SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

tmRNA, a product of ssrA gene, plays a crucial role in the quality control system that eliminates aberrant products of nonstop mRNAs in prokaryotes. Although tmRNA recycles ribosomes stalled at the 3′ end of nonstop mRNAs, the fate of ribosomes that stall at the 3′ end in the absence of tmRNA has not been extensively examined. Here we report our analysis of the translation status of nonstop mRNAs. Polysome analysis showed that nonstop mRNAs were translated efficiently, and peptidyl-tRNA was not found in any fraction in a ΔssrA strain. In vitro translation experiments using PURESYSTEM revealed that ribosomes translating nonstop mRNAs were dissociated from the 3′ end of mRNA, and the peptidyl-tRNA was only weakly hydrolyzed in the monosome. These results suggest that the peptidyl-tRNA of a nonstop mRNA is hydrolyzed by an unknown factor(s) in vivo, thereby allowing a nonstop mRNA to be translated as efficiently as a normal mRNA. Possible factors involved in the hydrolysis of the peptidyl-tRNAs of nonstop mRNAs are discussed.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The termination of translation at an aberrant position can result in the production of harmful protein products. Therefore, cells have quality control systems that recognize aberrant translation termination codons in order to eliminate these aberrant mRNAs (Isken & Maquat 2007). In eubacteria, tmRNA is recruited to the empty A-site of the ribosome, where it acts first as an alanyl-tRNA and then as an mRNA to direct the addition of a short peptide tail to the polypeptide (Keiler et al. 1996; Karzai et al. 2000). This trans-translation terminates at a stop codon contained in the tmRNA reading frame, and the tagged polypeptide is rapidly degraded by proteases (Gottesman et al. 1998).

The level of nonstop mRNAs increases when ssrA, which encodes a small stable tmRNA, is deleted, indicating that tmRNA dissociates the ribosome from nonstop mRNAs and facilitates the degradation of truncated mRNAs. We proposed that a tmRNA quality control system not only degrades aberrant polypeptides once produced but also prevents the production of aberrant polypeptides by rapidly eliminating damaged mRNAs (Yamamoto et al. 2003). We also found that similar levels of proteins were produced from a nonstop mRNA and a control mRNA in the absence of tmRNA (Sunohara et al. 2004a). These data suggest that nonstop mRNAs are translated as efficiently as normal mRNAs in the absence of tmRNA and that abnormal proteins are degraded by trans-translation to prevent their expression in vivo. This translatability of nonstop mRNAs indicates the necessity of efficiently dissociating ribosomes that are translating nonstop mRNAs. On the other hand, the stabilization of nonstop mRNAs in ΔssrA suggests that ribosomes may stall at the 3′ end of nonstop mRNA and protect them from exonuclease attack. The efficient translation and stability of nonstop mRNAs seem to be inconsistent. Therefore, the fate of a ribosome stalled at the 3′ end of a nonstop mRNA must be analyzed in order to understand the translatability and stability of nonstop mRNAs in Escherichia coli.

It has been shown that an aberrant mRNA lacking a termination codon (nonstop mRNA) is rapidly degraded by the nonstop mRNA decay pathway (NSD) in eukaryotes (Frischmeyer et al. 2002; van Hoof et al. 2002). In yeast, Ski7p is proposed to recognize a stalled 80S ribosome at the 3′ end of a nonstop mRNA and recruit the exosome and Ski complex, resulting in rapid degradation of the nonstop mRNA. We have previously shown that the level of products produced from a nonstop mRNA containing a poly(A) tail was reduced 100-fold and that this reduction was due to rapid mRNA degradation, translation repression and protein destabilization, at least in part, by the proteasome (Inada & Aiba 2005; Ito-Harashima et al. 2007). Insertion of a poly(A) tract upstream of a termination codon resulted in translation repression and protein destabilization but not rapid mRNA decay (Inada & Aiba 2005; Ito-Harashima et al. 2007), indicating that translation of the poly(A) sequence results in translation arrest followed by co-translational protein degradation, thereby repressing expression of the aberrant protein. Consistent with this finding, it was shown that a nonstop mRNA lacking a poly(A) tail was translated efficiently in yeast (Meaux & Van Hoof 2006), indicating that a nonstop mRNA without a poly(A) tail could be translated in yeast as well as in E. coli.

