RelE of Escherichia coli is a global inhibitor of translation that is activated by nutritional stress. Activation of RelE depends on Lon-mediated degradation of RelB, the antagonist that neutralizes RelE. In vitro, RelE cleaves synthetic mRNAs positioned at the ribosomal A-site. We show here that in vivo overexpression of RelE confers cleavage of mRNA and tmRNA in their coding regions. RelE-mediated cleavage depended on translation of the RNAs and occurred at both sense and stop codons. RelE cleavage of mRNA and tmRNA was also induced by amino acid starvation. An ssrA deletion strain was hypersensitive to RelE, whereas overproduction of tmRNA counteracted RelE toxicity. After neutralization of RelE by RelB, rapid recovery of translation required tmRNA, indicating that tmRNA alleviated RelE toxicity by rescuing ribosomes stalled on damaged mRNAs. RelE proteins from Gram-positive Bacteria and Archaea cleaved tmRNA with a pattern similar to that of E. coli RelE, suggesting that the function and target of RelE may be conserved across the prokaryotic domains.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In prokaryotes, the stringent response is defined as the physiological changes elicited by amino acid (aa) starvation. The concentration of the alarmone, ppGpp, increases rapidly after onset of aa starvation. The increased ppGpp level has profound consequences for the global pattern of gene expression. Transcription of genes encoding rRNA and tRNA is inhibited, whereas transcription of biosynthetic operons is stimulated (Cashel et al., 1996; Chatterji and Kumar, 2001). Furthermore, ppGpp increases the competitiveness of stress-related σ factors (Jishage et al., 2002; Laurie et al., 2003). Thus, most, if not all, the control exerted by ppGpp occurs at the level of transcription (Ishihama, 1999). However, inhibition of transcription during nutrient downshifts does not immediately result in reduction in translation, the process that consumes the most cellular energy (in the form of ATP and GTP). During nutritional stress, this could lead to detrimental depletion of NTP pools.
Most prokaryotic chromosomes encode a novel type of stress response elements, the toxin–antitoxin (TA) loci that modulate translation during nutritional stresses (Gerdes, 2000; Christensen et al., 2001; Pedersen et al., 2002). Escherichia coli relBE, a prototype TA locus, encodes the RelE toxin and RelB antitoxin (see Fig. 1). RelE is a very efficient inhibitor of translation, both in vivo and in vitro (Christensen et al., 2001; Pedersen et al., 2002). RelB neutralizes RelE toxicity by direct protein–protein interaction. Furthermore, RelB autoregulates transcription of relBE, and the RelB–RelE complex yields even better repression than RelB alone (Fig. 1). Thus, RelE is a co-repressor of relBE transcription (Gotfredsen and Gerdes, 1998). Consistently, the level of relBE mRNA was very low during exponential cell growth. However, shifts to growth media that induced aa starvation resulted in a rapid and dramatic increase in relBE transcription (Christensen et al., 2001). Concomitantly, relBE reduced the level of translation. Both these effects were readily explained by Lon-mediated degradation of RelB during the starvation period (Christensen et al., 2001).
RelE co-precipitated with ribosomes, suggesting that inhibition of translation by RelE may involve direct contact with the translation apparatus (Galvani et al., 2001). Recently, we used a fully reconstituted in vitro translation system to analyse in more detail how RelE inhibits translation. We found that RelE cleaved mRNAs positioned at the ribosomal A-site, between the second and third base of the A-codon (Fig. 1) (Pedersen et al., 2003). Cleavage was catalytic and highly dependent on the specific codon positioned at the A-site. Thus, a test mRNA with a UAG amber A-codon was cleaved 800-fold more efficiently than one with a UGA opal codon. An mRNA with a UAA ochre codon was cleaved at an intermediate rate. Sense codons placed at the A-site were also cleaved by RelE.
Escherichia coli tmRNA (also called SsrA or 10 Sa RNA) has two known functions: it releases stalled ribosomes from damaged mRNAs, and it tags the nascent polypeptides from such ribosomes for proteolysis (Muto et al., 1998; Karzai et al., 2000). Thus, tmRNA acts as both a tRNA and an mRNA. tmRNA has a domain that folds very similar to a part of tRNAAla and is amino-acylated at its CCA 3′ end (see Figs 1 and 2A). tmRNA also encodes a short tag sequence ANDENYALAA (Tu et al., 1995; Keiler et al., 1996). It is believed that tmRNA enters the empty A-site of stalled ribosomes and adds its charged alanine to the nascent polypeptide. Translation then shifts from the original mRNA to the ANDENYALAA reading frame within tmRNA, a process called trans-translation. As a result, the stalled ribosome is rescued from the damaged mRNA, and the tmRNA-encoded peptide tag is added to the C-terminus of the nascent polypeptide. The tmRNA tag is recognized by multiple proteases, resulting in rapid degradation of the tagged proteins (Keiler et al., 1996; Gottesman et al., 1998; Herman et al., 1998).
