The Inhibitory Mechanism of Protein Synthesis by YoeB, an Escherichia coli Toxin*

YoeB is a toxin encoded by the yefM-yoeB antitoxin-toxin operon in the Escherichia coli genome. Here we show that YoeB, a highly potent protein synthesis inhibitor, specifically blocks translation initiation. In in vivo primer extension experiments using two different mRNAs, a major band was detected after YoeB induction at three bases downstream of the initiation codon at 2.5 min. An identical band was also detected in in vitro toeprinting experiments after the addition of YoeB to the reaction mixtures containing 70 S ribosomes and the same mRNAs, even in the absence of \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{tRNA}_{\mathrm{f}}^{\mathrm{Met}}\) \end{document}. Notably, this band was not detected in the presence of YoeB alone, indicating that YoeB by itself does not have endoribonuclease activity under the conditions used. The 70 S ribosomes increased upon YoeB induction, and YoeB was found to be specifically associated with 50 S subunits. Using tetracycline and hygromycin B, we demonstrated that YoeB binds to the 50 S ribosomal subunit in 70 S ribosomes and interacts with the A site leading to mRNA cleavage at this site. As a result, the 3′-end portion of the mRNA was released from ribosomes, and translation initiation was effectively inhibited. These results demonstrate that YoeB primarily inhibits translation initiation.

YoeB is one of the toxins encoded by the genome of Escherichia coli, which is co-expressed with YefM, its cognate antitoxin (1). It has been shown that YefM is highly unstable in the cell and purified YefM exists as a denatured form without secondary structure (2). However, when YefM is co-expressed with YoeB, a dimer of YefM forms a stable complex with a single molecule of YoeB. The x-ray structure of this heterotrimer has been determined (3). Under normal growth conditions, YoeB presumably exists as a heterotrimer and is thus unable to exert its cellular toxicity. Lon is an ATP-dependent protease that is induced under various stress conditions, and has been implicated to be responsible for the degradation of YefM and release of YoeB. This consequently leads to inhibition of protein synthesis. These authors demonstrated that the yefM-yoeB system is involved in Lon-dependent lethality (1). They also showed that induction of YoeB leads to cleavage of the lpp mRNA at a few sites and concluded that YoeB functions similar to other E. coli toxins such as RelE, MazF, and ChpBK by cleaving translated mRNAs (1). Interestingly, RelE by itself has no endoribonuclease activity (4) and has been proposed to be a ribosome-associated factor, which stimulates the endogenous ribonuclease activity of ribosomes (4,5). On the other hand, MazF is a ribosome-independent endoribonuclease that cleaves mRNAs specifically at ACA sequences, and is thus termed an mRNA interferase (6). ChpBK was also found to be another mRNA interferase specifically cleaving mRNAs at UAC sequences (7). Therefore, it remains to be determined if YoeB is a ribosome-associating factor like RelE or if it functions as a ribosome-independent mRNA interferase like MazF and ChpBK or if it inhibits translation by a yet-unknown mechanism. Notably, purified YoeB has been shown to have an intrinsic endoribonuclease activity specific for purine residues in RNA (3).
In the present paper we demonstrate that YoeB is a 50 S ribosome-associating factor that inhibits translation initiation using a mechanism that is different from the mechanism proposed earlier by Christensen et al. (1) and Kamada and Hanaoka (3). The former authors proposed that YoeB induces cleavage of the translated mRNA and the latter authors concluded that YoeB is a purine-specific endoribonuclease. In in vivo primer extension experiments carried out with two different mRNAs, a major band was detected after YoeB induction at a position three bases downstream of the initiation codon at 2.5 min. An identical band was also detected in in vitro toeprinting experiments when YoeB was added to the reaction mixtures containing 70 S ribosomes and the same mRNAs even in the absence of initiator tRNA f Met . When tRNA f Met was added to the reaction mixtures in the absence of YoeB, toeprinting sites were shifted to a position 13 to 14 bases downstream of the initiation codon. When YoeB forms a complex with 70 S ribosomes and mRNA in vitro, partial cleavage of mRNAs is also observed at positions three and four bases downstream of the initiation codon as observed in in vivo experiments. Notably, this partial cleavage was not observed when YoeB was incubated with mRNA in the absence of ribosomes. However, when a nuclease-negative YoeB mutant (H83Q) (3) was used in in vitro toeprinting experiments, partial cleavage was observed. Using tetracycline and hygromycin B, we demonstrated that YoeB binds to the 50 S ribosomal subunit in 70 S ribosomes and interacts with the A site leading to mRNA cleavage at this site. As a result, the 3Ј-end portion of the mRNA was released from ribosomes. These results indicate that YoeB associates with 50 S subunits in 70 S ribosomes to effectively inhibit translation initiation by preventing the formation of the translation initiation complex on mRNAs.

