 10–23 DNAzyme (Santoro & Joyce, 1997) designed to cleave the full-length yeiP transcript within codon 45 (YeiP-DZM3) or the trxB transcript within codon 89 (TrxB-DZM2) (Table S3). The mixtures were heated to 85°C for 5 min, slowly cooled to 30°C, and chilled on ice for at least 1 min. T4 DNA ligase buffer (New England Biolabs) was added to a final concentration of 50 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 10 mM dithiothreitol, and 1 mM ATP, and the cleavage reactions were allowed to proceed at 37°C for 4 hr. The reaction was stopped by EDTA addition and phenol-chloroform extraction, and the reaction products were recovered by ethanol precipitation and analyzed by Northern blotting.

PABLO analysis

Samples of total cellular RNA (20 μg) were analyzed by PABLO as described previously (Celesnik et al., 2008). PABLO analysis of full-length yeiP mRNA was performed with oligonucleotides YyeiP and X32 and YeiP-DZM3. PABLO analysis of mini yeiP mRNA made use of oligonucleotides YyeiP and X32 but omitted the YeiP DNAzyme. PABLO analysis of the trxBWT and trxBCUG transcripts was performed with oligonucleotides YtrxB and X32 and TrxB-DZM2. For the trxBSDdown and trxBSDup transcripts, YtrxB was replaced with YtrxBdown or YtrxBup, respectively. In each case, the PABLO ligation products were detected by Northern blot analysis with a transcript-specific probe. For oligonucleotide sequences, see Table S3.

The percentage of a full-length transcript that was monophosphorylated was calculated by dividing its PABLO ligation yield by that for its fully monophosphorylated counterpart. All such measurements were performed in triplicate to obtain a mean and standard deviation. Fully monophosphorylated yeiP, mini yeiP, and trxB mRNA was generated by treating total cellular RNA (20 μg) with tobacco acid pyrophosphatase (TAP, 2 units, Epicentre Biotechnologies) in 50 mM sodium acetate (pH 6.0), 1 mM EDTA, 0.1% β-mercaptoethanol, and 0.01% Triton X-100 for 2 hr at 37°C. The reaction was stopped by EDTA addition and phenol-chloroform extraction, and the reaction products were recovered by ethanol precipitation.

Characterization of decay intermediates

The 5′ ends of yeiP and trxB decay intermediates were mapped by primer extension beside cognate sequence ladders. Total cellular RNA (1–5 μg) was treated with TURBO DNase (Ambion) and then reverse transcribed with SuperScript III (Invitrogen) and a 5′-end-labeled transcript-specific primer (YeiP-Rev or TrxB-Rev (Table S3)) according to the manufacturer instructions. Sequence ladders were generated by primer extension on a yeiP or trxB DNA template in the presence of a 2′, 3′-dideoxynucleoside triphosphate.

To determine whether cleavage of the yeiP 5′ UTR was RNase E-dependent, E. coli strains N3431 and N3433 bearing plasmid pYeiP1 were grown to mid-log phase at 30°C in MOPS medium containing 0.2% (w/v) glucose and 0.08% (w/v) casamino acids, and total cellular RNA was extracted before and after shifting the culture temperature to 44°C for 10 min. Equal amounts of each RNA sample were then cleaved with a yeiP-specific DNAzyme and analyzed by Northern blotting with a yeiP-specific probe.

β-galactosidase assays

E. coli BW25113 cells containing a plasmid-borne yeiP-lacZ or trxB-lacZ fusion were grown to mid-log phase at 37°C in defined MOPS medium containing 0.2% glucose (w/v), and the β-galactosidase activity of a 0.2-ml culture sample was determined spectrophotometrically as previously described (Miller, 1992). All such measurements were performed in triplicate to obtain a mean and standard deviation.