Our in vitro experiments predicted that activation of RelE in vivo should result in stalled ribosomes on damaged mRNAs. Therefore, we speculated that tmRNA might be involved in recovery of RelE-inhibited cells. We find that, indeed, RelE mediates cleavage of a representative model mRNA and of tmRNA itself. RelE cleavage was dependent on translation of the RNAs and, as observed in vitro, susceptibility to cleavage depended on the RNA primary sequence. Overexpression of tmRNA counteracted RelE toxicity, and tmRNA stimulated restart of translation in RelE-inhibited cells. Hence, tmRNA rescues ribosomes stalled on mRNAs cleaved by RelE. Interestingly, RelE homologues from Gram-positive Bacteria and Archaea exhibited cleavage activities very similar to that of E. coli RelE. Our results suggest that relBE and related systems function in general cellular stress adaptation.
RelE toxicity is enhanced in cells lacking tmRNA
We measured the effect of tmRNA on RelE-mediated inhibition of colony formation. The relE gene was transcribed from pA1/O4/O3, a synthetic, IPTG-inducible promoter present on the low-copy-number plasmid pMG223, which also carries lacIq (Gotfredsen and Gerdes, 1998). Cells of MG1655 (wt)/pMG223 and MG1655ΔssrA/pMG223 were grown exponentially, and relE transcription was induced at time zero. As seen from Fig. 2B, deletion of ssrA that encodes tmRNA rendered the cells more sensitive to RelE.
Overproduction of tmRNA and SmpB counteracts RelE toxicity
Plasmid pSC320 (pBR322 ssrA+smpB+) and the proper control plasmid were established in strains MG1655/pMG223 (pA1/O4/O3::relE) and MG1655 ΔssrA/pMG223. The smpB gene encodes a protein that interacts with tmRNA and is essential to tmRNA function (Karzai et al., 1999). As seen from Fig. 2C, the presence of the high-copy-number plasmid encoding wt ssrA and smpB severely reduced RelE-mediated inhibition of colony formation, in both a wild-type and a ΔssrA strain. In contrast, an ssrA gene in which the resume codon of tmRNA was changed to a UAA stop codon (Fig. 2A) did not suppress RelE toxicity at all (pssrA′ + smpB in Fig. 2C). The stop codon mutation in ssrA renders tmRNA non-translatable and therefore non-functional in trans-translation (Williams et al., 1999). Thus, the opposing effect of tmRNA on RelE toxicity depended on trans-translation.
Resuscitation of RelE-inhibited cells requires tmRNA
Recently, we showed that RelE overexpression induced a bacteriostatic condition (Pedersen et al., 2002). Thus, the presence of excess RelE prevented colony formation without killing of the cells. This was concluded from the fact that later induction of relB transcription in RelE-inhibited cells mediated complete recovery of colony formation. As deletion of ssrA increased the toxicity of RelE, we speculated that tmRNA might be required for resuscitation of relE-inhibited cells. Strains MG1655/pSC221 (pA1/O4/O3::relB)/pSC3035 (pBAD::relE) and MG1655ΔssrA/pSC221/pSC3035 were grown in synthetic medium supplemented with glucose. At time zero, transcription of relE was induced by transfer of the cells to a medium containing glycerol and arabinose. As seen from Fig. 3A, viable counts on solid medium without IPTG decreased immediately for both cultures. Again, RelE was most toxic in the ssrA deletion strain. However, plating the cells on medium containing IPTG (which induced relB transcription) resulted in complete recovery of wild-type cells. In contrast, recovery of cells lacking tmRNA was considerably reduced. Thus, tmRNA is required for resuscitation of RelE-inhibited cells.
tmRNA is required for rapid recovery of translation in RelE-inhibited cells
Next, we asked whether tmRNA was required for RelB-mediated counteraction of RelE-mediated inhibition of translation. Strains MG1655/pSC221 (pA1/O4/O3::relB)/pSC3035 (pBAD::relE) and MG1655ΔssrA/pSC221/pSC3035 were grown as described above. As expected, the rate of translation decreased dramatically in both cultures immediately after induction of relE transcription (Fig. 3B). Thirty minutes after induction, transcription of relE was halted by the addition of glucose, and the culture was split in two. One subculture received IPTG (to induce relB transcription). As expected, translation resumed in the cultures that received IPTG. Importantly, however, resumption of translation was significantly delayed in the ΔssrA strain. Thus, tmRNA enhanced restart of translation in RelE-inhibited cells.
