Aberrant messenger RNAs containing a premature termination codon (PTC) are eliminated by the nonsense‐mediated mRNA decay (NMD) pathway. Here, we show that a crucial NMD factor, up frameshift 1 protein (Upf1), is required for rapid proteasome‐mediated degradation of an aberrant protein (PTC product) derived from a PTC‐containing mRNA. Western blot and pulse–chase analyses revealed that Upf1 stimulates the degradation of specific PTC products by the proteasome. Moreover, the Upf1‐dependent, proteasome‐mediated degradation of the PTC product was also stimulated by mRNAs harbouring a faux 3′ untranslated region (3′‐UTR). These results indicate that protein stability might be regulated by an aberrant mRNA 3′‐UTR.
Gene expression is highly accurate because of quality control systems that prevent the expression of potentially harmful protein products. One such system, the nonsense‐mediated messenger RNA decay pathway (NMD), recognizes and eliminates aberrant mRNAs containing a premature termination codon (PTC). Three NMD factors, up frameshift 1 protein (Upf1), Upf2 and Upf3, are highly conserved, and aberrant termination at a premature nonsense codon is thought to allow the binding of NMD factors to the ribosome, triggering mRNA degradation. In mammalian cells, translation termination codons and exon–exon junctions are cis‐acting elements that allow the recognition of PTCs (Holbrook et al, 2004; Maquat, 2004). An abnormally long 3′ untranslated region (3′‐UTR) can also trigger NMD in yeast, mammalian, Drosophila melanogaster, Caenorhabditis elegans and plant cells (Kertesz et al, 2006; Schwartz et al, 2006; Behm‐Ansmant et al, 2007; Longman et al, 2007), although not all mRNAs with an extended 3′‐UTR are subjected to NMD. The poly(A)‐binding protein 1 (Pab1) interacts with the translation termination factor, eukaryotic release factor 3, to enhance normal termination efficiency (Cosson et al, 2002; Uchida et al, 2002). Premature termination, however, is thought to be characterized by the absence of proximal Pab1, leading to a reduction in translation termination efficiency and in the association of NMD factors with the termination complex (Amrani et al, 2004). This faux 3′‐UTR model is supported by the observation that NMD is suppressed by the tethering of Pab1 to the proximal position of a PTC in yeast (Amrani et al, 2004), Drosophila (Behm‐Ansmant et al, 2007) and mammalian cells (Ivanov et al, 2008), indicating that the role of Pabs in the identification of PTCs is conserved across eukaryotes. By contrast, it has been reported that PTC is recognized even in a pab1 deletion mutant (Meaux et al, 2008), suggesting that PTC could be recognized independently of Pab1 in yeast.
Furthermore, it has been shown that post‐translational regulation could have a crucial function in quality control (Inada & Aiba, 2005; Ito‐Harashima et al, 2007). These findings indicate that translation arrest and protein degradation are crucial for preventing potentially harmful products of aberrant mRNAs working together with mRNA degradation. Translation is essential for discriminating PTC‐containing mRNAs, and therefore the truncated products might be synthesized from PTC‐containing mRNAs until NMD factors are able to recognize and identify these aberrant mRNAs. It should be emphasized that NMD generally reduces the abundance of nonsense‐containing mRNAs to approximately 5–25% of the nonsense‐free level, indicating that aberrant products might be produced to some extent. Aberrant proteins can acquire dominant‐negative activities; therefore it is crucial that cells repress the production of aberrant proteins derived from PTC‐containing mRNAs. It has been shown that translation of PTC‐containing aberrant mRNAs is repressed at the translation initiation step (Isken et al, 2008). However, it is unknown whether other steps in gene expression have crucial functions in repressing the aberrant proteins of PTC‐containing mRNAs.
Here, we examined the levels of protein produced by aberrant PTC‐containing mRNAs to investigate the potential role of other regulatory processes in the quality control system that represses PTC products. By stimulating proteasome‐dependent protein degradation and mRNA degradation, we found that Upf1 downregulates the level of aberrant imidazoleglycerol‐phosphate dehydratase (His3) or 3‐phosphoglycerate kinase 1 (Pgk1) proteins derived from mRNAs containing a PTC at specific positions. The downregulation of specific PTC products by Upf1p requires aberrant 3′‐UTRs and proteasome activity, indicating that Upf1 stimulates the proteasome‐mediated degradation of specific PTC products, wherein the stability of these proteins is regulated by the length of the 3′‐UTR. These results lead us to propose that, working together with rapid mRNA degradation, protein degradation by the proteasome has a function in preventing the expression of abnormal proteins derived from aberrant mRNAs.