EXPERIMENTAL PROCEDURES
Strains and Plasmids-E. coli BL21(DE3), BW25113 (⌬araBAD) (8), and MRE600 (9) were used. The yefM-yoeB operon was amplified by PCR using E. coli genomic DNA as template and cloned into the NdeI-XhoI sites of pET21c (Novagen). This construction created an in-frame translation fusion with a (His) 6 tag at the yoeB C-terminal end. The plasmid was designated as pET21c-YefM-YoeB(His) 6 . The yoeB gene was ]methionine incorporation in toluene-treated cells. E. coli BW25113 cells containing pBAD-YoeB were grown at 37°C in glycerol-M9 medium. When the A 600 of the culture reached 0.4, arabinose was added to a final concentration of 0.2%. After incubation at 37°C for 10 min, the cells were treated with toluene (10,11). Using toluene-treated cells, ATP-dependent protein synthesis was carried out with [ 35 S]methionine as described previously (10). B, effect of YoeB on [␣-32 P]dCTP incorporation in toluene-treated cells (12). C, effect of YoeB on [␣-32 P]UTP incorporation in toluene-treated cells (13). D, growth curve of BW25113 cells containing pBAD-YoeB plasmid in LB medium. Arabinose was added at 175 min as indicated by the solid arrow. E, effect of YoeB on the rate of [ 35 S]methionine incorporation in vivo. [ 35 S]Methionine incorporation into E. coli BW25113 cells containing pBAD-YoeB was measured at various time points after YoeB induction as indicated. F, SDS-PAGE analysis of in vivo protein synthesis after the induction of YoeB. The same cultures in E were used. cloned into pBAD creating pBAD-YoeB to tightly regulate yoeB expression by the addition of arabinose (0.2%). The yefM gene was cloned into pGEX-4T-1 (Amersham Biosciences) creating pGEX-4T-1-YefM.
Assay of Protein, DNA and RNA Synthesis in Toluene-treated Cells-A 70-ml culture of E. coli BW25113 containing pBAD-YoeB plasmid was grown at 37°C in M9 medium with 0.5% glycerol (no glucose) and all the amino acids except for methionine and cysteine (1 mM each). When the A 600 of the culture reached 0.4, arabinose was added to a final concentration of 0.2%. After incubation at 37°C for 10 min, the cells were treated with 1% toluene (10,11). Using toluene-treated cells, protein synthesis was carried out with [ 35 S]methionine as described previously (10,11). The toluene-treated cells were washed once with 0.05 M potassium phosphate buffer (pH 7.4) at room temperature, and then re-suspended into the same buffer to examine DNA synthesis using [␣-32 P]dCTP as described previously (12). For assaying RNA synthesis, the toluenetreated cells were washed once with 0.05 M Tris-HCl buffer (pH 7.5) at room temperature, and then re-suspended into the same buffer to measure [␣-32 P]UTP incorporation into RNA as described previously (13).
Assay of in Vivo Protein Synthesis-E. coli BW25113 cells containing pBAD-YoeB were grown in M9 medium with 0.5% glycerol (no glucose) and all the amino acids except for methionine and cysteine (1 mM each). When the A 600 value of the culture reached 0.4, arabinose was added to a final concentration of 0.2% to induce YoeB expression. Cell cultures (0.6 ml) were taken at time intervals as indicated in Fig. 1F and mixed with 30 Ci of [ 35 S]methionine. After a 1-min incubation at 37°C, the rate of protein synthesis was determined as described previously (6). For SDS-PAGE analysis of the total cellular protein synthesis, samples were removed from the [ 35 S]methionine incorporation reaction mixture (500 l) at time intervals indicated in Fig. 1F and added into chilled test tubes containing 100 g/ml each of non-radioactive methionine and cysteine. Cell pellets collected by centrifugation were dissolved into 50 l of loading buffer and subjected to SDS-PAGE followed by autoradiography.