RelE cleaves lpp mRNA and tmRNA in vivo
To monitor the effect of RelE on mRNA metabolism, we chose lpp mRNA of E. coli as a model substrate. lpp mRNA is a well-characterized, small (322 nt) and relatively stable transcript that expresses the abundant but non-essential housekeeping lipoprotein (Nakamura et al., 1980; Nilsson et al., 1984). Figure 4A (centre) shows a Northern analysis of lpp mRNA before and after induction of relE transcription. As seen, shortly after induction of relE transcription, lpp mRNA disappeared. As RelE is an efficient and instantaneous inhibitor of translation, we also investigated the effect of chloramphenicol on the lpp mRNA band pattern. As seen from Fig. 4A (left), addition of chloramphenicol did not confer degradation of lpp mRNA.
tmRNA was analysed in a similar way (Fig. 5A). In this case, induction of relE transcription conferred a biphasic response, characterized by an initial rapid and severe drop in the amount of tmRNA followed by a slower reappearance of the RNA (Fig. 5A, right). Addition of chloramphenicol did not seriously affect the stability of tmRNA (Fig. 5A, left).
RelE cleavage of lpp mRNA and tmRNA depends on their translation
RelE cleaved mRNAs in vitro only when they were associated with the ribosomes (Pedersen et al., 2003). To prevent lpp mRNA associating with ribosomes, its ATG start codon was changed to AAG. As seen from Fig. 4A (right), induction of relE transcription had no effect on the band pattern of the mutated lpp mRNA. Thus, RelE cleaved lpp mRNA in vivo only when it was translated.
To corroborate this observation, we investigated whether RelE cleavage of tmRNA depended on translation of its tag sequence. The first codon (the resume codon) of tmRNA, GCA, was changed to a stop codon (UAA, Fig. 2A). Plasmid derivatives carrying either the wild-type ssrA smpB or the ssrA GCA→UAA smpB genes were established in MG1655ΔssrA/pMG223 (pA1/O4/O3::relE). Figure 5B shows a Northern analysis of the tmRNA band pattern produced by these two strains before and after induction of relE transcription. As seen, wild-type tmRNA was rapidly degraded (Fig. 5B, left). In contrast, tmRNA carrying the stop codon mutation was not degraded at all (Fig. 5B, right). This result indicated that RelE cleaved tmRNA only when it was translated.
RelE cleavage pattern is codon dependent
To obtain information on the specific RelE cleavage pattern, we performed primer extension analyses on the 5′ part of lpp mRNA before and after induction of relE transcription [Fig. 4B, marked TGC(wt)]. As seen, distinct cleavage bands appeared in the mRNA coding region, but not in the non-translated leader region. Specific cleavage sites are indicated in the mRNA sequence on the left of Figure 4B. As observed in vitro, cleavages were strongest in codons with a G base in their third position. Cleavage occurred between the second and third bases in all cases but one, consistent with the in vitro specificity of RelE cleavage (Pedersen et al., 2003). Strong cleavages were also present at the second codon (AAA), whereas the fifth codon, also AAA, was not cleaved at a detectable level (Fig. 4B).
Stop codons were particularly sensitive to RelE cleavage in vitro. The stop codon of lpp is located close to the 3′ end of the mRNA (Nakamura et al., 1980). Therefore, we introduced all three stop codons at codon 21 of lpp, in a region where RelE did not cleave. Primer extension reactions of the mutated lpp mRNAs before and after induction of relE transcription are shown in Fig. 4B. Introduction of the UAA and UAG stop codons resulted in strong RelE-mediated cleavages between the second and third bases of the stop codons. In contrast, the UGA codon conferred weak cleavage only, between the first and second codons. Thus, RelE cleavage was weakest at the UGA codon. In the case of UGA, but not UAA or UAG, cleavages were seen beyond the stop codon. We believe that these cleavages were a consequence of translational readthrough at UGA, the most inefficient of the three stop codons (Poole et al., 1995; Tate and Mannering, 1996).
RelE cleavage pattern of tmRNA is also codon dependent
The UAA stop codon of wild-type tmRNA was changed to either UAG or UGA (Fig. 2A). Figure 5C shows a primer extension analysis of wild-type and mutated tmRNAs before and after relE induction. The band patterns of tmRNAs carrying either UAA or UAG were similar, with prominent cleavages at the stop codons and at the UAC and UUA sense codons upstream from the stop codons (arrows in Fig. 5C). The cleavage sites in tmRNA are also shown in Fig. 2A. The tmRNA carrying the UGA opal codon exhibited cleavage at the two sense codons, whereas cleavage at the stop codon was greatly reduced. Again, UGA seemed to be a less efficient substrate for RelE-mediated cleavage than the other two stop codons. tmRNA was not cleaved by RelE outside its tag sequence (data not shown).