In this study, we directly examined the translation status of bacterial nonstop mRNAs in an ssrAΔ E. coli strain by polysome analysis and found that nonstop mRNAs were distributed mainly in the polysome fractions. The translation products of nonstop mRNAs were not detected in the polysome fractions but in the ribosome-free fractions. These data suggest that nonstop mRNAs are translated efficiently in the ssrAΔstrain and that a peptidyl-tRNA of the nonstop mRNA is hydrolyzed in vivo. In vitro translation with PURESYSTEM revealed that the peptidyl-tRNA of the nonstop mRNA was hydrolyzed less efficiently than that of the stop-containing mRNA and was distributed only in the monosome or polysome fractions. These results suggest that the peptidyl-tRNA of the nonstop mRNA is hydrolyzed by an unknown factor(s) in vivo that is not present in PURESYSTEM and indicate that nonstop mRNA is translated as efficiently as normal mRNA in vivo. Based on these results, we discuss possible factors and mechanisms involved in the hydrolysis of the peptidyl-tRNA of nonstop mRNA.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Polysome analysis of nonstop mRNA in vivo

Previous reports have shown that ribosome stalling during translation elongation results in endonucleolytic cleavage of the mRNA (Hayes & Sauer 2003; Sunohara et al. 2004a,b; Baker & Parker 2006) and this cleavage results in the production of a nonstop mRNA. We previously found that the nonstop mRNA was produced when ribosomes translates the crp-GIRAGP-crr reporter mRNA that contains the arrest-inducible sequence derived from secM, and the protein levels produced from a nonstop mRNA (truncated crp mRNA) were comparable to that of a control mRNA (Sunohara et al. 2004a). Because these results suggested that nonstop mRNAs could be translated efficiently in vivo, we examined the translation status of nonstop mRNAs in the absence of tmRNA in E. coli. Because polysome analysis is a well-established, standard method to analyze the translation status of mRNA, we used polysome analysis of E. coli cell extracts to determine the translation efficiency of truncated mRNAs in vivo.

Cells harboring the control pHA7 (WT) or pJK208 (nonstop) plasmid were grown in LB medium and chloramphenicol (100 mg/mL) to inhibit translation elongation. Collected cells were frozen in liquid nitrogen, and extracts were prepared by grinding. Polysome analysis was performed by centrifuging the cell extracts on 10%–40% sucrose gradients. After centrifugation, mRNA samples were prepared from each fraction and analyzed by Northern blotting. As shown in Fig. 1a, wild-type mRNAs were mainly distributed in the polysome fractions. Quantification of mRNA levels in each fraction revealed that more than 70% of wild-type crp mRNAs were distributed in the polysome fractions (Fig. 1a, bottom panel). These findings indicate that the crp mRNA is translated multiple times. The distribution of crp-crr and truncated mRNAs in extracts of ΔssrA cells harboring pJK208 was also examined. We found that the majority of crp-crr mRNAs (~60%) were distributed in the polysome or monosome fractions, indicating that the crp-crr mRNA is actively translated (Fig. 1d, lanes 10–15). We also found that more than 60% of the truncated mRNAs were distributed in the monosome or polysome fractions (Fig. 1d, lanes 6–15), while some were in the ribosome-free fractions. We found that the distribution of truncated mRNAs was essentially the same as that of crp mRNA (Fig. 1a, bottom panel, lanes 6–15), suggesting that truncated mRNAs, like wild-type mRNAs, are translated multiple times in vivo. These results suggest that ribosomes dissociate more efficiently from truncated mRNAs than non-truncated mRNAs. The significant level of truncated mRNAs in the ribosome-free fractions suggests that a ribosome efficiently dissociates from a nonstop mRNA and does not require stimulation by another ribosome on the same mRNA.

image

Figure 1. Polysome analysis of nonstop mRNA in the ΔssrA background. (a) Polysome analysis was performed with TA501 (ΔcrpΔssrA) cell extracts harboring the control pHA7 plasmid. Cell extracts were prepared as described in the Experimental procedures, and polysomes were separated by centrifugation on linear 10%–40% sucrose density gradients. Top panel: Ribosome profile with continuous monitoring of absorbance at 254 nm. Middle panel: Distribution of mRNAs in the sucrose gradients. mRNA samples were isolated from each fraction, and the crp mRNA levels in each fraction were quantified by Northern blot analysis with DIG-labeled crp probe. The DIG-labeled DNA probe used was a 576-bp probe corresponding to the crp coding region. Bottom panel: The relative levels of mRNAs in the fractions were shown as proportion to the total mRNA level, which was assigned a value of 100. The band intensities were quantified by LAS3000mini (Fuji Film, Japan). (b) Distribution of proteins in the sucrose gradients. Protein samples of TA501 (ΔcrpΔssrA) cell extracts harboring the control pHA7 plasmid were isolated from each fraction of the sucrose gradient. CRP proteins were detected by Western blotting. The asterisk denotes a nonspecific signal. The positions of the CRP proteins are indicated by arrows. (c) The samples used in (b) were analyzed by NuPAGE and subsequent Western blot analysis. The positions of CRP and peptidyl-tRNA proteins are indicated by arrows. The peptidyl-tRNA product derived from the nonstop crp mRNA with an in vitro translation reaction shown in Fig. 3b lane 10 was loaded in lane 14. (d) Polysome analysis was performed with TA501 (ΔcrpΔssrA) cell extracts harboring pJK208. Polysome analysis was performed as shown in (a). Middle panel: Distribution of the mRNAs in sucrose gradients. Bottom panel: Relative levels of mRNA in the fractions. The band intensities were quantified as in (a). (e) Distribution of proteins in the sucrose gradients. Protein samples of TA501 (ΔcrpΔssrA) cell extracts harboring pJK208 were isolated from each fraction of the sucrose gradient. CRP and CRP-GIRAGP-IIAGlc proteins were detected by Western blotting. (f) The samples in (e) were analyzed with NuPAGE followed by Western blot analysis.