Supplementary Material

Supp Figure S1-S3&Table S1-S3

Acknowledgments

We are grateful to Eyoel Yemanaberhan for preliminary studies and to Tricia Foley for helpful comments on the manuscript. This investigation was supported by a postdoctoral fellowship to J.R. from the Vilcek Endowment and by a graduate fellowship to D.J.L. (T32AI007180) and a research grant to J.G.B. (R01GM035769) from the National Institutes of Health. The content of this publication is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

References

  1. Arnold TE, Yu J, Belasco JG. mRNA stabilization by the ompA 5′ untranslated region: two protective elements hinder distinct pathways for mRNA degradation. RNA. 1998;4:319–330. [PMC free article] [PubMed] [Google Scholar]
  2. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol. 2006;2:2006.0008. doi: 10.1038/msb4100050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Babitzke P, Kushner SR. The Ams (altered mRNA stability) protein and ribonuclease E are encoded by the same structural gene of Escherichia coli. Proc Natl Acad Sci USA. 1991;88:1–5. doi: 10.1073/pnas.88.1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baker KE, Mackie GA. Ectopic RNase E sites promote bypass of 5′-end-dependent mRNA decay in Escherichia coli. Mol Microbiol. 2003;47:75–88. doi: 10.1046/j.1365-2958.2003.03292.x. [DOI] [PubMed] [Google Scholar]
  5. Belasco JG. All things must pass: contrasts and commonalities in eukaryotic and bacterial mRNA decay. Nat Rev Mol Cell Biol. 2010;11:467–478. doi: 10.1038/nrm2917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boni IV, Isaeva DM, Musychenko ML, Tzareva NV. Ribosome-messenger recognition: mRNA target sites for ribosomal protein S1. Nucleic Acids Res. 1991;19:155–162. doi: 10.1093/nar/19.1.155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bouvet P, Belasco JG. Control of RNase E-mediated RNA degradation by 5′-terminal base pairing in E. coli. Nature. 1992;360:488–491. doi: 10.1038/360488a0. [DOI] [PubMed] [Google Scholar]
  8. Bricker AL, Belasco JG. Importance of a 5′ stem-loop for longevity of papA mRNA in Escherichia coli. J Bacteriol. 1999;181:3587–3590. doi: 10.1128/jb.181.11.3587-3590.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Callaghan AJ, Marcaida MJ, Stead JA, McDowall KJ, Scott WG, Luisi BF. Structure of Escherichia coli RNase E catalytic domain and implications for RNA turnover. Nature. 2005;437:1187–1191. doi: 10.1038/nature04084. [DOI] [PubMed] [Google Scholar]
  10. Celesnik H, Deana A, Belasco JG. Initiation of RNA decay in Escherichia coli by 5′ pyrophosphate removal. Mol Cell. 2007;27:79–90. doi: 10.1016/j.molcel.2007.05.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Celesnik H, Deana A, Belasco JG. PABLO analysis of RNA 5′-phosphorylation state and 5′-end mapping. Methods Enzymol. 2008;447:83–98. doi: 10.1016/S0076-6879(08)02205-2. [DOI] [PubMed] [Google Scholar]
  12. Deana A, Belasco JG. Lost in translation: the influence of ribosomes on bacterial mRNA decay. Genes Dev. 2005;19:2526–2533. doi: 10.1101/gad.1348805. [DOI] [PubMed] [Google Scholar]
  13. Deana A, Celesnik H, Belasco JG. The bacterial enzyme RppH triggers messenger RNA degradation by 5′ pyrophosphate removal. Nature. 2008;451:355–358. doi: 10.1038/nature06475. [DOI] [PubMed] [Google Scholar]
  14. Dreyfus M. Killer and protective ribosomes. Prog Mol Biol Transl Sci. 2009;85:423–466. doi: 10.1016/S0079-6603(08)00811-8. [DOI] [PubMed] [Google Scholar]
  15. Ehretsmann CP, Carpousis AJ, Krisch HM. Specificity of Escherichia coli endoribonuclease RNase E: in vivo and in vitro analysis of mutants in a bacteriophage T4 mRNA processing site. Genes Dev. 1992;6:149–159. doi: 10.1101/gad.6.1.149. [DOI] [PubMed] [Google Scholar]
  16. Emory SA, Bouvet P, Belasco JG. A 5′-terminal stem-loop structure can stabilize mRNA in Escherichia coli. Genes Dev. 1992;6:135–148. doi: 10.1101/gad.6.1.135. [DOI] [PubMed] [Google Scholar]
