Multiple elements required for translation of plastid atpB mRNA lackin
http://www.100md.com
《核酸研究医学期刊》
1 Center for Gene Research, Nagoya University, Nagoya 464-8602, Japan and 2 Graduate School of Natural Sciences, Nagoya City University, Yamanohata, Mizuho, Nagoya 467-8501, Japan
* To whom correspondence shoud be addressed at Graduate School of Natural Sciences, Nagoya City University, Yamanohata, Mizuho, Nagoya 467-8501, Japan. Tel/Fax: +81 52 872 6021; Email: sugiura@nsc.nagoya-cu.ac.jp
ABSTRACT
The mechanism of translational initiation differs between prokaryotes and eukaryotes. Prokaryotic mRNAs generally contain within their 5'-untranslated region (5'-UTR) a Shine-Dalgarno (SD) sequence that serves as a ribosome-binding site. Chloroplasts possess prokaryotic-like translation machinery, and many chloroplast mRNAs have an SD-like sequence, but its position is variable. Tobacco chloroplast atpB mRNAs contain no SD-like sequence and are U-rich in the 5'-UTR (–20 to –1 with respect to the start codon). In vitro translation assays with mutated mRNAs revealed that an unstructured sequence encompassing the start codon, the AUG codon and its context are required for translation. UV crosslinking experiments showed that a 50 kDa protein (p50) binds to the 5'-UTR. Insertion of an additional initiation region (SD-sequence and AUG) in the 5'-UTR, but not downstream, arrested translation from the authentic site; however, no inhibition was observed by inserting only an AUG triplet. We hypothesize for translational initiation of the atpB mRNA that the ribosome enters an upstream region, slides to the start codon and forms an initiation complex with p50 and other components.
INTRODUCTION
Translational control is an important means by which cells govern gene expression. The initiation phase of translation usually represents the rate-limiting step, although the mechanism of translational initiation differs between prokaryotes and eukaryotes (1). The 30S ribosomal subunit in prokaryotes binds primarily to mRNA translational initiation regions via start codons combined with Shine-Dalgarno (SD) sequences that are commonly found between 4 and 12 nt upstream from initiation codons (2). On the other hand, the eukaryotic 40S ribosomal subunit enters at the 5' cap of the mRNA, not at the initiation codon. Then the subunit carrying initiator tRNA and initiation factors scans the 5'-untranslated region (5'-UTR) for the first AUG codon. Chloroplasts are plant-specific organelles that possess their own genome (3,4). It has been shown that chloroplast gene expression is primarily regulated at the post-transcriptional level, including translation (5–8). The translation machinery in chloroplasts generally resembles that found in prokaryotes, as chloroplasts contain 70S-type ribosomes and mRNAs lacking 5' caps and eukaryotic 3' poly(A) tails (9,10). Many of the higher plant chloroplast genes encoding polypeptides possess SD-like sequences in 5'-UTRs, although their distance from start codons is highly variable (11,12). In the tobacco chloroplast genome, 30 of the 79 polypeptide-coding genes contain no SD-like sequences (GGA, AGG, GAGG, GGAG or GGAGG) within 20 nt upstream from the initiation codon, and the remaining 49 have SD-like sequences but not at a conserved position (13).
Using chloroplast transformation in Chlamydomonas reinhardtii, mutation analysis of chloroplast SD-like sequences has yielded conflicting results; deletion of the SD-like sequence from its psbA 5'-UTR ceased translation (14), while replacement mutagenesis of the SD-like sequence in the 5'-UTRs of petD (15) and of atpB, atpE, rps4 and rps7 mRNAs (16) had little effect on translation. Further mutation studies suggested that the secondary structure in the 5'-UTR of rps7 mRNA plays an essential role in translational initiation (17) and that the ribosomal protein S7 is also involved in translational initiation (18). In vivo analysis by tobacco chloroplast transformation has shown that the 5'-UTR of psbA mRNAs is important for translation efficiency (19–22) and that sequences downstream of the AUG codon in rbcL and atpB mRNAs are also important determinants of translational efficiency in chloroplasts (23).
In order to dissect biochemical processes of translation unique to higher plant chloroplasts, we have developed an in vitro translation system from tobacco chloroplasts. Our in vitro system supports reproducibly accurate initiation of translation from a variety of chloroplast mRNAs. Hence it provides a powerful tool for the functional analysis of regulatory elements/factors and of the role of low molecular mass compounds (24). Using this in vitro system, we identified three distinct cis-acting elements in the 5'-UTR of tobacco psbA mRNA. Among them, RBS1 (AAG) and RBS2 (UGAUGAU) are complementary to the 3' terminal region of chloroplast 16S rRNA, and the AU-box (UAAAUAAA, –21 to –14) located between RBS1 and RBS2 is likely to be a binding site of putative trans-acting factor(s). An SD-like sequence (GGAG) located at –36 to –33 is not important for translation. We then assessed the role for translation of SD-like sequences in tobacco chloroplast mRNAs (25,26). The atpE, rps14 and rbcL mRNAs possess SD-like sequences (GGAG at –18 to –15, GGA at –14 to –12 and GGAGG at –10 to –6, respectively) at a position similar to the conserved region observed in Escherichia coli mRNAs, and these sequences were found to be essential for translation. On the other hand, SD-like sequences in the rps12 mRNA and in the petB mRNA are located far from (–44 to –42) and too close to (–5 to –2) the initiation codon, respectively, and these sequences are not essential for translation. Unexpectedly, an SD-like sequence (GGAG) at a proper position (–8 to –5) was not required and even inhibitory for translation of tobacco rps2 mRNA (27). In addition, translation of the tobacco ndhD mRNA starts only from the AUG codon created by RNA editing where no SD-like sequence is present in the 5'-UTR, suggesting that translational initiation of ndhD mRNAs depends on a novel sequence element(s) (28,29).
Here, we describe the existence of a novel mode of translational initiation unique to chloroplasts. Translation of the atpB mRNA lacking an SD-like sequence in the 5'-UTR requires an unstructured sequence to which trans-acting factor(s) binds. Based on assays which utilize artificial insertion of an additional translational initiation region and of an AUG triplet into the 5'-UTR of atpB mRNAs, we hypothesize that a chloroplast ribosome enters an upstream region and slides to the initiation codon for translational initiation in chloroplast atpB mRNAs.
MATERIALS AND METHODS
Plasmid construction
Portions of atpB and atpE genes were amplified by PCR from tobacco chloroplast DNA (30,31), cut with appropriate restriction enzymes, and cloned into pBluescript II SK–, in which atpB (–244 to +169) and atpE (–431 to +191) regions were fused in-frame to the 3' lacZ coding region (224 bp). As the atpB overlaps with the downstream atpE, the region encompassing the 3' part of atpB and the entire atpE was replaced by 'lacZ to eliminate possible effects of the atpE mRNA. The calculated size of the hybrid product was 14.3 kDa. The5' deletion constructs were prepared by inverse PCR using appropriate primer pairs and original constructs. Site-directed mutagenesis was performed using a TransformerTM Site Directed Mutagenesis kit Version 2 (Clontech) using primer pairs and original constructs. Constructs were amplified in E.coli XL-1 Blue, isolated by the alkaline lysis method and purified further by Qiagen midi tip (Qiagen). Plasmid DNAs were linearized with BglII, and treated with phenol/chloroform and precipitated with ethanol. mRNAs were synthesized with T3 RNA polymerase using T3 MEGASCRIPT (Ambion) and further purified by Sephadex G50 spin columns (Amersham Pharmacia).