Effect of YoeB and MazF on Protein Synthesis in Prokaryotic and Eukaryotic Cell-free Systems-Prokaryotic cell-free protein synthesis was carried out with an E. coli T7 S30 extract system (Promega). The reaction mixture consisted of 10 l of S30 premix, 7.5 l of S30 extract, and 2.5 l of an amino acid mixture (1 mM each of all amino acids except methionine), 1 l of [ 35 S]methionine, and different amounts of YoeB(His) 6 and GST-YefM in a final volume of 29 l. The different amounts of YoeB(His) 6 and GST-YefM were preincubated for 10 min at 25°C before the assay was started by adding 1 l of pET-11a-MazG plasmid-DNA (0.16 g/l) (15). The reaction was performed for 1.5 h at 37°C, and proteins were then precipitated with acetone and analyzed by SDS-PAGE followed by autoradiography. Eukaryotic cell-free protein synthesis was carried out with a rabbit reticulocyte lysates system, TNT T7 Coupled Reticulocyte Lysate System (Promega). A luciferase T7 control DNA was used as DNA template. The reaction was performed for 1.5 h at 30°C, and proteins were then precipitated with acetone and analyzed by SDS-PAGE.  6 and MazF(His) 6, on cell-free protein synthesis. A, purification of YoeB(His) 6 . Lane 1, purified YefM-YoeB(His) 6 complex, and lane 2, purified YoeB(His) 6 . B, purification of MazF(His) 6 . Lane 1, purified MazE-MazF(His) 6 complex, and lane 2, purified MazF(His) 6 . C, effect of YoeB(His) 6 on MazG synthesis in a prokaryotic cell-free protein synthesis system. Lane 1, without YoeB(His) 6 ; lanes 2-5, 30, 60, 240, and 960 nM YoeB(His) 6 were added, respectively; and lanes 6 and 7, 960 nM YoeB(His) 6 plus GST-YefM in its ratios to YoeB(His) 6 of 2.8 and 4.2, respectively. D, effect of MazF(His) 6 on MazG synthesis in a prokaryotic cell-free protein synthesis system. Lane 1, without MazF(His) 6 ; lanes 2-5, 161, 644, 1288, and 2576 nM MazF(His) 6 were added, respectively; lanes 6 and 7, 2576 nM MazF(His) 6 plus MazE(His) 6 in its ratios to MazF(His) 6 of 1.7 and 11.9, respectively. E, effect of MazF(His) 6 and YoeB(His) 6 on the synthesis of luciferase in a eukaryotic cell-free protein synthesis system. Lane 1, without protein; lane 2, 2.57 M MazF(His) 6 ; and lane 3, 31.5 M MazE(His) 6 and 2.57 M MazF(His) 6 (the ratio of (His) 6 MazE to MazF-(His) 6  Effect of YoeB on the yefM mRNA in Vitro-Different amounts of YoeB were incubated with 1 g of the yefM mRNA at 37°C for 15 min in a 10-l reaction mixture containing 10 mM Tris-Cl (pH 7.8), 60 mM NH 4 Cl, 10 mM MgCl 2 , and 1 mM DTT 2 as indicated in Fig. 2F. The reaction products were analyzed by 8% acrylamide native gel electrophoresis followed by ethidium bromide staining.
Preparation of E. coli 70 S Ribosomes-70 S ribosomes were prepared from E. coli MRE 600 as described previously (16 -18) with minor modifications. Bacterial cells (2 g) were suspended in buffer A (10 mM Tris-HCl (pH 7.8) containing 10 mM MgCl 2 , 60 mM NH 4 Cl, and 6 mM 2-mercaptoethanol). Cells were lysed by French Press. Cell debris was removed by centrifugation two times at 30,000 ϫ g for 30 min at 4°C with a Beckman 50Ti rotor. The supernatant (three-fourths volume from the top) was then layered over an equal volume of 1.1 M sucrose in buffer B (buffer A containing 0.5 M NH 4 Cl) and centrifuged at 45,000 ϫ g for 15 h at 4°C with a Beckman 50Ti rotor. After washing with buffer A, the ribosome pellets were resuspended in buffer A and applied to a linear 5 to 40% (w/v) sucrose gradient prepared in buffer A, and centrifuged at 35,000 ϫ g for 3 h at 4°C with a Beckman SW41Ti rotor. Gradients were fraction-ated and the 70 S ribosome fractions were pooled, pelleted at 45,000 ϫ g for 20 h at 4°C with a Beckman 50Ti rotor. The 70 S ribosome pellets were resuspended in buffer A before they were stored at Ϫ80°C.
Ribosome Profile Analysis-For polysome profile analysis, cells containing pBAD-YoeB plasmid were grown at 37°C in 150 ml of LB medium and at A 600 of 0.6, the arabinose was added to a final concentration of 0.2%, after 10 min induction, chloramphenicol was added to a final concentration of 200 g/ml. Cell pellets were suspended in 1 ml of buffer A (10 mM Tris-Cl, pH 7.8, 60 mM NH 4 Cl, 10 mM MgCl 2 , and 1 mM DTT). Extracts were prepared using liquid nitrogen freezing and thawing four times. After centrifugation at 45,000 ϫ g for 10 min at 4°C with a Beckman 100. 3 Ti rotor, about 10 A 260 units of the supernatant were layered onto a 5-40% sucrose gradient in buffer A and centrifuged at 35,000 ϫ g for 3 h at 4°C in a Beckman SW41 rotor. Gradients were analyzed with continuous monitoring at 254 nm. To analyze a 50 S and 30 S ribosome profile, another 10 A 260 units of supernatant as prepared above were dialyzed overnight against buffer B (10 mM Tris-HCl, pH 7.8, 60 mM NH 4 Cl, 0.5 mM MgCl 2 , and 1 mM DTT) (dialysis buffer was changed once) and then layered onto 10 -30% sucrose gradients in buffer B and centrifuged at 35,000 ϫ g for 3 h at 4°C in a Beckman SW41 rotor. Gradients were analyzed with continuous monitoring at 254 nm.