  17. Gentz R, Langner A, Chang AC, Cohen SN, Bujard H. Cloning and analysis of strong promoters is made possible by the downstream placement of a RNA termination signal. Proc Natl Acad Sci USA. 1981;78:4936–4940. doi: 10.1073/pnas.78.8.4936. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Goldblum K, Apirion D. Inactivation of the ribonucleic acid-processing enzyme ribonuclease E blocks cell division. J Bacteriol. 1981;146:128–132. doi: 10.1128/jb.146.1.128-132.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Komarova AV, Tchufistova LS, Supina EV, Boni IV. Protein S1 counteracts the inhibitory effect of the extended Shine-Dalgarno sequence on translation. RNA. 2002;8:1137–1147. doi: 10.1017/s1355838202029990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lee K, Bernstein JA, Cohen SN. RNase G complementation of rne null mutation identifies functional interrelationships with RNase E in Escherichia coli. Mol Microbiol. 2002;43:1445–1456. doi: 10.1046/j.1365-2958.2002.02848.x. [DOI] [PubMed] [Google Scholar]
  21. Lopez PJ, Marchand I, Yarchuk O, Dreyfus M. Translation inhibitors stabilize Escherichia coli mRNAs independently of ribosome protection. Proc Natl Acad Sci USA. 1998;95:6067–6072. doi: 10.1073/pnas.95.11.6067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mackie GA. Ribonuclease E is a 5′-end-dependent endonuclease. Nature. 1998;395:720–723. doi: 10.1038/27246. [DOI] [PubMed] [Google Scholar]
  23. Mackie GA. Stabilization of circular rpsT mRNA demonstrates the 5′-end dependence of RNase E action in vivo. J Biol Chem. 2000;275:25069–25072. doi: 10.1074/jbc.C000363200. [DOI] [PubMed] [Google Scholar]
  24. McDowall KJ, Lin-Chao S, Cohen SN. A+U content rather than a particular nucleotide order determines the specificity of RNase E cleavage. J Biol Chem. 1994;269:10790–10796. [PubMed] [Google Scholar]
  25. Meacock PA, Cohen SN. Partitioning of bacterial plasmids during cell division: a cis-acting locus that accomplishes stable plasmid inheritance. Cell. 1980;20:529–542. doi: 10.1016/0092-8674(80)90639-x. [DOI] [PubMed] [Google Scholar]
  26. Melefors Ö, von Gabain A. Genetic studies of cleavage-initiated mRNA decay and processing of ribosomal 9S RNA show that the Escherichia coli ams and rne loci are the same. Mol Microbiol. 1991;5:857–864. doi: 10.1111/j.1365-2958.1991.tb00759.x. [DOI] [PubMed] [Google Scholar]
  27. Miller JH. A Short Course in Bacterial Genetics. Cold Spring Harbor Laboratory Press; Cold Spring Harbor, NY: 1992. [Google Scholar]
  28. Mudd EA, Krisch HM, Higgins CF. RNase E, an endoribonuclease, has a general role in the chemical decay of E. coli mRNA: evidence that rne and ams are the same genetic locus. Mol Microbiol. 1990;4:2127–2135. doi: 10.1111/j.1365-2958.1990.tb00574.x. [DOI] [PubMed] [Google Scholar]
  29. Ono M, Kuwano M. A conditional lethal mutation in an Escherichia coli strain with a longer chemical lifetime of mRNA. J Mol Biol. 1979;129:343–357. doi: 10.1016/0022-2836(79)90500-x. [DOI] [PubMed] [Google Scholar]
  30. Santoro SW, Joyce GF. A general purpose RNA-cleaving DNA enzyme. Proc Natl Acad Sci USA. 1997;94:4262–4266. doi: 10.1073/pnas.94.9.4262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Spickler C, Stronge V, Mackie GA. Preferential cleavage of degradative intermediates of rpsT mRNA by the Escherichia coli RNA degradosome. J Bacteriol. 2001;183:1106–1109. doi: 10.1128/JB.183.3.1106-1109.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Taraseviciene L, Miczak A, Apirion D. The gene specifying RNase E (rne) and a gene affecting mRNA stability (ams) are the same gene. Mol Microbiol. 1991;5:851–855. doi: 10.1111/j.1365-2958.1991.tb00758.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp Figure S1-S3&Table S1-S3
NIHMS407907-supplement-Supp_Figure_S1-S3_Table_S1-S3.pdf (323KB, pdf)

RESOURCES