Download figure to PowerPoint

To examine the status of ribosomes that are translating nonstop mRNAs, we analyzed the protein products of the nonstop mRNA in each fraction. The proteins produced from the nonstop mRNA (Fig. 1e, lanes 1–2) as well as the wild type mRNA (Fig. 1b, lanes 1–2) were detected only in the ribosome-free fractions, indicating that there is no stable ribosome-nonstop mRNA-peptidyl-tRNA complex in the polysome fractions. To determine whether the products in the ribosome-free fractions were peptidyl-tRNA, we performed Western blot analysis after NuPAGE. It was previously shown that peptidyl-tRNA migrated stably in NuPAGE but not in conventional SDS-PAGE (Onouchi et al. 2005), indicating that NuPAGE is a simple method to detect the peptidyl-tRNA. We also detected the peptidyl-tRNA derived from the nonstop mRNA produced with an in vitro translation reaction using this procedure (Fig. 1c,f, lane 14). There was no significant difference between the SDS-PAGE (Fig. 1b,e) and NuPAGE (Fig. 1c,f) results, suggesting that there was no peptidyl-tRNA derived from the nonstop mRNA in the ribosome-free fractions. However, the migration of full-length proteins produced from the crp-crr mRNA was almost the same as that of the nascent peptidyl-tRNA of the nonstop mRNA by NuPAGE (Fig. 1f, lanes 1 and 14). Therefore, it is difficult to conclude that no peptidyl-tRNA is derived from nonstop mRNAs in vivo. To detect clearly the nascent peptidyl-tRNA of nonstop mRNA, we constructed a new reporter gene as shown in Fig. 2a.

image

Figure 2. The pth(ts) mutation had no effect on the protein levels of the nonstop mRNA. (a) A schematic diagram of the pIT815 reporter gene. The black box indicates the insert encoding GIRAGP and the ssrA tag sequence. The ssrA tag sequence is also underlined. (b) TA341 (Δcrp; AA), TA481(ΔcrpssrADD; DD), or TA501 (ΔcrpΔssrA; Δ) cells harboring pIT815 were grown in LB medium, and samples were prepared for Western blot analysis. (c) Growth curve of the pth(ts) mutant and parental strain. Parental TA341 (Δcrp) cells (boxes) or IT1534 (pth(ts)ΔssrA) (circles) harboring pIT815 were grown in LB medium supplemented with ampicillin. Cells were grown at 32 °C or 39 °C, and the optical density at 600 nm (OD600) was monitored. (d) Detection of the peptidyl-tRNA derived from the nonstop mRNA in the pth(ts) mutant. The nonstop mRNA product levels in pth(ts) (TS) and its parental strain (WT) were analyzed by Western blot with anti-CRP antibodies.

Download figure to PowerPoint

Role of Pth in releasing the peptide from the peptidyl-tRNA of a nonstop mRNA

To detect only the proteins produced from nonstop mRNAs, both an SsrA-tag sequence and the arrest sequences (GIRAGP) was inserted just upstream of the authentic termination codon, and the resulting plasmid is pIT815. When translation elongation is temporally arrested by the arrest sequences (GIRAGP), a nonstop mRNA is produced and a truncated protein is tagged by trans-translation and rapidly degraded. When translation of the reporter mRNA is terminated at the original stop codon, the protein product contains the SsrA-tag sequence at its carboxy terminus and is rapidly degraded. Therefore, we suspected that proteins produced only from nonstop mRNAs may be detected only in the absence of tmRNA. As expected, nonstop mRNA products were detected in the ΔssrA but not the ssrA+strain (Fig. 2b, lanes 1–3). We performed NuPAGE and subsequent Western blot analysis and found that no peptidyl-tRNA was derived from the nonstop mRNA in ΔssrA cells harboring the pIT815 plasmid (Fig. 2d, lanes 3 and 5).