Chloroplast S30 fraction and in vitro translation
Tobacco plants (Nicotiana tabacum var. Blight Yellow 4) were grown in a growth chamber (28°C, 16 h light and 8 h dark) for 4 weeks. Preparation of S30 fractions from intact tobacco chloroplasts and in vitro translation reactions were carried out as previously described (24). The dialyzed S30 fraction (0.5 ml, 15 mg/ml protein) could be stored at –70°C for up to 6 months. In vitro translation reactions were carried out at 30°C for 30 min in 50 μl solution containing 30 mM HEPES–KOH, pH 7.7, 10 mM magnesium acetate, 60 mM potassium acetate, 60 mM NH4Cl, 1% polyethyleneglycol 6000, 1 mM ATP, 0.1 mM GTP, 8 mM creatine phosphate, 0.4 mg/ml creatine phosphokinase (Type I; Sigma), 2 mM dithiothreitol, 0.4 mg/ml E.coli tRNAs (Boehringer), 0.4 mg/ml E.coli tRNAfmet (Sigma), 1 μg/ml leupepcin, 130 U RNase inhibitor (TaKaRa), 0.37 MBq L-methionine (>37 TBq/mmol; Amersham), 20 pmol mRNA and 24 μl (360 μg protein) S30 fraction. We adopted a template-excess condition which yielded high translation activity. The synthesized products were separated by 0.1% SDS-15 or 18% PAGE. Protein size markers were Rainbow markers (Amersham Pharmacia). The separated products were visualized and quantified by a Bio-imaging Analyzer BAS2000 (Fuji Photo Film Co.).
Enzymatic probing of RNA structure and primer extension
Partial digestion of mRNA by RNase T2 and primer extension was performed essentially as previously described (28). mRNAs were synthesized from constructs wt, SC5, SC8, A7 and U7, which were linearized with BssHII followed by purification by denaturing gel electrophoresis. RNA samples (10 pmol each) were preincubated for 20 min at 25°C in 30 mM Hepes–KOH (pH 7.7), 60 mM potassium acetate, 10 mM magnesium acetate and 60 mM NH4Cl, and digested with 0.1 U RNase T2 (Sigma) for 5 min at 25°C. Digested RNA samples were extracted by phenol/chloroform and precipitated with ethanol. The primer (positions +39 to +25) was labeled at its 5' end with T4 polynucleotide kinase and ATP, and annealed to the digested RNAs at 95°C for 1 min followed by cooling to 42°C. Reverse transcription was carried out using a cDNA cycle kit (Invitrogen). Primer extension samples were separated by 10% PAGE containing 7 M urea and visualized by a Bio-imaging Analyzer BAS2000.
UV crosslinking assays
Labeled RNAs were synthesized using an in vitro transcription kit with UTP (110 TBq/mmol) and purified by PAGE. Labeled RNA at a concentration of 10 fmol was incubated in a chloroplast S30 extract under the conditions of an in vitro translation reaction at 30°C for 15 min (24). The S30 fraction (360 μg protein) was excess with respect to the mRNA concentration. UV irradiation followed by RNase A digestion and SDS–PAGE were carried out as previously reported (32).
RESULTS
Sequences required for translation of atpB mRNAs
The tobacco chloroplast atpB gene encoding the ? subunit of chloroplast F0F1 ATP synthase is co-transcribed with its downstream atpE gene for its subunit (33), and transcription of the atpB/E operon starts from multiple sites (34,35). The 20 nt region immediately upstream of the start codon in the atpB 5'-UTR is highly U-rich and possesses no SD-like sequence, suggesting that the chloroplast ribosome recognizes the initiation site through a novel element(s). To identify sequences necessary for the initiation of translation, an atpB–lacZ fusion mRNA was constructed starting from –244 to +169 of the atpB mRNA, which contained an almost complete 5'-UTR of the major atpB mRNA accumulated in tobacco chloroplasts , and a set of 5' deletion mutant mRNAs were also generated (Figure 1A). As shown in Figure 1B, translation occurred at a relatively high level (80% of wt) by deletion to position –53 (5D1), and at a moderate level (40% of wt) by deletion to –25 (5D2). However, translation could scarcely be detected from the mRNA lacking the entire 5'-UTR (5D3). There is an upstream open reading frame (uORF) of 25 codons. The above assays also show that translation of the authentic coding region is hardly affected by the uORF. In vitro translation using two internal deletion mutants showed that the mRNA lacking the –25 to –1 region (ID1) could not be translated, whereas the mutant lacking the –53 to –25 region (ID2) produced translation products, although to a limited extent (20% of wt). Therefore, we concluded that the region from –25 to –1 (or part of it) is essential for translation of atpB mRNAs and the 5'-UTR from –53 to –25 is necessary for efficient translation.
Figure 1. Translation in vitro of tobacco chloroplast atpB mRNAs. The atpB mRNA was fused with lacZ mRNA. (A) Schematic representation of the atpB 5'-UTR (wt), and its 5' deletion (5D) and internal deletion (ID) mutants. The 25 nt sequence upstream of the start codon is shown above. uORF, upstream ORF of 25 codons. (B) 35S-products were separated by PAGE. Translation activity was calculated from band densities (wild type as 100%) and is shown.
In order to define the essential elements more in detail, in vitro translation experiments were carried out using a series of mutant mRNAs produced by scanning mutagenesis in the region (from –25 to +18) spanning the initiation codon (7 nt was sequentially substituted by C-heptamer with 3 nt overlapping, Figure 2A). As shown in Figure 2B, mutations to –11 (SC13) and from +11 (SC9) reduced translation to 20% or less, and from –10 to +10 (SC48) abolished translation almost completely (<5%). These results indicate that both upstream and downstream regions are required for translation of atpB mRNAs. The 20 nt region spanning AUG (–10 to +10) is essential and its flanking sequences are also necessary for efficient translation.
Figure 2. Translation in vitro of mutated atpB mRNAs. (A) Schematic representation of altered regions constructed by replacement mutagenesis using C-heptamers (SC). Hyphens represent identical nucleotides with the wild type (wt). (B) 35S-products were separated by PAGE. Translation activity was calculated from band densities (wt as 100%) and is shown.
Importance of an unstructured sequence for translation of atpB mRNAs
As the essential region (–10 to +10) of atpB mRNA was defined by substitution with C-heptamers, this sequence was then substituted with an A or U stretch (Figure 3A). Translation of A7 and U7 mRNAs was not impaired and that of U17 mRNA was reduced slightly, indicating that no specific sequence in this region is necessary for basic translation and C-stretches abolished translation (Figure 3B).
Figure 3. Translation in vitro of mutated atpB mRNAs and detection of secondary structures in these mRNAs. (A) Schematic representation of mutated regions constructed by replacement mutagenesis. Hyphens represent identical nucleotides with the wild type (wt). (B) 35S-products were separated by PAGE. (C) RNase T2 probing pattern of mRNAs. Vertical arrows on the right of gel patterns show protected sequences; + and –, with and without RNase T2 treatment; +1, the first position of AUG. Ladders appeared in RNase T2 lanes are probably due to slight degradation of mRNA during incubation because no RNase inhibitor was added and also due to cDNA synthesis artefacts.
Secondary structures of the wild-type atpB mRNA and its mutant mRNAs were determined by RNase T2 probing. As shown in Figure 3C, only the C-heptamers in SC5 and SC8 mRNAs were protected from RNase T2 digestion, indicating that these regions form secondary structures. Conversely, the corresponding region in wt, A7 and U7 mRNAs does not form a strong secondary structure, suggesting that an unstructured RNA sequence rather than a specific sequence is important for translation of atpB mRNA.