Results And Discussion
Downregulation of aberrant proteins by Upf1
It has been shown that PTC‐containing aberrant mRNAs are recognized by NMD factors and are rapidly degraded by general mRNA decay pathways. To examine whether other steps in gene expression have crucial functions in repressing proteins derived from aberrant mRNAs, we constructed FLAG–HIS3 reporter genes containing a PTC at different positions (Fig 1A), and determined the levels of aberrant mRNAs and proteins. As shown in Fig 1B, the decrease in the level of PTC‐containing mRNAs was successively less marked as the PTC was positioned further towards the 3′ end within the HIS3 open reading frame (ORF), as previously reported (Peltz et al, 1993; Cao & Parker, 2003; Meaux et al, 2008). We also determined the levels of protein derived from FLAG–HIS3 reporter genes containing a PTC at various codon positions. Western blot analysis using secondary antibodies conjugated to quantum dots (Bakalova et al, 2005), which allows the detection of proteins at low concentrations, revealed that Upf1 reduced protein levels more efficiently than it did mRNA levels, especially for constructs with PTCs at positions 100 and 157 (Fig 1B, lanes 7–8 and 15–16; Fig 1C, lanes 9–12). For example, the level of FLAG–his3‐100 mRNA relative to FLAG–HIS3 mRNA was 15% in wild‐type cells, whereas the protein level of FLAG–his3‐100 relative to FLAG–HIS3 was less than 1% (Fig 1B, lane 7; Fig 1C, lane 9). However, in the absence of Upf1, the relative FLAG–his3‐100 mRNA level was 122%, whereas the relative protein level was 40% (Fig 1B, lane 15; Fig 1C, lane 10). The decrease to 40% protein in comparison with 122% mRNA indicates that protein degradation or reduced translation also occurs in the absence of Upf1. These results show that truncated protein levels are significantly lower than mRNA levels, and indicate that Upf1 contributes to the downregulation of severely truncated PTC products at the translational or post‐translational level, as well as at the mRNA level.
To determine whether the proteasome is involved in this process, we measured the levels of protein derived from various reporter genes in UPF1 and upf1Δ cells after the addition of MG132. The FLAG‐His3‐100 and FLAG‐His3‐157 protein levels in wild‐type cells were low (0.6%) in comparison with the levels of FLAG‐His3 proteins in wild type (100%), but were increased significantly in the presence of 0.2 mM MG132, up to 4% of His3 (Fig 1D, lanes 9–12). By contrast, the levels of longer PTC proteins were minimally affected by the addition of MG132, whereas the levels of FLAG‐His3‐170 and FLAG‐His3‐183 protein slightly increased in the presence of MG132 (Fig 1D, lanes 13–20). Unexpectedly, the levels of FLAG‐His3‐50 protein in both UPF1 and upf1Δ cells were decreased in the presence of MG132 by an unknown mechanism (Fig 1D, lanes 7–8). Upf1 downregulated the FLAG–His3‐100 protein sixfold in the presence of MG132 (Fig 1D, lanes 9–10) to nearly the same degree as the downregulation of FLAG–his3‐100 mRNA (Fig 1B, lanes 7 and 15). It has been reported that proteasome inhibitors also inhibit translation through phosphorylation of eukaryotic initiation factor 2α, which interferes with NMD (Mazroui et al, 2007). Pulse‐label experiments indicate that the rate of protein synthesis was only slightly affects by MG132 treatment (supplementary Fig S1 online). Consistently, MG132 treatment did not affect either the steady‐state levels (supplementary Fig S2 online) or the stability of FLAG–his3‐100 mRNA in wild‐type and upf mutant yeast cells (Fig 2A). Therefore, we concluded that MG132 treatment minimally affects NMD under the conditions used in our study. In wild‐type cells, the level of FLAG–his3‐100 mRNA relative to FLAG–HIS3 mRNA was 15%, and that of the protein was only 4%, even though the proteasome was inhibited by the addition of MG132. This is consistent with previous observations showing that the translation of PTC‐containing mRNAs is repressed (Amrani et al, 2004; Sheth & Parker, 2006; Isken et al, 2008). However, we could not exclude the possibility that another type of proteolysis, in addition to proteasomal degradation, could be involved in the downregulation of the FLAG–His3‐100 protein.