It is well known that a short peptidyl-tRNA can prematurely dissociate from the ribosome during translation elongation, and peptidyl-tRNA can be hydrolyzed by Pth, a peptidyl-tRNA hydrolase (Garcia-Villegas et al. 1991; Ontiveros et al. 1997; Karimi et al. 1998; Heurgue-Hamard et al. 2000). It was shown that the level of peptidyl-tRNAs that dissociates from ribosomes increased in the ΔssrA strain and it was proposed that there is a connection between peptidyl-tRNA drop-off and trans-translation (Singh & Varshney 2004). Therefore, we suspected that the peptidyl-tRNA derived from the nonstop mRNA used in this study could drop off the ribosome in the absence of tmRNA. To address this possibility, we examined the expression of the nonstop mRNA in the temperature-sensitive pth mutant at the permissive and semi-restrictive temperatures. We confirmed the temperature sensitive growth of the pth(ts) mutant (Fig. 2c), and examined the expression of the protein produced from the nonstop mRNA by NuPAGE and subsequent Western blotting. As shown in Fig. 2d, the peptidyl-tRNA derived from the nonstop mRNA was minimally detected, even when samples were prepared from cells grown at the semi-restrictive temperature (Fig. 2d, lane 6). These results did not support the possibility that peptidyl-tRNA dissociates from ribosomes translating the nonstop mRNA used in this study in vivo.

Hydrolysis of peptidyl-tRNA during translation of nonstop mRNAs in vitro

Polysome analysis using cell extracts strongly suggested that translation might terminate without a termination codon in vivo. To determine the mechanism that terminates translation of nonstop mRNAs, we performed in vitro translation of the nonstop mRNA using the PURESYSTEM (Shimizu et al. 2001). In this system, a linear DNA fragment is used as a template and translation is coupled with transcription. The construct of the mRNAs expressed from the PCR products is shown in Fig. 3a, and we expected that ribosomes may stall just at the end of the nonstop mRNA with the CG dinucleotide in the A-site and the last codon, ACG, in the P-site.

image

Figure 3. Analysis of nascent peptides derived from the nonstop mRNA produced by in vitro translation. (a) Schematic drawing of the wild-type (stop) and nonstop (nonstop) mRNA used in the in vitro PURESYSTEM. The boxes indicate the open reading frame. The DNA sequences of the carboxyl terminal region are shown. The translation termination codon of the crp gene is indicated by an asterisk. (b) Top panel: kinetics of translation of wild type mRNA (stop) and nonstop mRNA (nonstop) with PURESYSTEM. Linear PCR fragments (10 pmole) were used for the in vitro translation reaction, and 8 mL of samples were analyzed by NuPAGE and subsequent Western blotting. In this neutral gel, the peptidyl-tRNA was stable and its migration was slower than that of a nascent peptide of 20 kDa. The asterisk indicates a nonspecific signal. Bottom panel: the band intensities were quantified, and the relative product levels are indicated. Open circles: the products of wild type mRNA; open boxes: total products of the nonstop mRNA; closed boxes: CRP-tRNA; closed circles: CRP. (c) Top panel: kinetics of the in vitro translation of the nonstop mRNA with PURESYSTEM (Δ2). Linear PCR fragments (10 pmole) were used for the in vitro translation reaction, and the products were analyzed as shown in (a). Bottom panel: the band intensities were quantified, and the relative product levels are indicated. Open boxes: total products of the nonstop mRNA; closed boxes: CRP-tRNA; closed circles: CRP. (d) Kinetics of ribosome association with the nonstop mRNA in vitro. Polysome analysis was performed with the in vitro translation reaction samples using the nonstop mRNA as a template. Ribosome profiles with continuous monitoring of absorbance at 254 nm.

Download figure to PowerPoint

We analyzed the protein products of the control (stop) or nonstop mRNA (nonstop) by Western blotting using anti-CRP antibodies. Because peptidyl-tRNA production was expected, the protein products were analyzed with NuPAGE. An analysis of the reaction revealed that the total amount of products, including CRP and peptidyl-tRNA, increased until 60 min in the experiments with two templates (Fig. 3b, top panel). When the control template was used, the peptidyl-tRNA (CRP-tRNA) was not detected, indicating that the peptidyl-tRNA was hydrolyzed efficiently during in vitro translation using PURESYSTEM (Fig. 3b, lanes 1–5 and bottom panel). When the nonstop template was used, peptidyl-tRNA (CRP-tRNA) production was followed by the production of polypeptide (CRP) resulting from the hydrolysis of peptidyl-tRNA (Fig. 3b, top panel). These data indicate that the peptidyl-tRNA of the nonstop mRNA was hydrolyzed less efficiently than that of the control mRNA (Fig. 3b, lanes 6–10 and bottom panel).