RelE cleaves tmRNA and lpp mRNA during amino acid starvation
Previously, we reported that aa starvation activates RelE and reduces the post-starvation level of translation by ≈ 50% (Christensen et al., 2001). We investigated the tmRNA band pattern before and after induction of aa starvation by the addition of serine hydroxamate (a competitive inhibitor of seryl tRNA synthetase). As seen in Fig. 6 (right), a strain carrying the wild-type relBE locus on a low-copy plasmid exhibited RelE-specific bands 60 min after the onset of starvation. tmRNA in the control strain lacking relBE was not cleaved during this period. RelE also cleaved lpp mRNA during aa starvation (data not shown).
RelE homologues from Gram-positive Bacteria and Archaea cleave tmRNA in E. coli
Chromosomes from Bacteria and Archaea encode relBE-homologous loci, often in multiple copies (Gerdes, 2000). For instance, Streptococcus pneumoniae and Methanocaldococcus jannaschii encode two and four relBE loci respectively. The six corresponding relE genes were cloned and overexpressed in E. coli. All six RelE proteins except homologue #1 from S. pneumoniae inhibited translation in E. coli (data not shown). Figure 7 shows a primer extension analysis of tmRNA from E. coli cells that overexpress heterologous RelEs. As seen, M. jannaschii homologues #2 and #3 and S. pneumoniae homologue #2 yielded patterns very similar to that of E. coli RelE, with prominent cleavages at the tmRNA stop and UUA codons. M. jannaschii homologue #4 yielded cleavage at the UUA codon and at the penultimate GCA codon. Thus, RelE proteins from Gram-positive Bacteria and Archaea exhibited or induced enzymatic activities very similar to that of E. coli RelE. By inference, this suggests that the toxic effect of overexpression of the heterologous RelE proteins resulted from inhibition of translation by mRNA cleavage.
Here, we present evidence that E. coli RelE mediates cleavage of a representative model mRNA and tmRNA in vivo. On purified ribosome complexes, RelE mediated cleavage of mRNA codons positioned at the ribosomal A-site (Pedersen et al., 2003). RelE mediated cleavage within the coding regions of lpp and tmRNA and mutations that abolished translation of the RNAs also prevented cleavage. Together, these results show that RelE inhibits translation by mediating cleavage of translated RNAs.
RelE cleavage mechanism and specificity
In vitro, RelE cleaved mRNAs in a highly codon-dependent fashion (Pedersen et al., 2003). Thus, a G or C at the third codon position enhanced cleavage, whereas a G or C at the second position reduced cleavage. For example, the UAG amber codon was cleaved 800-fold more specifically than the UGA opal codon (expressed as relative Kcat/Km values).
To validate the in vitro measurements, we constructed lpp and tmRNAs with all three stop codons and compared their in vivo RelE cleavage patterns (Figs 4 and 5). In lpp mRNA, RelE conferred cleavage at all three stop codons. As also observed in vitro, UGA was the least sensitive codon. Similarly, the RelE-mediated cleavages of tmRNAs carrying the three different stop codons yielded an overall qualitative pattern that was consistent with the quantitative values obtained in vitro.
RelE also mediated cleave of sense codons of lpp and tmRNA. In lpp, the cleavages occurred between the second and third codon positions with a preference for a G in the third position. However, in vivo, the second codon in lpp, AAA, was cleaved efficiently, whereas the AAA codon at the fifth position was not cleaved at a detectable level (Fig. 4B). Consistently, an AAA codon at the fifth position of a test mRNA was cleaved very inefficiently in vitro (Pedersen et al., 2003). Thus, it seems that stop codons and codons adjacent to the start codon are particularly sensitive to RelE-mediated cleavage. The reason for this is not yet known. The processes of initiation and termination of translation are considerably slower than those of elongation (Freistroffer et al., 2000). Hence, the A-site may be open for longer times during initiation and termination than during elongation. These observations suggest that an open A-site increases the probability of RelE-promoted mRNA cleavage.
It is not known how RelE mediates RNA cleavage. In vitro, RelE cleavage at stop codons was inhibited by the addition of cognate class I release factors (RFI or RFII) to the cleavage reaction, suggesting that RelE competes with release factors for access to the ribosomal A-site (Pedersen et al., 2003). In vivo, access to the A-site is determined by a number of factors including the ternary aa-tRNA/EF-Tu/GTP complex, RF1/RF2 and RF3, RRF and perhaps others. It has been suggested that translation factors recognize the A-site by tRNA mimicry (Nakamura et al., 2000), leaving open the possibility that RelE might also mimic the structure of tRNAs.
RelE mediated cleavage of mRNA associated with ribosomes only, both in vitro and in vivo. Thus, as yet unidentified components of the ribosome may take part in the cleavage reaction. The ribosomal A-site consists almost entirely of RNA (Carter et al., 2000), raising the interesting possibility that ribosomal RNA participates directly or indirectly in the catalytic reaction.