FLAG–His3‐100 is more stable in upf mutants
To address the possibility that Upf1 might stimulate proteasome‐dependent protein degradation of truncated PTC products, the stability of these proteins was examined directly by pulse–chase analysis. We found that the FLAG–His3‐100 protein was more labile in W303 cells (t1/2=7 min) than in upf1Δ (t1/2=30 min), upf2Δ or upf3Δ cells (Fig 2B,C). In addition, FLAG–His3‐100 proteins were significantly stabilized in wild‐type (t1/2 ≈90 min) and upf1Δ cells (t1/2>120 min) in the presence of MG132 (Fig 2B,C). These results indicate that Upf1 destabilizes FLAG–His3‐100 proteins by stimulating their degradation, which is at least partly proteasome dependent. There is a possibility that protease(s) other than the proteasome might be involved in destabilizing truncated proteins, because MG132 does not completely inhibit proteasome function. We propose that Upf1 stimulates proteasome‐dependent degradation of the FLAG–His3‐100 protein, at least partly, to reduce PTC product levels together with the degradation of mRNA.
One possible explanation for the Upf1‐mediated destabilization of FLAG–His3‐100 protein (partly through a proteasome‐dependent pathway) is that proteasome activity is generally impaired in the upf1Δ mutant. To investigate this possibility, we first measured the levels of ubiquitinated proteins by western blot analysis using an FK2 antibody (anti‐ubiquitinylated proteins, clone FK2) that recognizes both mono‐ubiquitin and poly‐ubiquitin forms, but not free ubiquitin (Fujimuro et al, 1994). Rpn6 is a component of the proteasome lid, and the activity of the proteasome is impaired in an rpn6‐2 temperature‐sensitive mutant (Saeki et al, 2005). We found that levels of ubiquitinated proteins increased markedly in the rpn6‐2 mutant but not in the upf1Δ mutant (supplementary Fig S3 online). Furthermore, the half‐life of the well‐known substrate of the proteasome, Sic1 (Petroski & Deshaies, 2005), was almost the same in wild type and upf1Δ mutants (supplementary Fig S4 online). These results also indicate that proteasome activity is not impaired in upf1Δ mutant strains.
Long 3′‐UTR is required for downregulation of His3‐100
To investigate the mechanism by which Upf1 mediates the efficient downregulation of the FLAG–His3‐100 protein, we constructed reporter genes as shown in Fig 3A. Upf1 downregulated FLAG–his3‐100 mRNA levels by 7.5‐fold (Fig 3B, lanes 7–8), and this reduction was significantly suppressed when the long 3′‐UTR was replaced by wild‐type 3′‐UTR (FLAG–his3‐100‐WUTR; Fig 3B, lanes 11–12). Upf1 also reduced the levels of a FLAG–HIS3‐LUTR mRNA that contained an intact ORF flanked by a long 3′‐UTR (Fig 3B, lanes 9–10). These results are consistent with the faux 3′‐UTR model that states that NMD requires improperly long 3′‐UTRs (Amrani et al, 2004). We found that Upf1 reduced both FLAG–HIS3‐LUTR mRNA and protein levels by almost the same extent (Fig 3B,C, lanes 9–10). These results indicate that Upf1 mainly downregulates the levels of the protein derived from FLAG–HIS3‐LUTR by reducing mRNA levels, rather than by proteolysis. On the basis of these results, we propose that post‐translational downregulation by proteolysis occurs only when the protein is truncated. In the case of FLAG–HIS3‐LUTR, Upf1 regulated the levels of mRNA but not of the properly folded wild‐type protein. We also found that Upf1 moderately downregulated the mRNA levels of a short ORF flanked by wild‐type 3′‐UTR (Fig 3B, lanes 11–12), but minimally affected the mRNA and protein levels derived from FLAG–HIS3 (Fig 3B,C, lanes 5–6). Therefore, we suspect that the shorter ORF might stimulate the effects of Upf1 on mRNA levels, and the wild‐type HIS3 3′‐UTR may not be long enough to repress NMD when the ORF is shorter than normal. To our surprise, the downregulation of the FLAG–His3‐100 protein was significantly suppressed by replacing the long 3′‐UTR with a wild‐type 3′‐UTR (FLAG–his3‐100‐WUTR; Fig 3C, lanes 11–12). Pulse–chase experiments also revealed that Upf1 minimally destabilized the FLAG–His3‐100 protein expressed from FLAG–his3‐100‐WUTR (Fig 3D). These results indicate that the Upf1 downregulation of the levels of protein derived from FLAG–his3‐100 depends more strongly on a faux 3′‐UTR than does the Upf1 downregulation of mRNA levels. We also found that Upf1 downregulates both protein and mRNA levels derived from FLAG–pgk1‐300, a construct with a PTC at codon 300 of PGK1, depending on the length of the 3′‐UTR (Fig 3E–G). In addition, Upf1 stimulates proteasome‐dependent degradation of the FLAG–Pgk1‐300 protein, depending on the faux 3′‐UTR (Fig 4). Considering these results, we propose that protein stability could be regulated by the length of the 3′‐UTR of an mRNA.