How is the peptidyl-tRNA derived from the nonstop mRNA hydrolyzed during in vitro translation with PURESYSTEM? It is possible that the peptide derived from the nonstop mRNA is released from the tRNA by conventional translation termination factors, but that the amount of release factors was enough to release only part of the peptides from the tRNA. To examine this possibility, we performed in vitro translation with PURESYSTEM (Δ2), which does not contain peptide chain release factors (RF1 and RF2). PCR products of the normal or nonstop reporter gene were incubated with PURESYSTEM (Δ2). The rate of peptidyl-tRNA hydrolysis was similar to that with in vitro translation using PURESYSTEM (Fig. 3c, bottom panel), suggesting that the hydrolysis of the peptidyl-tRNA of the nonstop mRNA occurs independently of these peptide release factors.

To investigate the translation status of nonstop mRNAs in vitro, especially the release of ribosomes from nonstop mRNAs, we performed a polysome analysis of the in vitro translation samples. The in vitro translation was stopped by the addition of chloramphenicol at the indicated times, and the samples were analyzed by centrifugation over 10%–40% sucrose gradients. As shown in Fig. 3d, several ribosomes were associated with an individual nonstop mRNA at 10 min after the initiation of the reaction, and the amount of polysome fractions increased until 20 min but decreased at 30 min and was almost completely abolished at 60 min. These results clearly indicate that a ribosome translating a nonstop mRNA dissociates from the mRNA. Because the proteins were synthesized most efficiently 10–20 min after initiation (Fig. 3b), the amount of polysomes correlates with the protein synthesis activity. Because the decrease in polysomes indicates that a new round of translation has stopped, these data suggest that a ribosome efficiently dissociates from a nonstop mRNA and does not require stimulation by another ribosome on the same mRNA.

Identification of a stalled ribosome complexed with nonstop mRNA and peptidyl-tRNA

To determine the fate of the peptidyl-tRNA derived from the nonstop mRNA, the distribution of the peptidyl-tRNAs and polypeptides derived from the nonstop mRNA in the polysome fractions was examined. Samples from a 30-min in vitro translation with control or nonstop mRNA were examined by polysome analysis followed by Western blotting. As expected, the ribosomes do not co-sediment with peptidyl-tRNAs or polypeptides derived from the control mRNA. This clearly indicates that the peptidyl-tRNA is rapidly and efficiently hydrolyzed by peptide chain release factors (RFs) on the ribosome, and indicates that hydrolysis of the peptidyl-tRNAs derived from nonstop mRNAs is much less efficient than that of normal mRNAs in vitro. Northern analysis revealed that the nonstop mRNAs were distributed in the monosome and polysome fractions (Fig. 4b). Therefore, the monosome fractions must contain ribosomes that stall at the 3′ end of the nonstop mRNA and contain the peptidyl-tRNA at the P-site (Fig. 4d, lanes 7–8). The other ribosome complex, which contains only peptidyl-tRNA but not mRNA, may exist in the monosome fractions, but the population of these two ribosome complexes is difficult to estimate under our conditions. We found that CRP proteins derived from the nonstop mRNA were not only in the ribosome-free fractions but also in the monosome fractions (Fig. 4d, lanes 7–8). This result clearly indicates that the peptidyl-tRNA is hydrolyzed on the ribosome during in vitro translation and that the nascent peptide is temporally associated with the ribosome after hydrolysis. We also conclude that the peptidyl-tRNA is not dissociated, but hydrolyzed on the ribosome probably after the mRNA is released from the ribosome in vitro.

image

Figure 4. Identification of a stalled ribosome complexed with the nonstop mRNA and peptidyl-tRNA. (a,b) Polysome analysis was performed with the in vitro translation samples using the normal mRNA (a) or nonstop mRNA (b) as a template. The distribution of the mRNAs was analyzed by Northern blotting. Top panels: ribosome profile with continuous monitoring of absorbance at 254 nm. Bottom panels: mRNA samples were isolated from each fraction, and crp was detected by Northern blot analysis. (c,d) The distribution of protein products in the sucrose gradients. Samples of the in vitro translation reactions with the normal mRNA (c) and nonstop mRNA (d) were analyzed by NuPAGE. Protein products were analyzed by Western blotting as described in Fig. 1B.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