Importance of the initiation codon and its context for translation of atpB mRNAs
The tobacco atpB gene contains 15 in-frame ATG codons in the coding region and one ATG triplet in the 5'-UTR (up to –244). Our in vitro translation system started translation only from the first AUG codon in the coding region. It is difficult to speculate that an unstructured RNA sequence around the initiation codon is the sole determinant for accurate translation initiation; therefore, an additional determinant(s) is expected to be present within the essential region. We examined first the importance of the context of initiation codon (–3 to +6). Each nucleotide from –3 to +6 (except AUG) was systematically altered by site-directed mutagenesis. As shown in Figure 4A, in vitro translation of these mutant mRNAs revealed that the context is important for translation efficiency. No translation occurred by substitution of +4A with other nucleotides. As substitution of +4A by U creates a termination codon (UGA), both +4A and +6A were substituted by U's. This double mutant almost abolished translation, confirming that U cannot substitute +4A. The –1U can be substituted by A but not by C or G. The +5G can be changed by A or U but not by C. The substitution of +5G with A or U and that of –2U with C increased translation more than 2-fold, suggesting that the wild-type sequence is not always the most efficient substrate, at least in vitro. Taken together, the sequence context from –3 to +6 for the most efficient translation is (A/G)C(A/U)AUGA(A/U)(A/U) (underline indicates a start codon).
Figure 4. Determination of the efficient context and start codon for translation of tobacco chloroplast atpB mRNAs. (A) Translation in vitro of point mutants in the AUG flanking nucleotides. Downward arrows indicate mutated nucleotides. 35S-products were separated by PAGE. Translation activity was calculated from band densities (wt as 100%) and is shown. (B) Effect of replacement from AUG to GUG on translation in vitro. The atpE mRNA with a functional SD sequence (26) was used as a control. 35S-products were separated by PAGE. Translation activity was shown (wt as 100%).
The initiation codon AUG of atpB mRNA was then altered to GUG and, as a control, the AUG initiation codon of atpE mRNA which possesses a functional SD sequence was replaced by GUG. As shown in Figure 4B, the mutant atpE mRNA was translated at 10% of the control efficiency, whereas no translation was observed from the mutant atpB mRNA. This indicates that the AUG codon itself is one of the essential determinants for translation of atpB mRNAs.
Insertion of a translational initiation region into the 5'-UTR arrests authentic translation
Eukaryotic translation is believed to be initiated by the binding of a 40S ribosomal subunit at the 5' end of mRNA mediated by the cap structure, followed by scanning to the first AUG codon (1). To examine the possibility of ribosomal scanning on chloroplast mRNAs, a 13 nt sequence including the SD sequence, spacer and start codon derived from the tobacco chloroplast rbcL mRNA (SD-AUG) was inserted before (–37 to –25) or after (+10 to +22) the authentic initiation codon of atpB mRNA (Figure 5A). The SD sequence (GGAGG) in the rbcL 5'-UTR was found to be functional (26,36). As shown in Figure 5B, insertion of the rbcL initiation region into the 5'-UTR (SD-AUG-atpB) abolished translation from the authentic initiation codon, while translation from the inserted AUG was detected at a low level. Insertion of only an AUG triplet to the same site caused no effect on translation from the authentic site. On the other hand, insertion of the SD-AUG into the downstream region (atpB-SD-AUG) had no effect on translation from the authentic site, and translation from the inserted SD-AUG was also detected. Translation occurred exclusively from the SD-AUG inserted in the downstream region when the authentic AUG was altered to UAC, indicating that translational initiation from the rbcL initiation region is independent of surrounding sequences.
Figure 5. Effect on translation in vitro of an additional initiation region inserted in the upstream or downstream region from the AUG in atpB mRNAs. (A) Schematic representation of mRNA constructs. The inserted sequence from the tobacco chloroplast rbcL mRNA is shown. UAC-SD-AUG mRNA possesses UAC in place of the authentic AUG. (B) 35S-products were separated by PAGE. Asterisks indicate products from the inserted SD-AUG site.
Protein factor(s) specifically binds to the U-rich upstream element
In order to detect proteins that associate with the translational initiation region of atpB mRNA, UV crosslinking experiments were carried out using the in vitro translation extract. As shown in Figure 6, several chloroplast proteins were detected by crosslinking to the 32P-labeled atpB mRNA (lane wt). Experiments with a set of mutant atpB mRNAs used for in vitro translation showed that a 50 kDa protein (p50) could not bind to the mRNAs whose 25 nt upstream region was deleted (ID1) or replaced by a C-heptamer (SC5). On the other hand, p50 could crosslink with the mRNA in which AUG was replaced by UAC (UAC) or the downstream region was replaced by a C-heptamer (SC8). A protein of 40 kDa and several other proteins of 28–33 kDa bind to all mutant mRNAs as well as the wild-type mRNA, suggesting that these proteins are non-specific RNA-binding proteins in chloroplasts (37). This result suggests that p50 preferentially binds to the 25 nt upstream region and promotes translational initiation.
Figure 6. Detection of tobacco chloroplast proteins bound to atpB mRNAs by UV crosslinking. mRNA constructs are depicted above: the detail of wt and ID1 mRNAs is shown in Figure 1A and that of SC5 and SC8 mRNAs is in Figure 2A, and UAC mRNA possesses UAC in place of AUG. C7, C-heptamer. Uniformly labeled atpB mRNAs were incubated with a chloroplast extract and UV irradiated. After RNase A treatment, 32P-proteins were separated by PAGE. Size markers are shown to the right. Translation activity was shown (wt as 100%).
DISCUSSION
Distinct mechanisms of translational initiation in chloroplasts
Previously we examined the effect on translation of SD-like sequences in tobacco chloroplast mRNAs, and found that SD-containing mRNAs can be categorized into two groups (25,26); (i) translation dependent on SD-like sequences and (ii) translation independent of SD-like sequences. Unlike these cases, sequences which lack the SD motif are found in the 5'-UTR of the tobacco atpB mRNA.
Mutation analysis of its 5'-UTR and N-terminal protein-coding region revealed that at least three critical elements affect translation of atpB mRNAs. An unstructured sequence rather than a specific sequence encompassing the initiation codon is important for efficient translation. The tobacco atpB mRNA possesses a characteristic upstream sequence (–1 to –20) consisting of 17 U and 3 A residues (see Figure 1A). This 5'-UTR is relatively similar to that of the rbcL mRNA in Euglena chloroplasts, and in addition its translational initiation was reported to be affected either by its unstructured or weakly structured region rather than by a specific sequence (38). The context of the initiation codon affects translational efficiency of atpB mRNAs. Among the 79 tobacco chloroplast protein-coding genes, only 30 genes have A at +4, and only ndhD perfectly matches the above context (A C U AUG A A U) whose AUG is created from ACG by RNA editing (28). The ndhD mRNA has no SD-like sequence but the U-rich sequence in the 5'-UTR, suggesting that translation initiates in a similar manner to that of atpB mRNA. The other chloroplast mRNAs (e.g. psbA and atpE) which are efficiently translated in vitro have no context similar to that of atpB mRNA, suggesting that the context is specific for a set of mRNAs with U-rich leaders. Recently, in vivo evidence has been reported that the extended codon-anticodon interaction (between the nucleotide 3' to the initiator tRNA anticodon and the base immediately preceding the initiation codon) is important for initiation of Chlamydomonas petA mRNA (39). Only 35 of the 79 mRNAs in tobacco chloroplasts possess U at the –1 position. Therefore, the expanded codon-anticodon model is unlikely for translation of the rest of the tobacco mRNAs. AUG is essential for atpB mRNA translation, which cannot be substituted by GUG, whereas atpE mRNA can be translated from a GUG codon. In tobacco chloroplasts, psbC, rps19 and ycf15 possess GTG as initiation codons (13). All three mRNAs are preceded by typical SD-like elements, suggesting that the chloroplast ribosome can recognize GUG as an initiation codon when a functional SD sequence is present.