Recent results, combined with the results reported here, indicate a possible mechanism for the stimulation of protein degradation by Upf1. Recently, it was reported that Upf1 potentially acts as an E3 ubiquitin ligase by its association with Upf3 in yeast (Takahashi et al, 2008). Therefore, Upf1 might act as an E3 ligase of the truncated protein, leading to the degradation of ubiquitinated protein by the proteasome. As Upf1 is recruited to aberrant mRNAs by an interaction with eukaryotic release factor 3 (Amrani et al, 2004; Kashima et al, 2006; Behm‐Ansmant et al, 2007; Ivanov et al, 2008; Singh et al, 2008), the ubiquitination of the aberrant protein might occur after the binding of eukaryotic release factors to the ribosome. It is unknown how quickly the peptide is released from the ribosome after hydrolysis of the peptidyl‐tRNA bond. To understand the mechanism by which Upf1 stimulates protein degradation, the kinetics of translation termination must be analysed more precisely in vivo. Mutational analysis of Upf1 is required for clarifying the mechanism of stimulation of protein degradation by Upf1.
We previously reported that translation of the poly(A) tail has a crucial function in repressing the production of aberrant proteins from nonstop mRNAs by both translation repression and proteasome‐dependent nascent protein destabilization in yeast (Inada & Aiba, 2005; Ito‐Harashima et al, 2007). This clearly indicates that translation arrest and protein degradation, in addition to mRNA degradation, are crucial for repressing the expression of nonstop mRNAs. We also recently reported that translation arrest caused by the presence of consecutive basic amino acids in a nascent protein induces Not4‐dependent co‐translational protein degradation by the proteasome (Dimitrova et al, 2009). In this study, we found that Upf1 prevents the accumulation of aberrant proteins by stimulating the degradation of specific PTC products. Taking these results together, we propose that protein degradation by the proteasome, working together with rapid mRNA degradation, has an important role in preventing the expression of abnormal proteins derived from aberrant mRNAs.
Yeast strains and plasmids. Details of the yeast strains and plasmids used in this study are described in supplementary Table S1 online, and information about the oligonucleotides is in supplementary Table S2 online. The construction of plasmids and other methods are described in the supplementary information online.
Determining relative protein levels by western blot with quantum dots. To quantify the protein levels, protein samples equivalent to 0.1 OD600 (optical density at 600 nm) were analysed by western blotting using secondary antibodies conjugated to quantum dots (Invitrogen, Carlsbad, CA, USA). The band intensities for diluted samples (quantum dot fluorescence was detected by FLA‐9000; FujiFilm, Tokyo, Japan) were compared with a standard curve, and the level of proteins relative to the control was determined.
Pulse–chase experiments. Yeast cells were grown exponentially at 30°C in a minimal medium lacking methionine and cysteine uracil. Yeast cells (10 ml) were labelled with 100 μCi [35S] methionine and cysteine (PerkinElmer NEG072, Waltham, MA, USA) for 1 min, followed by the addition of cold amino acids (final density of 40 μg/ml). At indicated times, cells were collected and cell extracts were prepared by Complete Lysis‐Y (Roche, NJ, USA). Cell extracts were incubated with an anti‐FLAG M2 resin in IXA‐100 buffer (Inada & Aiba, 2005), and then washed three times and eluted with 0.4 mg/ml FLAG peptide. Samples to undergo immunoprecipitation were separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS–PAGE). The radioactivity of precipitated proteins was measured using Typhoon9400 (GE, Healthcare, NJ, USA).
Supplementary information is available at EMBO reports online (http://www.emboreports.org).
Conflict of Interest
The authors declare that they have no conflict of interest.
We thank Dr Allan Jacobson and Dr John Hershey for critically reading this paper and for their valuable comments and suggestions. We also thank Dr Toshiya Endo, Dr Tohru Yoshihisa, Dr Takumi Kamura and Dr Satoru Mimura for yeast strain plasmids and helpful discussions. This study was supported by Grants‐in‐Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to T.I.).
- Copyright © 2009 European Molecular Biology Organization