Cells lacking tmRNA exhibited increased sensitivity to RelE (Fig. 2B), and simultaneous overproduction of tmRNA and SmpB efficiently counteracted RelE toxicity (Fig. 2C). Furthermore, without tmRNA, RelB could not fully revive cells inhibited by RelE (Fig. 3A). The reduced resuscitation of cells lacking tmRNA was correlated with a lower translation rate (Fig. 3B). These results show that tmRNA alleviated the toxic effect of RelE by counteracting inhibition of translation. tmRNA may counteract RelE in several ways. For example, tmRNA could out-titrate RelE. However, a tmRNA with its resume codon changed to a stop codon did not counteract RelE toxicity (Fig. 2C). Instead, we therefore favour an explanation that involves the biological function of tmRNA. It has been well described that tmRNA rescues ribosomes stalled at damaged mRNAs (Tu et al., 1995; Keiler et al., 1996). As RelE inhibits translation by mRNA cleavage, our results suggest that tmRNA counteracts RelE by promoting ribosome recycling. In RelE-poisoned cells, tmRNA can only do this to a limited extent (because tmRNA is also cleaved by RelE). However, after RelE neutralization by RelB, replenishment of tmRNA should lead to a rapid restart of translation and rescue of cell viability. This was precisely what was observed (Fig. 3A and B). Thus, the result of the resuscitation experiment is consistent with the proposal that tmRNA counteracts RelE activity by promoting translational restart. One prediction from this model is that a large proportion of the nascent peptides should be tagged by tmRNA immediately after translational restart in RelE-inhibited cells.
What is the physiological role of RelE? We showed previously that RelE was activated during the stringent response and reduced the post-starvation level of translation (Christensen et al., 2001). Activation resulted from Lon-mediated degradation of RelB. Consistently, we found here that aa starvation triggered RelE cleavage of tmRNA (Fig. 6) and lpp mRNA (data not shown). Thus, one consequence of RelE activation is a reduced global level of translation. Reduction in translation may possibly have several beneficial functions. (i) An overall reduction in energy consumption (in the form of ATP and GTP) during limited supply of nutrients. A reduction in translation contributes to maintain a high amino acid pool. In turn, this may (ii) reduce the rate of translational errors and (iii) sustain a high rate of synthesis of those proteins that are produced by mRNAs escaping RelE cleavage. This, in turn, may allow the cell to react more adeptly to signals that demand rapid changes in gene expression. (iv) The large variation in cleavage specificities exhibited by different codons raises the possibility that certain mRNAs or even classes of mRNAs may be resistant to inactivation by RelE. If true, translation of mRNAs encoding proteins increasing in fitness during stressful conditions may be favoured by the presence of relBE. Hence, the beneficial physiological effect of relBE may thus be a combination of several additive mechanisms.
RelE homologues from Gram-positive Bacteria and Archaea inhibited translation in E. coli (data not shown) and cleaved tmRNA (Fig. 6). For several of the RelE homologues, the tmRNA cleavage pattern was very similar to that of E. coli RelE, suggesting that both target recognition and cleavage specificity are conserved across the prokaryotic domains. This is surprising as most antibiotics that inhibit translation in Bacteria do not inhibit translation in Archaea. Although RelE proteins from Bacteria and Archaea are clearly homologous, their sequence similarities are modest (e.g. RelE from E. coli and M. jannaschii homologue #2 share 18% identical and 40% similar aa only). In this light, the identical reaction specificity seems even more surprising and is consistent with a more conserved element taking part in it (e.g. ribosomal RNA). The function of relBE loci in Gram-positive Bacteria and Archaea is not known but, as in the case of E. coli relBE, may be related to regulation of translation during nutritional stress. It should be noted that archaeal tmRNA homologues or analogues have not been identified.
Here, we give evidence of an unprecedented example of how a small protein can regulate the capacity of the entire protein-synthesizing apparatus by cleavage of ribosome-associated mRNAs, a highly surprising and dramatic mechanism. We also show that tmRNA alleviates the inhibitory effect of RelE during restart of translation. RelB antagonizes RelE activity. In turn, the RelB level is regulated by Lon protease. It will be interesting to learn whether a cellular signal triggers Lon to degrade RelB.
Growth conditions and media
Cells were grown in either Luria–Bertani (LB) broth or M9 minimal medium supplemented with aa in defined concentrations and 0.2% glucose or 0.5% glycerol at 37°C. When appropriate, the medium was supplemented with ampicillin (30 or 100 µg ml−1), chloramphenicol (25 or 50 µg ml−1) or tetracycline (10 µg ml−1). [35S]-methionine (1175 Ci mmol−1) was obtained from NEN and serine hydroxamate from Sigma.