Our UV crosslinking assay revealed the interaction of a 50 kDa RNA-binding protein (p50) with the 25 nt upstream region of atpB mRNA. An RNA-binding protein of 47 kDa was reported to bind to the spinach psbA 5' UTR and it was identified as the chloroplast homolog of the E.coli ribosomal protein S1 (40). The tobacco p50 is similar in size to the spinach S1 protein and might be its tobacco homolog. A maize mutant, atp1, was shown to reduce synthesis of the atpB product, and atp1 function was suggested to be required during translational initiation (41). Hence, p50 may be encoded by atp1. Chloroplast RNA-binding proteins 50 kDa in size have been reported as factors possibly involved in endonucleolytic mRNA processing in mustard and in mRNA 3' end formation in spinach . These proteins preferentially bind to U-rich sequences, suggesting that the chloroplast possesses a large family of RNA-binding proteins with high affinity to U-rich sequences.
Ribosomal scanning in chloroplasts?
Elements required for atpB mRNA translation are (i) an AUG start codon, (ii) an unstructured RNA sequence surrounding the AUG and (iii) the context of the initiation codon, which apparently corresponds to the important elements involved in cytoplasmic translational initiation (1,46,47). Translation usually starts at the first AUG in cytoplasmic mRNAs. The 40S ribosomal subunit binds to the 5'-end of mRNA and then scans the 5'-UTR, searching for the initiation codon. An AUG triplet inserted upstream of the authentic initiation codon replaces it as the initiation codon.
Translation from the authentic initiation codon of atpB mRNA was impaired neither by the presence of uORF (see Figure 1) nor by an AUG triplet inserted (–27 to –25) upstream of wt initiation region (see Figure 5), suggesting that the scanning model cannot simply be applied to explain translational initiation in chloroplasts. However, insertion of an SD sequence plus AUG derived from tobacco rbcL mRNA into the same site completely arrested translation from the authentic initiation site; instead, translation initiated from the inserted site though low in efficiency (see Figure 5). On the other hand, the insertion of the same sequence into the downstream region (+10 to +22) did not affect translation from the authentic site. This could exclude the possibility that the inserted SD-AUG and the authentic initiation region compete to capture ribosomes. A plausible explanation is that ribosomal binding to the SD-AUG sequence blocks sliding of the 30S ribosomal subunit from the upper side. As translation activity of tobacco rbcL mRNA is slow in vitro (our unpublished data), the mRNA–ribosome complex via the SD-AUG sequence is likely to be relatively stable, which may stall ribosomal subunits at this point. Taken together, we favor the hypothesis that the 30S subunit enters a 5' distal region, slides to the AUG initiation codon and forms an initiation complex with the aid of p50 and other components. As chloroplast mRNAs have no cap structure, the ribosome entry site, if any, can either be at the 5' end, within the 5'-UTR, or both. One intriguing possibility is that ribosomes primarily bind to a sequence between –53 and –25, because a deletion up to –53 still possesses near full activity but insertion of the SD-AUG at –25 blocks translation. However, other possibilities cannot be ruled out and we need further analyses.
Our in vitro translation experiments revealed the existence of a novel mechanism of translational initiation in chloroplasts. In E.coli, translation of most mRNAs is thought to be initiated via the SD sequence. On the other hand, sequencing of a cyanobacterial ribosomal gene cluster showed that the SD sequence is not always a common sequence element in the prokaryotic genes (13). The mechanism reported here might operate both in chloroplasts and some eubacteria.
ACKNOWLEDGEMENTS
We thank Dr Takahiro Nakamura for his critical reading of this manuscript. This work was supported in part by a Grant-in-Aid for Scientific Research (B) (No. 12440226) and Exploratory Research (No. 13874111) from the Ministry of Education, Science, Sports and Culture.
REFERENCES
Kozak,M. ( (1999) ) Initiation of translation in prokaryotes and eukaryotes. Gene, , 234, , 187–208.
McCarthy,J.E.G. and Gualerzi,C. ( (1990) ) Translational control of prokaryotic gene expression. Trends Genet., , 6, , 78–85.
Sugiura,M. ( (1992) ) The chloroplast genome. Plant Mol. Biol., , 19, , 149–168.
Wakasugi,T., Tsudzuki,T. and Sugiura,M. ( (2001) ) The genomics of land plant chloroplasts: gene content and alternation of genomic information by RNA editing. Photosynthesis Res., , 70, , 107–118.
Stern,D.B., Higgs,D.C. and Yang,J.J. ( (1997) ) Transcription and translation in chloroplasts. Trends Plant Sci., , 2, , 308–315.
Bruick,R.K. and Mayfield,S.P. ( (1999) ) Light-activated translation of chloroplast mRNAs. Trends Plant Sci., , 4, , 190–195.
Zerges,W. ( (2000) ) Translation in chloroplasts. Biochimie, , 82, , 583–601.
Choquet,Y. and Wollmann,F.-A. ( (2002) ) Translational regulations as specific traits of chloroplast gene expression. FEBS Lett., , 529, , 39–42.
Yamaguchi,K., von Knoblauch,K. and Subramanian,A.R. ( (2000) ) The plastid ribosomal proteins: identification of all the proteins in the 30 S subunit of an organelle ribosome (chloroplast). J. Biol. Chem., , 275, , 28455–28465.
Yamaguchi,K., Prieto,S., Beligni,M.V., Haynes,P.A., McDonald,W.H., Yates,J.R. and Mayfield,S.P. ( (2002) ) Proteomic characterization of the small subunit of Chlamydomonas reinhardtii chloroplast ribosome: identification of a novel S1 domain-containing protein and unusually large orthologs of bacterial S2, S3, and S5. Plant Cell, , 14, , 2957–2974.
Ruf,M. and K?ssel,H. ( (1988) ) Occurrence and spacing of ribosome recognition sites in mRNAs of chloroplasts from higher plants. FEBS Lett., , 240, , 41–44.
Bonham-Smith,P.C. and Bourque,D.P. ( (1989) ) Translation of chloroplast-encoded mRNA: potential initiation and termination signals. Nucleic Acids Res., , 17, , 2057–2080.
Sugiura,M., Hirose,T. and Sugita,M. ( (1998) ) Evolution and mechanism of translation in chloroplasts. Annu. Rev. Genet., , 32, , 437–459.
Mayfield,S.P., Cohen,A., Danon,A. and Yohn,C.B. ( (1994) ) Translation of the psbA mRNA of Chlamydomonas reinhardtii requires a structured RNA element contained within the 5' untranslated region. J. Cell Biol., , 127, , 1537–1545.
Sakamoto,W., Chen,X., Kindle,K.L. and Stern,D.B. ( (1994) ) Function of the Chlamydomonas reinhardtii petD 5' untranslated region in regulating the accumulation of subunit IV of the cytochrome b6/f complex. Plant J., , 6, , 503–512.
Fargo,D.C., Zhang,M., Gillham,N.W. and Boynton,J.E. ( (1998) ) Shine-Dalgarno-like sequences are not required for translation of chloroplast mRNAs in Chlamydomonas reinhardtii chloroplasts or in Escherichia coli. Mol. Gen. Genet., , 257, , 271–282.
Fargo,D.C., Boynton,J.E. and Gillham,N.W. ( (1999) ) Mutations altering the predicted secondary structure of a chloroplast 5' untranslated region affect its physical and biochemical properties as well as its ability to promote translation of reporter mRNAs in both the Chlamydomonas reinhardtii chloroplast and in Escherichia coli. Mol. Cell. Biol., , 19, , 6980–6990.
Fargo,D.C., Boynton,J.E. and Gillham,N.W. ( (2001) ) Chloroplast ribosomal protein S7 of Chlamydomonas binds to chloroplast mRNA leader sequences and may be involved in translation initiation. Plant Cell, , 13, , 207–218.