It has been proposed that tmRNA is recruited to the empty A-site of the stalled ribosome that is irretrievably stuck on the 3′ end of nonstop mRNA, and acts as an aminoacyl-tRNA and then as an mRNA to rescue the stalled ribosome in eubacteria. However, we found that the level of proteins produced from nonstop mRNA and normal mRNA was similar (Sunohara et al. 2004a). Because this strongly suggests that nonstop mRNAs are translated as efficiently as normal mRNAs in the absence of tmRNA, we clarified the fate of ribosomes that are translating nonstop mRNAs in the absence of tmRNA in E. coli in this study. To analyze the translation status of nonstop mRNAs directly, we performed polysome analysis in vivo. As shown in Fig. 1, nonstop mRNAs as well as normal mRNAs were distributed in the monosome and polysome fractions, and the proteins produced from the nonstop mRNA were distributed only in the ribosome-free fractions. Therefore, the stalled ribosome complex containing both the nonstop mRNA and peptidyl-tRNA was not detected in our polysome analysis. There are two possibilities to explain this result. One possibility is that the peptidyl-tRNA in the stalled ribosome complex may be dissociated during sample preparation, centrifugation over sucrose gradients or electrophoresis. However, the stalled complex could be stably isolated when we used the in vitro translation products of the nonstop mRNA (Fig. 3). This indicates that the stalled ribosome itself is stable at least during electrophoresis and that the method used to prepare E. coli cell extracts must be improved. The other possibility is that the nascent peptide is released from the tRNA and the ribosome is dissociated from the nonstop mRNA in vivo, allowing the nonstop mRNA to be translated multiple times as normal mRNA.

Translation is normally terminated at a termination codon upon the release of the nascent peptide from the tRNA, and hydrolysis follows from the interaction of release factor with the ribosome. Our results indicate that nonstop mRNA is translated as normal mRNA, however, the mechanism of translation termination without a termination codon is still largely unknown. An important and unsolved issue is how nascent peptides are released from the tRNA and the ribosomes are dissociated from the nonstop mRNA. A weak but significant peptide release activity in in vitro translation experiments was detected when a nonstop mRNA was used as a template (Fig. 3a), and neither RF1 nor RF2 was responsible for this activity (Fig. 3c). Polysome analysis during in vitro translation revealed that a ribosome translating the nonstop mRNA dissociates from this mRNA (Fig. 3d). We also found that nonstop mRNAs were distributed in the polysome fractions, indicating that these mRNAs are actively translated in vivo (Fig. 1). Based on these results, we propose that ribosomes can be released from the nonstop mRNA and that the peptidyl-tRNA is hydrolyzed on ribosomes in vivo. When the ribosome stalls at the end of a truncated mRNA, peptidyl-tRNA may not correctly base-pair with the codon on the mRNA and may not locate at a position suitable for the peptide transfer reaction. Translational by-passing or frame-shifting may involve a similar process but without hydrolysis of the peptidyl-tRNA (Herr et al. 2000). The peptidyl-tRNA exits the P-site codon and moves to another cognate codon on the same mRNA, resulting in translational by-passing; similarly the peptidyl-tRNA takes off from the nonstop mRNA but the nonstop mRNA dissociates because the peptidyl-tRNA can not find its proper site. Slippage of the peptidyl-tRNA in the P-site usually results in frame-shifting, while stalling of the peptidyl-tRNA at the end of mRNA results in the dissociation of the mRNA. It has been suggested that an incorrect peptidyl-tRNA in a ribosome preferentially disassociates (Caplan & Menninger 1979), probably because of weak codon–anticodon pairing. This may be similar to the peptidyl-tRNA that is in a ribosome stalled at the end of a nonstop mRNA, and the thermodynamic stability of the codon–anticodon pairing should depend on the length of the 3′-extention of the nonstop mRNA from the P-site.

It has been proposed that peptidyl-tRNAs derived from nonstop mRNAs can dissociate from the ribosome in the absence of tmRNA (Singh & Varshney 2004), suggesting a connection between peptidyl-tRNA drop-off and trans-translation. However, there is no clear evidence to show that a peptidyl-tRNA derived from a nonstop mRNA, a substrate for trans-translation, indeed drops off the ribosome in the absence of tmRNA. It is important to emphasize that the drop-off of peptidyl-tRNAs is only evident with short peptides and the drop-off of long peptidyl-tRNAs has never been shown experimentally (Menez et al. 2000). When the length of a nascent peptide is longer than 40 amino acid residues, the nascent peptide may pass through the ribosome tunnel backward in order to disassociate from the ribosome. However, this process might be inefficient because the peptide has been partially folded by chaperones outside the peptide tunnel. Therefore, we propose that it would be difficult for the P-site tRNA with a long nascent peptide to easily detach from the ribosome by passing through the peptide tunnel. Thus, it is logical to assume that the peptide may be released from the ribosome after hydrolysis of the peptidyl-tRNA. Taken together, we suspect that Pth may act on the ribosome after the release of the nonstop mRNA without the drop-off of the peptidyl-tRNA from the ribosome. Pth could attack the peptidyl-tRNA on the ribosome, assuming that the tRNA moiety temporarily exits the P-site to allow a few residues of the nascent peptide to be removed from the tunnel. Singh and Varshney (2004) have shown that overexpression of tmRNA suppresses the temperature-sensitive phenotype of the Pth mutant. This strongly suggests that Pth is involved in rescuing ribosomes that are stalled at the end of nonstop mRNAs. We could not exclude the possibility that the temperature-sensitive (ts) Pth mutant is partially damaged but still active in the cell at the semi-restrictive temperature, and the activity of purified Pth in the PURESYSTEM in vitro translation reactions should be addressed.