Bacterial strains and plasmids (listed in Table 1)
ssrA was deleted from the chromosome of MG1655 by the procedure described by Datsenko and Wanner (2000). A polymerase chain reaction (PCR) product was synthesized using primers delta ssrA-1 (5′-CGAATAAAAA TCAGGCTACATGGGTGCTAAATCTTTAACGATAACGC C G T GTAGGCTGGAGCTGCTTC) and delta ssrA-2 (5′-CTTAG GACTTCATCGGATGACTCTGGTAATCACCGATGGA G AATT TTGCATATGAATATCCTCCTTTAG) with pKD3 as template. The PCR product was electroporated into BW25113/pKD46, and cells were spread on LA plates containing 25 µg ml−1 chloramphenicol. Deletion of the ssrA locus was verified by PCR. The cat allele was then transduced into MG1655 and removed as described by Datsenko and Wanner (2000).
A PCR product was synthesized using primers delta lpp-1 (5′-GTTCGATGCTTCTTTGAGCGAACGAT CA A A AATAAGTGCCTTCCCATCGTGTAGGCTGGAGCTGC TTC) and delta lpp-2 (5′-GAATGGTGAACCAGAGCAAG GGAATATGTTACGCGTGACGCAGTAGCGGCATATG A ATATC CTCCTTAG) with pKD3 as template. The PCR product was electroporated into BW25113/pKD46, and the cells were spread on LA plates containing 25 µg ml−1 chloramphenicol. The cat allele was then transduced into MG1655 and removed, resulting in MG1655Δlpp. Deletion of cat and lpp was verified by PCR.
pMG71 and pNDM71.
pOU71 is a mini-R1 runaway replication plasmid carrying unique BamHI and EcoRI restriction sites useful for cloning purposes (Larsen et al., 1984). Two oligonucleotides, RBJ1 (5′-AATTCCTCGAGCACGTGGGTA CCG) and RBJ2 (5′-GATCCGGTACCCACGTGCTCGAGG), were annealed and inserted between the BamHI and EcoRI sites of pOU71, resulting in pMG71. This plasmid contains unique restriction sites KpnI, PmlI and XhoI between the unique BamHI and EcoRI sites. An AatII–AspI (Tth111I) DNA fragment carrying the par locus of R1 (Dam and Gerdes, 1994) was inserted between the AatII and AspI sites of pMG71, resulting in pNDM71.
pMG1923 and pMG3323.
A DNA fragment encoding relE was generated by PCR using primers relE1B (5′-CCCCCG GATCCATAAGGAGTTTTATAATGGCGTATTTTCTGGATTTT GACG) and relE2 (5′-CCCCCCTCGAGGTCGCTCAGAG AATGCGTTTGACCGC). The fragment was restricted with BamHI and SalI and cloned into the multiple cloning site (MCS) of pUC19, resulting in pMG1923. A SacI–PstI fragment of pMG1923 was cloned into pBAD33 also digested with SacI and PstI. The resulting plasmid, pMG3323, contains the relE gene with an efficient ribosome binding site (from parM of plasmid R1) downstream of the pBAD promoter.
relEII of M. jannaschii was amplified from chromosomal DNA of DSM strain 2661 with primers RelE MJ#2 BamHI, XbaI Op SD (5′-CCCCGGATCCTCTAGATAAAGG GAGGAAAAAACCATGAAAGTGTTATTTGCTAAAAC-3′) and RelE MJ#2-2 EcoRI, HindIII (5′-CCCCGAATTCAAGCTTT TATGGGAAATAATCATATATGC -3′). The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM220.
relEIII of M. jannaschii was amplified from chromosomal DNA with primers relE MJ#3 opSD, BamHI, XbaI (5′-CCCCGGATCCTCTAGATAAAGGGAGGAAAAAACCATGAA ACAATGGAAGTATCTTTTAAAAAAATC) and relE MJ#3-2 EcoRI, HindIII (5′-CCCCGAATTCAAGCTTTTATGGAAATCT TTTATAAATTTGTTTTC). The PCR product was digested with EcoRI and BamHI and inserted into pNDM220.
relEIV of M. jannaschii was amplified from chromosomal DNA with primers relE MJ#4-1 opSD, BamHI (5′-CCCCGGATCCTAAAGGGAGGAAAAAACCATGTATGAAAT CGAAATAATGCC) and relE MJ#4-2 EcoRI, HindIII (5′-CCCCGAATTCAAGCTTGAGTTAATCTTTAAATAGCTTTTTC). The PCR product was digested with EcoRI and BamHI and inserted into pNDM220.