Staub,J.M. and Maliga,P. ( (1993) ) Accumulation of D1 polypeptide in tobacco plastids is regulated via the untranslated region of the psbA mRNA. EMBO J., , 12, , 601–606.
Staub,J.M. and Maliga,P. ( (1994) ) Translation of psbA mRNA is regulated by light via the 5'-untranslated region in tobacco plastids. Plant J., , 6, , 547–553.
Eibl,C., Zou,Z., Beck,A., Kim,M., Mullet,J. and Koop,H.-U. ( (1999) ) In vivo analysis of plastid psbA, rbcL and rpl32 UTR elements by chloroplast transformation: tobacco plastid gene expression is controlled by modulation of transcript levels and translation efficiency. Plant J., , 19, , 333–345.
Zou,Z., Eibl,C. and Koop,H.-U. ( (2003) ) The stem–loop region of the tobacco psbA 5'UTR is an important determinant of mRNA stability and translation efficiency. Mol. Gen. Genomics, , 269, , 340–349.
Kuroda,H. and Maliga,P. ( (2001) ) Sequences downstream of the translation initiation codon are important determinants of translation efficiency in chloroplasts. Plant Physiol., , 125, , 430–436.
Hirose,T. and Sugiura,M. ( (1996) ) Cis-acting elements and trans-acting factors for accurate translation of chloroplast psbA mRNAs: development of an in vitro translation system from tobacco chloroplasts. EMBO J., , 15, , 1687–1695.
Hirose,T., Kusumegi,T. and Sugiura,M. ( (1998) ) Translation of tobacco chloroplast rps14 mRNA depends on a Shine-Dalgarno-like sequence in the 5'-untranslated region but not on internal RNA editing in the coding region. FEBS Lett., , 430, , 257–260.
Hirose,T. and Sugiura,M. ( (2004) ) Functional Shine-Dalgarno-like sequences for translational initiation of chloroplast mRNAs. Plant Cell Physiol., , 45, , 114–117.
Plader,W. and Sugiura,M. ( (2003) ) The Shine-Dalgarno-like sequence is a negative regulatory element for translation of tobacco chloroplast rps2 mRNA: an additional mechanism for translational control in chloroplasts. Plant J., , 34, , 377–382.
Hirose,T. and Sugiura,M. ( (1997) ) Both RNA editing and RNA cleavage are required for translation of tobacco chloroplast ndhD mRNA: a possible regulatory mechanism for the expression of a chloroplast operon consisting of functionally unrelated genes. EMBO J., , 16, , 6804–6811.
Sasaki,T., Yukawa,Y., Miyamoto,T., Obokata,J. and Sugiura,M. ( (2003) ) Identification of RNA editing sites in chloroplast transcripts from the maternal and paternal progenitors of tobacco (Nicotiana tabacum): comparative analysis shows the involvement of distinct trans-factors for ndhB editing. Mol. Biol. Evol., , 20, , 1028–1035.
Shinozaki,K., Ohme,M., Tanaka,M., Wakasugi,T., Hayashida,N., Matsubayashi,T., Zaita,N., Chungwongse,J., Obokata,J., Yamaguchi-Shinozaki,K. et al. ( (1986) ) The complete nucleotide sequence of the tobacco chloroplast genome: its gene organization and expression. EMBO J., , 5, , 2043–2049.
Sugiura,M., Shinozaki,K., Zaita,N., Kusuda,M. and Kumano,M. ( (1986) ) Clone bank of the tobacco (Nicotiana tabacum) chloroplast genome as a set of overlapping restriction endonuclease fragments: mapping of eleven ribosomal protein genes. Plant Sci., , 44, , 211–216.
Hirose,T. and Sugiura,M. ( (2001) ) Involvement of a site-specific trans-acting factor and a common RNA-binding protein in the editing of chloroplast mRNAs: development of a chloroplast in vitro RNA editing system. EMBO J., , 20, , 1144–1152.
Shinozaki,K., Deno,H., Kato,A. and Sugiura,M. ( (1983) ) Overlap and cotranscription of the genes for the beta and epsilon subunits of tobacco chloroplast ATPase. Gene, , 24, , 147–155.
Kapoor,S., Wakasugi,T., Deno,H. and Sugiura,M. ( (1994) ) An atpE-specific promoter within the coding region of the atpB gene in tobacco chloroplast DNA. Curr. Genet., , 26, , 263–268.
Kapoor,S., Suzuki,J.Y. and Sugiura,M. ( (1997) ) Identification and functional significance of a new class of non-consensus-type plastid promoters. Plant J., , 11, , 327–337.
Kim,J. and Mullet,J.E. ( (1994) ) Ribosome-binding sites on chloroplast rbcL and psbA mRNAs and light-induced initiation of D1 translation. Plant Mol. Biol., , 25, , 437–448.
Li,Y. and Sugiura,M. ( (1990) ) Three distinct ribonucleoproteins from tobacco chloroplast: each contains a unique amino terminal acidic domain and two ribonucleoprotein consensus motifs. EMBO J., , 9, , 3059–3066.
Koo,J.S. and Spremulli,L.L. ( (1994) ) Effect of the secondary structure in the Euglena gracilis chloroplast ribulose-bisphosphate carboxylase/oxygenase messenger RNA on translational initiation. J. Biol. Chem., , 269, , 7501–7508.
Esposito,D., Fey,J.P., Eberhard,S., Hicks,A.J. and Stern,D.B. ( (2003) ) In vivo evidence for the prokaryotic model of extended codon-anticodon interaction in translation initiation. EMBO J., , 22, , 651–656.
Alexander,C., Faber,N. and Klaff,P. ( (1998) ) Characterization of protein-binding to the spinach chloroplast psbA mRNA 5' untranslated region. Nucleic Acids Res., , 26, , 2265–2272.
McCormac,D.J. and Barkan,A. ( (1999) ) A nuclear gene in maize required for the translation of the chloroplast atpB/E mRNA. Plant Cell, , 11, , 1709–1716.
Nickelsen,J. and Link,G. ( (1991) ) RNA–protein interactions at transcript 3' ends and evidence for trnK-psbA cotranscription in mustard chloroplasts. Mol. Gen. Genet., , 228, , 89–96.
Chen,H. and Stern,D.B. ( (1991) ) Specific binding of chloroplast proteins in vitro to 3' untranslated region of spinach chloroplast petD mRNA. Mol. Cell. Biol., , 11, , 4380–4388.
Hayes,R., Kudla,J., Schuster,G., Gabay,L., Maliga,P. and Gruissem,W. ( (1996) ) Chloroplast mRNA 3'-end processing by a high molecular weight protein complex is regulated by nuclear encoded RNA binding proteins. EMBO J., , 15, , 1132–1141.
Stern,D.B., Jones,H. and Gruissem,W. ( (1989) ) Function of plastid mRNA 3' inverted repeats: RNA stabilization and gene-specific protein binding. J. Biol. Chem., , 264, , 18742–18750.
Pestova,T.V. and Hellen,C.U.T. ( (1999) ) Ribosome recruitment and scanning: what's new? Trends Biochem. Sci., , 24, , 85–87.