Primer Extension Analysis in Vivo-For primer extension analysis of mRNA cleavage sites in vivo, total RNAs were extracted from the E. coli BW25113 cells containing pBAD-YoeB at different time points as indicated in Fig. 3, C and D. Primer extension reactions were carried out for 1 h using different primers at 42°C using reverse transcriptase-avian myeloblastosis virus (2 units) as described previously (6).
Toeprinting Assays-Toeprinting was carried out as described previously (19) with a minor modification. The mixture for primer-template annealing containing mRNA and 32 P-end labeled DNA primer was incubated at 70°C for 5 min, and then cooled slowly to room temperature. The ribosome-binding mixture contained 2 l of 10ϫ buffer (100 mM Tris-HCl, pH 7.8, containing 100 mM MgCl 2 , 600 mM NH 4 Cl, and 60 mM 2-mercaptoethanol), different amounts of YoeB(His) 6   added. The cDNA synthesis was carried out at 37°C for 15 min. The reaction was stopped by adding 12 l of the sequencing loading buffer (95% formamide, 20 mM EDTA, 0.05% bromphenol blue, and 0.05% xylene cyanol EF). The sample was incubated at 90°C for 5 min prior to electrophoresis on a 6% polyacrylamide sequencing gel. The lpp and ompA mRNAs were synthesized in vitro from their individual DNA fragments containing a T7 promoter and part of their opening reading frames using T7 RNA polymerase. These ϳ200-bp DNA fragments for lpp (194 bp) and ompA (248 bp), all of which had the initiation codon at the center, were amplified by PCR using appropriate primers using chromosome DNA as templates, respectively. The 5Ј-end primers for lpp and ompA contained the T7 promoter sequence. The mazG DNA fragment containing a T7 promoter and part of the mazG open reading frame as described earlier (6) was obtained by PCR amplification using the pET-11a-MazG plasmid (15) as DNA template. mRNA produced from this DNA fragment was 151 bases in length.
RNA Isolation and Northern Blot Analysis-E. coli BW25113 containing pBAD-YoeB was grown at 37°C in LB medium. When the A 600 value reached 0.6, arabinose was added to a final concentration of 0.2%. The samples were taken at different intervals as indicated in Fig. 3, A and B. Total RNA was isolated using the hot-phenol method as described previously (20). Northern blot analysis was carried out as described previously (21).
Effect of YoeB(His) 6 Fig. 5F) were preincubated at 37°C for 10 min, and then 0.035 M 32 P-labeled 30-base RNA was added. The reaction mixture was incubated at 37°C for another 10 min and then the reaction product was loaded on a 20% sequencing gel.

YoeB Inhibits Protein Synthesis Immediately After Its
Induction-To identify the cellular functions inhibited by YoeB, a cell-free system prepared from E. coli BW25113 cells The sequence ladders shown at the right-hand side in all cases were obtained using the same primers used for toeprinting with their corresponding genes cloned in pCR 2.1-TOPO as template. All mRNA sequences shown are complementary to the sequencing ladders. The initiation codon, AUG, is indicated with an arrowhead. TP(Y) is the band where toeprinting was stopped in the presence of YoeB(His) 6 . FL, the full-length of the mRNA; and TP(r), the toeprinting site due to normal ribosome binding to mRNA in the absence of YoeB(His) 6 . The band at TP(S) is likely to be formed due to a secondary structure of the mRNA.
carrying arabinose-inducible pBAD-YoeB vector was used. The cells were permeabilized by toluene treatment (10,11). ATPdependent [ 35 S]methionine incorporation was completely inhibited when cells were preincubated for 10 min in the presence of arabinose before toluene treatment (Fig. 1A). However, the incorporation of [␣-32 P]dCTP (Fig. 1B) and [␣-32 P]UTP (Fig. 1C) was not significantly affected under similar conditions (12,13). These results demonstrate that YoeB inhibits protein synthesis, but not DNA replication or RNA synthesis. Notably, cell growth was inhibited almost immediately after the addition of arabinose (Fig. 1D). The in vivo incorporation of [ 35 S]methionine was also dramatically inhibited within 5 min after YoeB induction (Fig. 1, E and F). MazF is another potent E. coli toxin that functions as an mRNA interferase that specifically cleaves mRNAs at ACA triplet sequences and consequently leads to inhibition of protein synthesis (6). However, in contrast to YoeB, complete inhibition of protein synthesis by MazF requires significantly longer time (15-20 min) after its induction (6).