relEII of S. pneumoniae strain R61 was amplified from chromosomal DNA with primers relE sp2-1 opSD, BamHI (5′-CCCCGGATCCTAAAGGGAGGAAAAAACCAT GAATAATTTGTATAAATTAGTTCC) and sp2 ccw (5′-CCCCGAATTCGAATGAAAATTTACTTGAAAAAAG). The PCR product was digested with EcoRI and BamHI and inserted into pNDM220.
relB was amplified by PCR from pBD2430 with primers relBopSDBamH1 (5′-CCCGGATCCTAAAGGGAG GAAAAAACCATGGGTAGCATTAACCTGCGTATTG-3′) and relB2 (3′-CCCCCCTCGAGGTCGACTCAGAGTTCATCCA GCGTCACACGTACTGG-5′). The PCR product was digested with BamHI and SalI and inserted into BamHI–XhoI-treated pNDM220. Cells carrying pSC221 produce RelB upon addition of IPTG.
smpB and ssrA were amplified from MG1655 using primers smpB−1 EcoRI (5′-CCCCGAATTCCAGTGCGGTC CGGCTAATC) and ssrA-2 (5′-CCCCCTGCAGGGTTCG GATTTAATTAGTTCTC). The PCR product was digested with EcoRI and PstI and inserted into pBR322.
smpB and ssrA were amplified from MG1655 with primers: smpB−1 EcoRI and ssrA stop 1-2 (5′-GTAGTTTTCGTCGTTTTAGACTATTTTTTGCGG) in addition to ssrA stop 1-1 (5′-CCGCAAAAAATAGTCTAAAACGAC GAAAACTAC). The resulting two overlapping PCR products were used as templates in a second PCR reaction with the smpB−1 EcoRI and ssrA−2 primers. The new PCR product was digested with EcoRI and PstI and inserted into pBR322. pSC321 encodes smpB and ssrA with its resume codon changed to a UAA stop codon.
lpp was amplified by PCR from MG1655 with lpp-1 BamHI (5′-CCCCGGATCCGGAGATTAACTCAATCTA GAGG) and lpp-2 EcoRI (5′-CCCCGAATTCGCGCCATT TTTCACTTCACAG). The resulting PCR product was digested with EcoRI and BamHI and inserted into pGEM3.
ssrA was amplified by PCR from MG1655 with ssrA−1 BamHI (5′-CCCCGGATCCCTTAATCGAATAAAAAT CAGGCTAC) and ssrA−2 EcoRI (5′-CCCCGAATTCGTG GAGCTGGCGGGAGTTG). The resulting PCR product was digested with EcoRI and BamHI and inserted into pGEM3.
The lpp gene was amplified by PCR with primers lpp−136 BamHI (5′-CCCCGGATCCGGAAGCATC CTGTTTTCTCTC) and lpp+101 EcoRI (5′-CCCCGAATTC GAACCAGAGCAAGGGAATATG). The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71.
The lpp gene was amplified by PCR from MG1655 with primers: lpp−136 BamHI and lpp 1AAG-2 (5′-GTTTAG TAGCTTTCTTTATTAATACCCTC) in addition to lpp 1AAG-1 (5′-GAG GGTATTAATAAAGAAAGCTACTAAAC) and lpp+101 EcoRI. The two overlapping PCR products were used as template in a PCR with the lpp−136 BamHI and lpp+101 EcoRI external PCR primers. The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71. pSC711 encodes the lpp messenger with an AAG codon replacing the natural ATG start codon, which prevents translation of the lpp mRNA.
smpB and ssrA were amplified from MG1655 using primers smpB −1 EcoRI (5′-CCCCGAATTCCAGTGCGGTC CGGCTAATC) and ssrA−2 BamHI (5′-CCCCGGATCCGGT TCGGATTTAATTAGTTCTC). The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71.
smpB and ssrA were amplified from MG1655 with primers: smpB−1 EcoRI and ssrA UGA-2 (5′-CTCTAAG CAGGTTATCAAGCTGCTAAAGCG) in addition to ssrA UGA-1 (5′-CGCTTTAGCAGCTTGATAACCTGCTTAGAG) and ssrA-2 BamHI. The two overlapping PCR products were used as templates in a PCR with the smpB −1 EcoRI and ssrA−2 BamHI primers. The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71. pSC715 encodes ssrA with a UGA stop codon instead of the natural UAA stop codon.
smpB and ssrA were amplified from MG1655 with primers: smpB −1 EcoRI and ssrA UAG-2 (5′-CTCTAAG CAGGTTACTAAGCTGCTAAAGC) in addition to ssrA UAG-1 (5′-GCTTTAGCAGCTTAGTAACCTGCTTAGAG) and ssrA-2 BamHI. The two PCR products were used as templates in a PCR with the smpB −1 EcoRI and ssrA−2 BamHI primers. The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71. pSC715 encodes ssrA with a UAG stop codon instead of the natural UAA stop codon.