We showed that the level of nonstop mRNAs increases in ΔssrA, indicating that tmRNA dissociates the ribosome from the nonstop mRNA and facilitates the degradation of truncated mRNAs. We have proposed that the tmRNA quality control system not only degrades aberrant polypeptides once they are produced but also prevents production of aberrant polypeptides by rapidly eliminating damaged mRNAs (Yamamoto et al. 2003). In the wild-type strain, tmRNA stimulates the degradation of nonstop mRNAs, which otherwise are stable and actively translated. We propose that tmRNA plays crucial roles in the mRNA surveillance system that identifies nonstop mRNAs by rapidly decaying abnormal mRNAs, in addition to protein degradation by trans-translation. It has been proposed that RNase R is required for the tmRNA-dependent degradation of nonstop mRNAs (Richards et al. 2006), and these findings are consistent with our results that nonstop mRNAs are as stable and translatable as normal mRNAs in the absence of tmRNA.

It has been shown that a nonstop mRNA is rapidly degraded by NSD (nonstop mRNA decay pathway in eukaryotes (Frischmeyer et al. 2002; van Hoof et al. 2002). In yeast, Ski7p is proposed to recognize a stalled 80S ribosome at the 3′ end of a nonstop mRNA and recruit the exosome and the Ski complex, resulting in rapid degradation of nonstop mRNA. However, there is no experimental result to show that Ski7p has an activity to rescue ribosome stalled at the end of nonstop mRNAs. We previously proposed a model for the translation status and decay pathways of nonstop mRNA in yeast (Inada & Aiba 2005). A stalled ribosome at the 3′ end of an mRNA inhibits multi-round translation of a nonstop mRNA by blocking the movement of subsequent ribosomes. Recently, we found that translation of the poly(A) tail plays crucial roles in nonstop mRNA surveillance via translation repression and protein destabilization (Ito-Harashima et al. 2007). The protein product of a nonstop mRNA was reduced 100-fold and the aberrant products were destabilized by the proteasome. Furthermore, insertion of poly(A) sequences upstream of a termination codon down-regulated the expression via translation repression and proteasome-mediated protein destabilization without affecting the mRNA levels (Ito-Harashima et al. 2007). This is consistent with the fact that a tRNA with a function equivalent to tmRNA has not been identified in eukaryotes because translation of the poly(A) tail itself results in protein destabilization and translation repression. The poly(A) tail is a common modification for eukaryotic mRNA and is a critical difference between eukaryotic and prokaryotic mRNAs. In prokaryotes, translation of the nonstop mRNA is not repressed without tmRNA probably due to a lack of poly(A) tail and this is why tmRNA is necessary to reduce the abnormal proteins of nonstop mRNA in prokaryotes. This analysis and further characterizations of prokaryotic and eukaryotic ribosomes during translation termination are required to understand not only translation control but also the mRNA surveillance systems in both kingdoms.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Plasmids and E. coli strains

The Escherichia coil strains and plasmids used in this study are described in Table 1. Strains were constructed by P1 transduction using W3110 as the parental strain. Recombinant DNA procedures were performed as described previously (Sambrook & Russell 2001). The pIT815 plasmid was constructed by inserting two annealed oligonucleotides (5′-CGCGTGGCATCCGTGCTGGCCCTA AAGCAAACGACGAAAACTACGCTTTAGCAGCTTAAT-3′ and 5′-CGCGATTAAGCTGCTAAAGCGTAGTTTTCGTCG TTTGCTTTAGGGCCAGCACGGATGCCA-3′) into the MluI site of pHA7MK. RNA preparation and Western blot analysis were performed as described.

Table 1.  Strains and plasmids used in this study
StrainDescriptionReference
TA341W3110ΔcrpSunohara et al. 2004a
TA481W3110Δcrp ssrAΔΔSunohara et al. 2004a
TA501W3110ΔcrpΔssrASunohara et al. 2004a
IT1621W3110ΔssrAThis study
IT1520W3110 pthtsThis study
IT1534W3110ΔssrA pthtsThis study
pHA7AmpRpBR322crpAiba et al. 1982
pJK208AmpRcrp-crr fusion gene encoding CRP-GIRAGP-IIAGlcSunohara et al. 2004a
pIT815AmpRcrp-ssrA fusion gene encoding CRP-SsrAtagThis study
pIT732CmRpSTV28 pthThis study
pIT733CmRpSTV28 pth(ts)This study
pIT736CmRpSTV28 pth(ts)3′-UTR::kanThis study