Inhibitory Effect of Purified YoeB on Cell-free Protein Synthesis-Next, we examined the effect of purified YoeB on E. coli cellfree protein synthesis. YoeB was purified from cells co-expressing both YefM and YoeB as described under "Experimental Procedures" (Fig. 2A). MazF was purified (Fig. 2B) as described previously (6). The synthesis of MazG protein (30 kDa) (15) from plasmid pET-11a-MazG was tested at 37°C for 1 h in the absence and presence of YoeB using an E. coli T7 S30 extract system (Promega) (Fig. 2C). MazG synthesis was almost completely blocked at YoeB concentrations of 240 nM or above. Similar inhibition was observed when purified MazF was added (Fig. 2D) consistent with the previous result (6).
We then tested the effect of YefM antitoxin on the YoeBmediated inhibition of MazG synthesis. The addition of GST-YefM rescued MazG synthesis in a dose-dependent manner (Fig. 2C). Importantly, YoeB did not inhibit eukaryotic cell-free protein synthesis (Fig. 2E, lane 4) in contrast to MazF (Fig. 2E, lane 2), which is known to inhibit both prokaryotic, and eukaryotic protein synthesis (6). These results indicate that YoeB is a protein synthesis inhibitor specific to prokaryotes. This is also consistent with the observation that YoeB induction in yeast cells had no effect on cell growth, whereas MazF induction blocked yeast cell growth (data not shown). It should be noted that YoeB alone could not degrade YefM mRNA under our experimental condition (Fig. 2F).
YoeB Induction Does Not Affect the Stability of Cellular mRNAs-Despite the abrupt inhibition of protein synthesis by YoeB, its induction did not severely affect the stability of cellular mRNAs. On the other hand, the induction of MazF significantly decreased stability of cellular mRNAs (6). Most significantly, full-length lpp and ompA mRNAs very quickly disappeared after MazF induction (10 min) (6), whereas all these mRNAs appeared to be more stable after YoeB induction (Fig. 3, A and B).
In Vivo Cleavage of the lpp and ompA mRNAs-As shown in Fig. 3, A and B, cellular mRNAs were partially degraded upon induction of YoeB. Therefore we next attempted to identify the cleavage sites of the chromosomally encoded lpp and ompA mRNAs by primer extension experiments. For this purpose, total RNA was extracted from E. coli BW25113 cells harboring pBAD-YoeB at different time intervals following induction of YoeB. The primer extension analyses of lpp (Fig. 3C) and ompA (Fig. 3D) mRNAs demonstrated that the distinct major bands exhibiting the specific cleavage sites in each mRNA appeared as early as 2.5 min after YoeB induction (Fig. 3, C and D, lane 2). The band intensities for ompA (Fig. 3D) mRNA further increased from 2.5 to 10 -20 min after YoeB induction. Interestingly, in both cases, the major bands resulted from cleavage of the mRNAs at three bases downstream of AUG, and most notably no other bands were observed in the regions between 5Ј-ends of the mRNAs and the initiation codon suggesting that YoeB may function only when it associates with the ribosomal translation machine. It is interesting to note that the substantial amounts of the full-length mRNAs for lpp (Fig. 3C) and ompA (Fig. 3D) mRNAs remained uncleaved even 30 min after induction of YoeB. This observation is again consistent with the fact that unlike MazF, YoeB does not significantly affect the stability of cellular mRNAs.
YoeB Binds to the Translation Initiation Complex in Vitro-Next, we examined if similar cleavage of lpp and ompA mRNAs was observed in vitro by toeprinting (TP) (19). In this experiment, we also used the mazG mRNA (Fig. 4C), which we previously used in our experiments with MazF (6) in addition to lpp (Fig. 4A) and ompA (Fig. 4B) mRNAs. Primer extension analysis of the mRNAs alone yielded the full-length bands (Fig. 4, A-C, lane 1). Importantly, the addition of YoeB alone did not result in formation of new bands indicating that YoeB by itself has no endoribonuclease activity under the conditions used (Fig. 4, A-C, lane 2). When 70 S ribosomes and initiator tRNA f Met were added to the reaction mixtures, typical toeprinting band (TP(r)) downstream of the initiation codon appeared (Fig. 4, A-C, lane  3). On the other hand, when YoeB was added together with 70 S ribosomes and initiator tRNA f Met , a new band TP(Y) (Fig. 4, A-C, lane 4) appeared with concomitant disappearance of the TP(r) band in all cases. This new band, TP(Y), is 10 bases upstream of the normal toeprinting band (13-14 bases downstream of the initiation codon) as judged from the sequence  6 . The band at TP(S) in Fig. 5, A-C, is likely to be formed due to a secondary structure of the mRNA. D, toeprinting of the ompA mRNA after phenol extraction and toeprinting of the ompA mRNA in the presence of YoeB(His) 6 and 70 S ribosomes with and without tRNA f Met . The experiment was carried out in the same way as described in the legend to

YoeB Inhibits Translation Initiation
ladder at the right-hand side. It should be noted that this band is observed only in the presence of 70 S ribosomes, initiator tRNA f Met , and YoeB (Fig. 4, A-C, lane 4). Interestingly, this TP(Y) band disappeared with concomitant appearance of the TP(r) band when YefM, the YoeB antitoxin, was added together with YoeB (Fig. 4C, lane 5). The band below the full-length band shown by TP(S) was probably due to the secondary structure present in the mazG mRNA because this band was detected in the absence of any factors (Fig. 4C, lane 1).