lpp was amplified by PCR from MG1655 with primers lpp-136 BamHI and lpp 21 TAA-2 (5′-GATTTTAGCGT TGCTGGATTAACCTGCCAGCAGAGTAG-3′) in addition to lpp 21 TAA-1 (5′-CTACTCTGCTGGCAGGTTAATCCAG CAACGCTAAAATC-3′) and lpp+101 EcoRI. The two PCR products were used as templates in a PCR with the lpp−136 BamHI and lpp+101 EcoRI primers. The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71. pSC719 encodes the lpp gene with a UAA stop codon at codon number 21 in the lpp reading frame.
The lpp gene was amplified by PCR from MG1655 with primers lpp−136 BamHI and lpp 21 TGA-2 (5′-GATTTTAGCGTTGCTGGATCAACCTGCCAGCAGAGTAG) in addition to lpp 21 TGA-1 (5′-CTACTCTGCTGGCAGGT TGATCCAGCAACGCTAAAATC) and lpp+101 EcoRI. The two PCR products were used as templates in a PCR with the lpp−136 BamHI and lpp+101 EcoRI primers. The resulting PCR product was digested with EcoRI and BamHI and inserted into pNDM71. pSC720 encodes the lpp gene with a UGA stop codon at codon number 21 in the lpp reading frame.
The lpp gene was amplified by PCR from MG1655 with primers lpp−136 BamHI and lpp 21 TAG-2 (5′-GATTTTAGCGTTGCTGGACTAACCTGCCAGCAGAGTAG) in addition to lpp 21 TAG-1 (5′-CTACTCTGCTGGCAGGT TAGTCCAGCAACGCTAAAATC) and lpp+101 EcoRI. The two PCR products were used as templates in a PCR with the lpp−136 BamHI and lpp+101 EcoRI primers. This PCR product was digested with EcoRI and BamHI and inserted into pNDM71. pSC721 encodes the lpp gene with a UAG stop codon at codon number 21 in the lpp reading frame.
Northern and primer extension analyses
Cells were grown in LB at 37°C. At an OD450 of ≈ 0.5, aa starvation was induced by the addition of serine hydroxamate (1 mg ml−1), which prevented further cell growth. Alternatively, relE was induced by the addition of either 0.2% arabinose (pBAD derivatives) or 2 mM IPTG (pNDM220 derivatives). For Northern analysis, total RNA was fractionated by PAGE (6% low-bis acrylamide), blotted to a Zeta-probe nylon membrane and hybridized with a locus-specific single-stranded 32P-labelled riboprobe. Radioactively labelled RNA probes complementary to ssrA or lpp were generated using linearized plasmid DNA of pSC334 or pSC333 respectively. Semi-quantitative primer extension analysis was performed essentially according to the method of Dam and Gerdes (1994) with primer lpp 21 (5′-CTGAACGTCAGAAGACAGCT GATCG-3′) or 10SA-2 (5′-GCCCCTCGGCATGCACC-3′).
Rates of protein synthesis
Cells were grown at 37°C in M9 minimal medium plus 0.2% glucose and an aa mixture in a defined concentration to an OD450 of 0.5. The cultures were then diluted 10 times. At an OD450 of ≈ 0.4, the cells were harvested, washed and resuspended in M9 + 0.5% glycerol. At time zero, 0.2% arabinose was added to induce transcription from the pBAD promoter. After 30 min, transcription from the pBAD promoter was inhibited by the addition of 0.2% glucose. Simultaneously, 0.5 mM IPTG was added to half the culture to induce transcription from the pA1/O4/O3 promoter. Cell samples of 0.5 ml were taken at the time points indicated and added to 5 µCi of 35S-methionine containing 100-fold excess of unlabelled methionine. In addition, a sample was taken for OD450 measurement. After 1 min of incorporation, samples were chased for 10 min with 0.5 mg of unlabelled methionine. The samples were harvested, resuspended in 200 µl of cold 20% TCA, centrifuged at 20 000 g for 30 min at 4°C, washed twice with 200 µl of cold 96% ethanol, precipitates transferred to vials and counted in a liquid scintillation counter. The amount of incorporated radioactivity per OD450 (specific rate of synthesis) was plotted against time.
We thank Måns Ehrenberg for suggestions and Manuel Espinosa for the donation of S. pneumoniae chromosomal DNA. We also thank Marie and Nikolaj D. Mikkelsen for the construction of pMG71, pNDM71, pMG1923 and pMG3323. This work was supported by EU-grants QLK3-CT-2001-00277, BIO4-98-0283 and The Danish Biotechnology Instrument Center (DABIC).