pth(ts) strain construction

The pth(ts) strain was constructed as follows. A pth EcoRI-BamHI fragment was PCR amplified using primers (5′-GGAATTCACC AGTAACGGTCGCACACCTGTTC-3′ and 5′-CCGGATCC AATACCGGCTTTGGTCAGCGCGTTG-3′) and inserted into pSTV28 (Takara, Japan) to generate pIT732. Then, the pth(ts) mutation, which confers a temperature sensitive phenotype, was constructed by PCR-based site-directed mutagenesis using two oligonucleotides (5′-CTGCCTCCTGACGTCGCCAAA-3′ and 5′-TTTGGCGACGTCAGGAGGCAG-3′) to obtain pIT733. The 3′-UTR fragment of pth was amplified with two oligonucleotides (5′-CCGGATCCTGCACCATTGAGCCGAACACAGG-3′ and 5′-AACTGCAGCAGGAAGCTCAGGTAACGAATAGCCG-3′), inserted into the BamHI-PstI sites and then a BamHI-BamHI kanamycin resistant gene cassette fragment was inserted into the BamHI site to generate pIT736. To replace the wild-type allele with the ts allele, an EcoRI-PstI fragment was purified and transformed into the BW25113 strain, and kanr strains were selected. The ts strains were screened using the replica plating method, and one ts strain was confirmed to be pth(ts) by complementation with the pIT732 plasmid that contains the pth gene.

Preparation of E. coli extracts and sucrose gradient separation

Escherichia coli cells were grown exponentially at 37 °C and harvested by centrifugation after the addition of chloramphenicol. Cells were washed once with buffer A (10 mm Tris-HCl (pH7.5), 60 mm KCl, 10 mm MgCl2, 1 mm DTT, 1 mm PMSF) and extracts were prepared as described previously(Inada et al. 2002). The equivalent of 50 A260 units was then layered onto linear 10%–40% sucrose density gradients. Sucrose gradients (10%–40% sucrose in 10 mm Tris-HCl (pH7.5), 50 mm NH4Cl, 10 mm MgCl2) were prepared in 25 × 89 mm polyallomer tubes (Beckman Coulter) using a gradient master. Crude extracts were layered on top of the sucrose gradients and centrifuged at 27 000 rpm in a P28S rotor (Hitachi Koki) for 3 h at 4 °C. The gradients were then fractionated (Biocomp, New Brunswick). Polysome profiles were generated by continuous absorbance measurement at 254 nm using a single path UV-1 optical unit (ATTO Biomini UV-monitor) connected to a chart recorder (ATTO, digital mini-recorder). Where indicated, equal volume fractions were collected and processed for total RNA purification as described previously (Inada & Aiba 2005).

In vitro translation with PURESYSTEM

PURESYSTEM (PostGenome Lab, Tokyo) was used for in vitro translation experiments. Template DNA fragments were prepared using a two-step PCR reaction. To prepare wild type mRNA, two oligonucleotides (5′-AAGGAGATATACCAATGGTGCTTGGC AAACCGC-3′ and 5′-AAGCATAAAAAAATGGCGCCGATG GGCGCC-3′) were used for the first PCR, followed by two oligonucleotides (5′-GAAATTAATACGACTCACTATAGGG AGACCACAACGGTTTCCCTCTAGAAATAATTTTGTTTAA CTTTAAGAGGAGATATACCA-3′ and 5′-AAGCATAAAAAAAT GGCGCCGATGGGCGCC-3′) for the second PCR. To prepare nonstop mRNA, two oligonucleotides (5′-AAGGAGATATAC CAATGGTGCTTGGCAAACCGC-3′ and 5′-CGCGTGCC GTAAACGACGATGG-3′) were used for the first PCR, followed by two oligonucleotides (5′-GAAATTAATACGACTCACTATA GGGAGACCACAACGGTTTCCCTCTAGAAATAATTTTGT TTAACTTTAAGAGGAGATATACCA-3′ and 5′-CGCGTGC CGTAAACGACGATGG-3′) for the second PCR. The DNA fragments contained T7 RNA polymerase promoter sequences, and the transcription and translation reactions were coupled in this system. The DNA fragments were incubated with 500 mL of Classic II PURESYSTEM, at 37 °C, and some portion of the samples was mixed with 2× sample buffer [62.5 mm Tris-HCl (pH6.8), 2% SDS, 5% Glycerol, 5%β-mercaptoethanol, 0.1% Bromophenol Blue] and analyzed by Western blot after NuPAGE (Invitrogen). The band intensities were quantified by LAS3000-mini (Fuji Film, Japan).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The authors thank Dr Yoshikazu Nakamura and Dr Koichi Ito for helpful discussions and for critically reading this manuscript. We also thank all the members of the lab. This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and Naito foundation (to T.I.).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References