Next, we tested whether the TP(Y) and TP(r) bands resulted from the cleavage of mRNAs using mazG and ompA mRNAs. These mRNAs were first preincubated without (Fig. 5, A, lane 3,  and D, lane 7) or with YoeB (Fig. 5, A, lane 5 for the mazG mRNA, and D, lane 8 for the ompA mRNA) in the presence of 70 S ribosomes and initiator tRNA f Met . RNAs were then phenol extracted and used for primer extension shown in Fig. 5A (lanes  4 and 6) for the mazG mRNA and Fig. 5D (lanes 9 and 10) for the ompA mRNA. The TP(Y) band was observed even after phenol extraction (compare Fig. 5, A, lanes 5 and 6 for the mazG mRNA, and D, lanes 8 and 9 for the ompA mRNA), indicating that the TP(Y) band resulted from at least partial cleavage of the mazG and ompA mRNAs. In contrast, the TP(r) band disappeared after phenol extraction, indicating that this band was not due to mRNA cleavage, but was caused by binding of ribosomes to the mRNAs. It is important to note that the mRNAs were cleaved three and four bases downstream of the translation initiation codon only in the presence of ribosomes, initiator tRNA f Met , and YoeB, but not in the presence of YoeB alone. Therefore, the latent endogenous endoribonuclease activity of 70 S ribosomes (5) may be enhanced only when YoeB binds to the ribosome initiation complex. The ribosome-dependent endoribonuclease activity induced by RelE preferentially cleaves mRNA at stop codons (UAG Ͼ UAA Ͼ UGA) and at certain sense codons (UCG and CAG) in in vitro experiments (4). Notably, the primary sequence of YoeB has only 15% identity with that of RelE. Comparison of the x-ray structures of the YoeB-YefM complex (3) and the RelE-RelB complex (26) demonstrates little homology between these complexes, although they share a microbial RNase fold (3). The possibility that YoeB may function as an endoribonuclease when it binds to the ribosome initiation complex is unlikely because a nuclease-negative YoeB mutant H83Q (3) was still able to block protein synthesis as shown later.
When the Shine-Dalgarno (SD) sequences, GGAG in the mazG mRNA (Fig. 5B, lanes 1-4) and GGAG in the ompA mRNA (Fig. 5E, lanes 1-4) were mutated to AAUG (Fig. 5B,  lanes 5-8) and GAAA (Fig. 5E, lanes 5-8), respectively, both TP(Y) and TP(r) completely disappeared (compare Fig. 5, B,  lanes 4 and 8, for the mazG mRNA, and E, for the ompA  mRNA). These results indicate that the binding of 70 S ribosomes to the mRNA is absolutely required for the YoeB-mediated cleavage of mRNAs.  (Fig. 5D). These results suggest that tRNA f Met was not required for YoeB-mediated A site cleavage. Note that both TP(Y) bands for mazG and ompA (Fig. 5, C and D, lane 6, respectively) were still detectable after phenol extraction (data not shown).
Cleavage of a Synthetic 30-Base RNA-To further unambiguously demonstrate that the mRNA cleavage occurring downstream of the initiation codon requires both ribosomes and YoeB, we repeated a similar experiment as above using a 30-base synthetic RNA that contains an SD sequence and AUG (see Fig. 5F). When YoeB was incubated with 70 S ribosomes, tRNA f Met , and the 5Ј-end 32 P-labeled 30-base RNA at 37°C for 10 min, a new band appeared (Fig. 5F, lanes 6 -8) (lane 5) were incubated with the 30-base RNA. As judged by using an RNA ladder, the new band is 7 bases shorter than the 30-base RNA. This result is consistent with the results seen with mazG (Fig. 5, A-C) and ompA (Fig. 5D)  Met from the ribosome in vitro. For this purpose, YoeB was added at the minimal concentration that is just enough to produce the TP(Y) band without forming the TP(r) band. As the amounts of YoeB added in the reaction mixture were reduced (Fig. 6A, from lanes 4 to 10), the intensities of the TP(Y) band were progressively reduced, whereas the intensities of the TP(r) band increased gradually to reach the normal level.