Gallie,D.R. ( (2002) ) Protein–protein interactions required during translation. Plant Mol. Biol., , 50, , 949–970.(Tetsuro Hirose1 and Masahiro Sugiura1,2,)
* To whom correspondence shoud be addressed at Graduate School of Natural Sciences, Nagoya City University, Yamanohata, Mizuho, Nagoya 467-8501, Japan. Tel/Fax: +81 52 872 6021; Email: sugiura@nsc.nagoya-cu.ac.jp
ABSTRACT
The mechanism of translational initiation differs between prokaryotes and eukaryotes. Prokaryotic mRNAs generally contain within their 5'-untranslated region (5'-UTR) a Shine-Dalgarno (SD) sequence that serves as a ribosome-binding site. Chloroplasts possess prokaryotic-like translation machinery, and many chloroplast mRNAs have an SD-like sequence, but its position is variable. Tobacco chloroplast atpB mRNAs contain no SD-like sequence and are U-rich in the 5'-UTR (–20 to –1 with respect to the start codon). In vitro translation assays with mutated mRNAs revealed that an unstructured sequence encompassing the start codon, the AUG codon and its context are required for translation. UV crosslinking experiments showed that a 50 kDa protein (p50) binds to the 5'-UTR. Insertion of an additional initiation region (SD-sequence and AUG) in the 5'-UTR, but not downstream, arrested translation from the authentic site; however, no inhibition was observed by inserting only an AUG triplet. We hypothesize for translational initiation of the atpB mRNA that the ribosome enters an upstream region, slides to the start codon and forms an initiation complex with p50 and other components.
INTRODUCTION
Translational control is an important means by which cells govern gene expression. The initiation phase of translation usually represents the rate-limiting step, although the mechanism of translational initiation differs between prokaryotes and eukaryotes (1). The 30S ribosomal subunit in prokaryotes binds primarily to mRNA translational initiation regions via start codons combined with Shine-Dalgarno (SD) sequences that are commonly found between 4 and 12 nt upstream from initiation codons (2). On the other hand, the eukaryotic 40S ribosomal subunit enters at the 5' cap of the mRNA, not at the initiation codon. Then the subunit carrying initiator tRNA and initiation factors scans the 5'-untranslated region (5'-UTR) for the first AUG codon. Chloroplasts are plant-specific organelles that possess their own genome (3,4). It has been shown that chloroplast gene expression is primarily regulated at the post-transcriptional level, including translation (5–8). The translation machinery in chloroplasts generally resembles that found in prokaryotes, as chloroplasts contain 70S-type ribosomes and mRNAs lacking 5' caps and eukaryotic 3' poly(A) tails (9,10). Many of the higher plant chloroplast genes encoding polypeptides possess SD-like sequences in 5'-UTRs, although their distance from start codons is highly variable (11,12). In the tobacco chloroplast genome, 30 of the 79 polypeptide-coding genes contain no SD-like sequences (GGA, AGG, GAGG, GGAG or GGAGG) within 20 nt upstream from the initiation codon, and the remaining 49 have SD-like sequences but not at a conserved position (13).
Using chloroplast transformation in Chlamydomonas reinhardtii, mutation analysis of chloroplast SD-like sequences has yielded conflicting results; deletion of the SD-like sequence from its psbA 5'-UTR ceased translation (14), while replacement mutagenesis of the SD-like sequence in the 5'-UTRs of petD (15) and of atpB, atpE, rps4 and rps7 mRNAs (16) had little effect on translation. Further mutation studies suggested that the secondary structure in the 5'-UTR of rps7 mRNA plays an essential role in translational initiation (17) and that the ribosomal protein S7 is also involved in translational initiation (18). In vivo analysis by tobacco chloroplast transformation has shown that the 5'-UTR of psbA mRNAs is important for translation efficiency (19–22) and that sequences downstream of the AUG codon in rbcL and atpB mRNAs are also important determinants of translational efficiency in chloroplasts (23).
In order to dissect biochemical processes of translation unique to higher plant chloroplasts, we have developed an in vitro translation system from tobacco chloroplasts. Our in vitro system supports reproducibly accurate initiation of translation from a variety of chloroplast mRNAs. Hence it provides a powerful tool for the functional analysis of regulatory elements/factors and of the role of low molecular mass compounds (24). Using this in vitro system, we identified three distinct cis-acting elements in the 5'-UTR of tobacco psbA mRNA. Among them, RBS1 (AAG) and RBS2 (UGAUGAU) are complementary to the 3' terminal region of chloroplast 16S rRNA, and the AU-box (UAAAUAAA, –21 to –14) located between RBS1 and RBS2 is likely to be a binding site of putative trans-acting factor(s). An SD-like sequence (GGAG) located at –36 to –33 is not important for translation. We then assessed the role for translation of SD-like sequences in tobacco chloroplast mRNAs (25,26). The atpE, rps14 and rbcL mRNAs possess SD-like sequences (GGAG at –18 to –15, GGA at –14 to –12 and GGAGG at –10 to –6, respectively) at a position similar to the conserved region observed in Escherichia coli mRNAs, and these sequences were found to be essential for translation. On the other hand, SD-like sequences in the rps12 mRNA and in the petB mRNA are located far from (–44 to –42) and too close to (–5 to –2) the initiation codon, respectively, and these sequences are not essential for translation. Unexpectedly, an SD-like sequence (GGAG) at a proper position (–8 to –5) was not required and even inhibitory for translation of tobacco rps2 mRNA (27). In addition, translation of the tobacco ndhD mRNA starts only from the AUG codon created by RNA editing where no SD-like sequence is present in the 5'-UTR, suggesting that translational initiation of ndhD mRNAs depends on a novel sequence element(s) (28,29).
Here, we describe the existence of a novel mode of translational initiation unique to chloroplasts. Translation of the atpB mRNA lacking an SD-like sequence in the 5'-UTR requires an unstructured sequence to which trans-acting factor(s) binds. Based on assays which utilize artificial insertion of an additional translational initiation region and of an AUG triplet into the 5'-UTR of atpB mRNAs, we hypothesize that a chloroplast ribosome enters an upstream region and slides to the initiation codon for translational initiation in chloroplast atpB mRNAs.
MATERIALS AND METHODS
Plasmid construction
Portions of atpB and atpE genes were amplified by PCR from tobacco chloroplast DNA (30,31), cut with appropriate restriction enzymes, and cloned into pBluescript II SK–, in which atpB (–244 to +169) and atpE (–431 to +191) regions were fused in-frame to the 3' lacZ coding region (224 bp). As the atpB overlaps with the downstream atpE, the region encompassing the 3' part of atpB and the entire atpE was replaced by 'lacZ to eliminate possible effects of the atpE mRNA. The calculated size of the hybrid product was 14.3 kDa. The5' deletion constructs were prepared by inverse PCR using appropriate primer pairs and original constructs. Site-directed mutagenesis was performed using a TransformerTM Site Directed Mutagenesis kit Version 2 (Clontech) using primer pairs and original constructs. Constructs were amplified in E.coli XL-1 Blue, isolated by the alkaline lysis method and purified further by Qiagen midi tip (Qiagen). Plasmid DNAs were linearized with BglII, and treated with phenol/chloroform and precipitated with ethanol. mRNAs were synthesized with T3 RNA polymerase using T3 MEGASCRIPT (Ambion) and further purified by Sephadex G50 spin columns (Amersham Pharmacia).
Chloroplast S30 fraction and in vitro translation
Tobacco plants (Nicotiana tabacum var. Blight Yellow 4) were grown in a growth chamber (28°C, 16 h light and 8 h dark) for 4 weeks. Preparation of S30 fractions from intact tobacco chloroplasts and in vitro translation reactions were carried out as previously described (24). The dialyzed S30 fraction (0.5 ml, 15 mg/ml protein) could be stored at –70°C for up to 6 months. In vitro translation reactions were carried out at 30°C for 30 min in 50 μl solution containing 30 mM HEPES–KOH, pH 7.7, 10 mM magnesium acetate, 60 mM potassium acetate, 60 mM NH4Cl, 1% polyethyleneglycol 6000, 1 mM ATP, 0.1 mM GTP, 8 mM creatine phosphate, 0.4 mg/ml creatine phosphokinase (Type I; Sigma), 2 mM dithiothreitol, 0.4 mg/ml E.coli tRNAs (Boehringer), 0.4 mg/ml E.coli tRNAfmet (Sigma), 1 μg/ml leupepcin, 130 U RNase inhibitor (TaKaRa), 0.37 MBq L-methionine (>37 TBq/mmol; Amersham), 20 pmol mRNA and 24 μl (360 μg protein) S30 fraction. We adopted a template-excess condition which yielded high translation activity. The synthesized products were separated by 0.1% SDS-15 or 18% PAGE. Protein size markers were Rainbow markers (Amersham Pharmacia). The separated products were visualized and quantified by a Bio-imaging Analyzer BAS2000 (Fuji Photo Film Co.).