We also tested if tRNA f Met bound to the ribosomes is displaced with YoeB under the condition as that used in Fig. 6A. For this purpose, [ 35 S]Met-tRNA f Met was first incubated with 70 S ribosomes and mRNA at 37°C for 10 min, and then different amounts of YoeB as used in Fig. 6A were added. The reaction mixtures were incubated at 37°C for another 10 min. The reaction mixtures were then applied to nitrocellulose filters (Millpore 0.45 M HA) and the filters were washed twice with 2 ml of buffer A before measuring the radioactivity retained on the filters. As shown in Fig. 6C Met was released at 10 nM YoeB (the ratio of YoeB to ribosome is 1:5). It should be noted that ribosomebound initiator tRNA drops from 72 to about 42, upon addition of 0.04 nM YoeB and further to about 30 upon addition of 3 nM YoeB. Because the major part of the decrease in ribosome bound initiator tRNA occurred at substoichiometric additions of YoeB, this result is not due to direct competition between initiator tRNA and YoeB. It is more likely due to indirect competition caused by mRNA cleavage. The binding stability of initiator tRNA to 70 S ribosomes is known to be coupled to that of mRNA, and therefore if mRNA is cleaved, the binding of mRNA to 70 S ribosomes should be reduced, which in turn, likely destabilizes initiator tRNA binding (Fig. 6C).
In a reciprocal experiment, where increasing amounts of tRNA f Met were added in the reaction mixture, the YoeB-mediated mRNA cleavage activity was diminished (Fig. 6B), However, the TP(r) band did not appear, further suggesting that YoeB and initiator tRNA appear to indirectly compete for binding to 70 S ribosomes. It should be noted that the use of Tetracycline and Hygromycin B Inhibit YoeB-mediated mRNA Cleavage-To identify the exact YoeB-acting site, we next examined the effect of tetracycline and hygromycin B on YoeB-mediated mRNA cleavage. Both tetracycline and hygromycin B are known to bind to the A site in ribosomes (23). tRNA f Met was first incubated with 70 S ribosomes and mRNA at 37°C for 10 min to form the translation initial complex, and then tetracycline (Fig. 6G, lanes 2 and 4) and different amounts of hygromycin B (Fig. 6H, lanes 2, 3, 5, and 6) were added. The reaction mixtures were incubated at 37°C for another 10 min. YoeB was then added to the reaction mixtures for another 10 min (Fig. 6, G, lanes 3 and 4, and H, lanes 4 -6). Reverse transcriptase was finally added to initiate cDNA synthesis. As shown in Fig. 6G, 160 M tetracycline (the ratio of tetracycline to 70 S ribosomes is 3250 to 1) inhibited YoeB-mediated mRNA cleavage at the A site with concomitant appearance of the TP(r) band (Fig. 6G, lane 4). With hygromycin B, two different concentrations (0.64 mM for lanes 2 and 5 and 1.6 mM for lanes 3 and 6) were used. As shown in Fig. 6H Fig. 6H). However, 1.6 mM hygromycin B (the ratio of hygromycin B to 70 S ribosomes is 32500) almost completely blocked YoeB-mediated mRNA cleavage activity (Fig. 6H, compare lane 4 with 6). At 0.64 mM hygromycin B, the mRNA cleavage was partially inhibited with concomitant appearance of the TP(r) band (Fig. 6H, lane 5). Note that at 1.6 mM hygromycin B, the intensity of the TP(r) band was almost identical to that in the absence of both hygromycin and YoeB (Fig. 6H, compare lane 6 with 1). These results are consistent with the result obtained with tetracycline in Fig.  6G, suggesting that hygromycin B binding to A sites interferes with YoeB-mediated mRNA cleavage at the A site.
Cleavage of mRNA at the A Site Codons Releases the 3Ј-End Portion of the mRNA from 70 S Ribosomes-To examine if the cleavage of mRNA results in the release of the 3Ј-end portion of the mRNA from 70 S ribosomes, the reaction mixture after complex formation with the use of the ompA mRNA was centrifuged at 90,000 ϫ g at 4°C for 1 h to pellet ribosomes. Subsequently, we determined if the 3Ј-end portion of the mRNA was released into the supernatant by primer extension reaction. As shown in Fig. 6E, the TP(r) band was not detectable in the supernatant fraction (lane 3), whereas the TP(Y) band was almost fully recovered in the supernatant fraction (lane 5).
These results indicate that the TP(Y) band detected in the toeprinting experiments was derived from the 3Ј-end portion of the ompA mRNA, subsequently released into the soluble fraction (or into the cytoplasmic fraction in the cells). Using the same 30-base synthetic RNA as used in Fig. 5F, we examined if the 5Ј-end portion of the RNA dissociates from ribosomes in the presence of YoeB. Although the RNA is cleaved at 7 bases from the 3Ј end (see Fig. 5F), the remaining 23-base RNA (labeled at its 5Ј end with 32 P) was bound to the ribosomes even at a very high concentration of YoeB (70 times higher concentration than that used in Fig. 6C; see Fig. 6F), indicating that the 5Ј-end portion of mRNA stays bound to 70 S ribosomes even after cleavage of the mRNA.