Enzymatic probing of RNA structure and primer extension
Partial digestion of mRNA by RNase T2 and primer extension was performed essentially as previously described (28). mRNAs were synthesized from constructs wt, SC5, SC8, A7 and U7, which were linearized with BssHII followed by purification by denaturing gel electrophoresis. RNA samples (10 pmol each) were preincubated for 20 min at 25°C in 30 mM Hepes–KOH (pH 7.7), 60 mM potassium acetate, 10 mM magnesium acetate and 60 mM NH4Cl, and digested with 0.1 U RNase T2 (Sigma) for 5 min at 25°C. Digested RNA samples were extracted by phenol/chloroform and precipitated with ethanol. The primer (positions +39 to +25) was labeled at its 5' end with T4 polynucleotide kinase and ATP, and annealed to the digested RNAs at 95°C for 1 min followed by cooling to 42°C. Reverse transcription was carried out using a cDNA cycle kit (Invitrogen). Primer extension samples were separated by 10% PAGE containing 7 M urea and visualized by a Bio-imaging Analyzer BAS2000.
UV crosslinking assays
Labeled RNAs were synthesized using an in vitro transcription kit with UTP (110 TBq/mmol) and purified by PAGE. Labeled RNA at a concentration of 10 fmol was incubated in a chloroplast S30 extract under the conditions of an in vitro translation reaction at 30°C for 15 min (24). The S30 fraction (360 μg protein) was excess with respect to the mRNA concentration. UV irradiation followed by RNase A digestion and SDS–PAGE were carried out as previously reported (32).
RESULTS
Sequences required for translation of atpB mRNAs
The tobacco chloroplast atpB gene encoding the ? subunit of chloroplast F0F1 ATP synthase is co-transcribed with its downstream atpE gene for its subunit (33), and transcription of the atpB/E operon starts from multiple sites (34,35). The 20 nt region immediately upstream of the start codon in the atpB 5'-UTR is highly U-rich and possesses no SD-like sequence, suggesting that the chloroplast ribosome recognizes the initiation site through a novel element(s). To identify sequences necessary for the initiation of translation, an atpB–lacZ fusion mRNA was constructed starting from –244 to +169 of the atpB mRNA, which contained an almost complete 5'-UTR of the major atpB mRNA accumulated in tobacco chloroplasts , and a set of 5' deletion mutant mRNAs were also generated (Figure 1A). As shown in Figure 1B, translation occurred at a relatively high level (80% of wt) by deletion to position –53 (5D1), and at a moderate level (40% of wt) by deletion to –25 (5D2). However, translation could scarcely be detected from the mRNA lacking the entire 5'-UTR (5D3). There is an upstream open reading frame (uORF) of 25 codons. The above assays also show that translation of the authentic coding region is hardly affected by the uORF. In vitro translation using two internal deletion mutants showed that the mRNA lacking the –25 to –1 region (ID1) could not be translated, whereas the mutant lacking the –53 to –25 region (ID2) produced translation products, although to a limited extent (20% of wt). Therefore, we concluded that the region from –25 to –1 (or part of it) is essential for translation of atpB mRNAs and the 5'-UTR from –53 to –25 is necessary for efficient translation.
Figure 1. Translation in vitro of tobacco chloroplast atpB mRNAs. The atpB mRNA was fused with lacZ mRNA. (A) Schematic representation of the atpB 5'-UTR (wt), and its 5' deletion (5D) and internal deletion (ID) mutants. The 25 nt sequence upstream of the start codon is shown above. uORF, upstream ORF of 25 codons. (B) 35S-products were separated by PAGE. Translation activity was calculated from band densities (wild type as 100%) and is shown.
In order to define the essential elements more in detail, in vitro translation experiments were carried out using a series of mutant mRNAs produced by scanning mutagenesis in the region (from –25 to +18) spanning the initiation codon (7 nt was sequentially substituted by C-heptamer with 3 nt overlapping, Figure 2A). As shown in Figure 2B, mutations to –11 (SC13) and from +11 (SC9) reduced translation to 20% or less, and from –10 to +10 (SC48) abolished translation almost completely (<5%). These results indicate that both upstream and downstream regions are required for translation of atpB mRNAs. The 20 nt region spanning AUG (–10 to +10) is essential and its flanking sequences are also necessary for efficient translation.
Figure 2. Translation in vitro of mutated atpB mRNAs. (A) Schematic representation of altered regions constructed by replacement mutagenesis using C-heptamers (SC). Hyphens represent identical nucleotides with the wild type (wt). (B) 35S-products were separated by PAGE. Translation activity was calculated from band densities (wt as 100%) and is shown.
Importance of an unstructured sequence for translation of atpB mRNAs
As the essential region (–10 to +10) of atpB mRNA was defined by substitution with C-heptamers, this sequence was then substituted with an A or U stretch (Figure 3A). Translation of A7 and U7 mRNAs was not impaired and that of U17 mRNA was reduced slightly, indicating that no specific sequence in this region is necessary for basic translation and C-stretches abolished translation (Figure 3B).
Figure 3. Translation in vitro of mutated atpB mRNAs and detection of secondary structures in these mRNAs. (A) Schematic representation of mutated regions constructed by replacement mutagenesis. Hyphens represent identical nucleotides with the wild type (wt). (B) 35S-products were separated by PAGE. (C) RNase T2 probing pattern of mRNAs. Vertical arrows on the right of gel patterns show protected sequences; + and –, with and without RNase T2 treatment; +1, the first position of AUG. Ladders appeared in RNase T2 lanes are probably due to slight degradation of mRNA during incubation because no RNase inhibitor was added and also due to cDNA synthesis artefacts.
Secondary structures of the wild-type atpB mRNA and its mutant mRNAs were determined by RNase T2 probing. As shown in Figure 3C, only the C-heptamers in SC5 and SC8 mRNAs were protected from RNase T2 digestion, indicating that these regions form secondary structures. Conversely, the corresponding region in wt, A7 and U7 mRNAs does not form a strong secondary structure, suggesting that an unstructured RNA sequence rather than a specific sequence is important for translation of atpB mRNA.
Importance of the initiation codon and its context for translation of atpB mRNAs
The tobacco atpB gene contains 15 in-frame ATG codons in the coding region and one ATG triplet in the 5'-UTR (up to –244). Our in vitro translation system started translation only from the first AUG codon in the coding region. It is difficult to speculate that an unstructured RNA sequence around the initiation codon is the sole determinant for accurate translation initiation; therefore, an additional determinant(s) is expected to be present within the essential region. We examined first the importance of the context of initiation codon (–3 to +6). Each nucleotide from –3 to +6 (except AUG) was systematically altered by site-directed mutagenesis. As shown in Figure 4A, in vitro translation of these mutant mRNAs revealed that the context is important for translation efficiency. No translation occurred by substitution of +4A with other nucleotides. As substitution of +4A by U creates a termination codon (UGA), both +4A and +6A were substituted by U's. This double mutant almost abolished translation, confirming that U cannot substitute +4A. The –1U can be substituted by A but not by C or G. The +5G can be changed by A or U but not by C. The substitution of +5G with A or U and that of –2U with C increased translation more than 2-fold, suggesting that the wild-type sequence is not always the most efficient substrate, at least in vitro. Taken together, the sequence context from –3 to +6 for the most efficient translation is (A/G)C(A/U)AUGA(A/U)(A/U) (underline indicates a start codon).