Effect of a Nuclease-negative Mutation (H83Q) in YoeB on the Cell Growth and Protein Synthesis-H83Q, a nuclease negative
YoeB mutant, previously reported by Kamada and Hannoka (3) was also constructed to examine its effect on cell growth. Induction of the YoeB(H83Q) mutant inhibited cell growth and the addition of YoeB(H83Q) also inhibited MazG protein synthesis in vitro (data not shown). When the toeprinting experiment was carried out using the purified YoeB(H83Q), toeprinting bands (TP(Y)) three and four bases downstream of the initiation codon were detected in the manner identical to that of the wild-type YoeB (data not shown). Thus, YoeB(H83Q), a nuclease negative mutant, does not affect YoeB-mediated mRNA cleavage at the A site. Notably a higher concentration of the mutant YoeB than the wild-type YoeB was required to result in the same toeprinting effect as caused with the wild-type YoeB, indicating the mutant YoeB seems to have a weaker affinity to 70 S ribosomes than the wild-type YoeB.
YoeB Associates with 50 S Ribosomal Subunits-Because YoeB binds to the translation initiation complex (Fig. 4), we examined whether YoeB affects the polysome profiles and also specifically associates with one of the ribosomal subunits. When the polysome pattern of E. coli BW25113 cells carrying the pBAD-YoeB plasmid was analyzed by sucrose density gradient at 10 min after induction of YoeB by arabinose, a significant increase in the 70 S ribosomal fraction was observed without change in the 30 S and 50 S ribosomal profiles (compare Fig.  7B with Fig. 7A). Western blotting analysis showed that YoeB was associated with both 70 S ribosome and 50 S ribosome fractions, but not with 30 S ribosome (Fig. 7B). Cell lysates were prepared in 0.5 mM Mg 2ϩ to dissociate 70 S ribosomes to 50 S and 30 S ribosomes and analyzed by sucrose density gradient centrifugation (Fig. 7C), where YoeB was detected only in the 50 S ribosome fraction, demonstrating that YoeB is a 50 S ribosome associating protein.

DISCUSSION
In the present paper, we demonstrated that YoeB is a potent inhibitor of protein synthesis, which functions in a manner that is distinctly different from that of MazF, an mRNA interferase (6). Upon YoeB induction, [ 35 S]methionine incorporation into cellular proteins was abruptly stopped within less than 5 min (Fig. 1, E and F). This is in sharp contrast to MazF induction, which stops [ 35 S]methionine incorporation only after 15-20 min of its induction (6). Another important difference between YoeB and MazF is that significant amounts of full-length cellular mRNAs remain in the YoeBinduced cells even 30 min after its induction (Fig. 3, A and B), whereas they rapidly disappear after 10 min upon MazF induction (6). In addition, MazF inhibits protein synthesis in both prokaryotes and eukaryotes as expected from its function as an mRNA interferase (6), whereas YoeB was found to be a prokaryote-specific inhibitor (Fig. 2, C and E).
On the basis of these results, we speculated that YoeB does not primarily function as a sequence-specific endoribonuclease or an mRNA interferase as originally proposed by Christensen et al. (1) and Kamada and Hanaoka (3). To substantiate our hypothesis, we carried out in vivo primer extension experiments using chromosomally encoded mRNAs as targets rather than mRNAs encoded by genes cloned in a multicopy plasmid to avoid artifacts (Fig. 3, C and D). Interestingly, both mRNAs tested were detected either as the corresponding full-length mRNAs or as a product cleaved at three bases downstream of the initiation codon (Fig. 3, C and D). Notably, in the ϳ200nucleotide 5Ј leader sequences for both mRNAs, no other cleavage bands were detected. These mRNAs encoded outer membrane proteins and were used in our experiment because of their abundance in the cell. However, the mRNA from the glnH gene that encodes a cytoplasmic protein was also cleaved at three bases downstream of AUG (data not shown), indicating that the cleavage of mRNAs by YoeB does not depend on the type of these mRNAs.
Importantly, this in vivo observation was consistent with the in vitro toeprinting experiments using 70 S ribosomes, tRNA f Met , purified YoeB, and the same mRNAs (Fig. 4). We demonstrated that the addition of YoeB caused the upstream shift of the stop site for primer extension (toeprinting site) from normal 13 and 14 bases downstream of AUG to three and four bases from AUG. This apparent 10-base upward shift resulted from the cleavage of the mRNA at the A site mediated by YoeB bound to 70 S ribosomes (Fig. 5, A, lane 6, and D,