Figure 4. Determination of the efficient context and start codon for translation of tobacco chloroplast atpB mRNAs. (A) Translation in vitro of point mutants in the AUG flanking nucleotides. Downward arrows indicate mutated nucleotides. 35S-products were separated by PAGE. Translation activity was calculated from band densities (wt as 100%) and is shown. (B) Effect of replacement from AUG to GUG on translation in vitro. The atpE mRNA with a functional SD sequence (26) was used as a control. 35S-products were separated by PAGE. Translation activity was shown (wt as 100%).
The initiation codon AUG of atpB mRNA was then altered to GUG and, as a control, the AUG initiation codon of atpE mRNA which possesses a functional SD sequence was replaced by GUG. As shown in Figure 4B, the mutant atpE mRNA was translated at 10% of the control efficiency, whereas no translation was observed from the mutant atpB mRNA. This indicates that the AUG codon itself is one of the essential determinants for translation of atpB mRNAs.
Insertion of a translational initiation region into the 5'-UTR arrests authentic translation
Eukaryotic translation is believed to be initiated by the binding of a 40S ribosomal subunit at the 5' end of mRNA mediated by the cap structure, followed by scanning to the first AUG codon (1). To examine the possibility of ribosomal scanning on chloroplast mRNAs, a 13 nt sequence including the SD sequence, spacer and start codon derived from the tobacco chloroplast rbcL mRNA (SD-AUG) was inserted before (–37 to –25) or after (+10 to +22) the authentic initiation codon of atpB mRNA (Figure 5A). The SD sequence (GGAGG) in the rbcL 5'-UTR was found to be functional (26,36). As shown in Figure 5B, insertion of the rbcL initiation region into the 5'-UTR (SD-AUG-atpB) abolished translation from the authentic initiation codon, while translation from the inserted AUG was detected at a low level. Insertion of only an AUG triplet to the same site caused no effect on translation from the authentic site. On the other hand, insertion of the SD-AUG into the downstream region (atpB-SD-AUG) had no effect on translation from the authentic site, and translation from the inserted SD-AUG was also detected. Translation occurred exclusively from the SD-AUG inserted in the downstream region when the authentic AUG was altered to UAC, indicating that translational initiation from the rbcL initiation region is independent of surrounding sequences.
Figure 5. Effect on translation in vitro of an additional initiation region inserted in the upstream or downstream region from the AUG in atpB mRNAs. (A) Schematic representation of mRNA constructs. The inserted sequence from the tobacco chloroplast rbcL mRNA is shown. UAC-SD-AUG mRNA possesses UAC in place of the authentic AUG. (B) 35S-products were separated by PAGE. Asterisks indicate products from the inserted SD-AUG site.
Protein factor(s) specifically binds to the U-rich upstream element
In order to detect proteins that associate with the translational initiation region of atpB mRNA, UV crosslinking experiments were carried out using the in vitro translation extract. As shown in Figure 6, several chloroplast proteins were detected by crosslinking to the 32P-labeled atpB mRNA (lane wt). Experiments with a set of mutant atpB mRNAs used for in vitro translation showed that a 50 kDa protein (p50) could not bind to the mRNAs whose 25 nt upstream region was deleted (ID1) or replaced by a C-heptamer (SC5). On the other hand, p50 could crosslink with the mRNA in which AUG was replaced by UAC (UAC) or the downstream region was replaced by a C-heptamer (SC8). A protein of 40 kDa and several other proteins of 28–33 kDa bind to all mutant mRNAs as well as the wild-type mRNA, suggesting that these proteins are non-specific RNA-binding proteins in chloroplasts (37). This result suggests that p50 preferentially binds to the 25 nt upstream region and promotes translational initiation.
Figure 6. Detection of tobacco chloroplast proteins bound to atpB mRNAs by UV crosslinking. mRNA constructs are depicted above: the detail of wt and ID1 mRNAs is shown in Figure 1A and that of SC5 and SC8 mRNAs is in Figure 2A, and UAC mRNA possesses UAC in place of AUG. C7, C-heptamer. Uniformly labeled atpB mRNAs were incubated with a chloroplast extract and UV irradiated. After RNase A treatment, 32P-proteins were separated by PAGE. Size markers are shown to the right. Translation activity was shown (wt as 100%).
DISCUSSION
Distinct mechanisms of translational initiation in chloroplasts
Previously we examined the effect on translation of SD-like sequences in tobacco chloroplast mRNAs, and found that SD-containing mRNAs can be categorized into two groups (25,26); (i) translation dependent on SD-like sequences and (ii) translation independent of SD-like sequences. Unlike these cases, sequences which lack the SD motif are found in the 5'-UTR of the tobacco atpB mRNA.
Mutation analysis of its 5'-UTR and N-terminal protein-coding region revealed that at least three critical elements affect translation of atpB mRNAs. An unstructured sequence rather than a specific sequence encompassing the initiation codon is important for efficient translation. The tobacco atpB mRNA possesses a characteristic upstream sequence (–1 to –20) consisting of 17 U and 3 A residues (see Figure 1A). This 5'-UTR is relatively similar to that of the rbcL mRNA in Euglena chloroplasts, and in addition its translational initiation was reported to be affected either by its unstructured or weakly structured region rather than by a specific sequence (38). The context of the initiation codon affects translational efficiency of atpB mRNAs. Among the 79 tobacco chloroplast protein-coding genes, only 30 genes have A at +4, and only ndhD perfectly matches the above context (A C U AUG A A U) whose AUG is created from ACG by RNA editing (28). The ndhD mRNA has no SD-like sequence but the U-rich sequence in the 5'-UTR, suggesting that translation initiates in a similar manner to that of atpB mRNA. The other chloroplast mRNAs (e.g. psbA and atpE) which are efficiently translated in vitro have no context similar to that of atpB mRNA, suggesting that the context is specific for a set of mRNAs with U-rich leaders. Recently, in vivo evidence has been reported that the extended codon-anticodon interaction (between the nucleotide 3' to the initiator tRNA anticodon and the base immediately preceding the initiation codon) is important for initiation of Chlamydomonas petA mRNA (39). Only 35 of the 79 mRNAs in tobacco chloroplasts possess U at the –1 position. Therefore, the expanded codon-anticodon model is unlikely for translation of the rest of the tobacco mRNAs. AUG is essential for atpB mRNA translation, which cannot be substituted by GUG, whereas atpE mRNA can be translated from a GUG codon. In tobacco chloroplasts, psbC, rps19 and ycf15 possess GTG as initiation codons (13). All three mRNAs are preceded by typical SD-like elements, suggesting that the chloroplast ribosome can recognize GUG as an initiation codon when a functional SD sequence is present.
Our UV crosslinking assay revealed the interaction of a 50 kDa RNA-binding protein (p50) with the 25 nt upstream region of atpB mRNA. An RNA-binding protein of 47 kDa was reported to bind to the spinach psbA 5' UTR and it was identified as the chloroplast homolog of the E.coli ribosomal protein S1 (40). The tobacco p50 is similar in size to the spinach S1 protein and might be its tobacco homolog. A maize mutant, atp1, was shown to reduce synthesis of the atpB product, and atp1 function was suggested to be required during translational initiation (41). Hence, p50 may be encoded by atp1. Chloroplast RNA-binding proteins 50 kDa in size have been reported as factors possibly involved in endonucleolytic mRNA processing in mustard and in mRNA 3' end formation in spinach . These proteins preferentially bind to U-rich sequences, suggesting that the chloroplast possesses a large family of RNA-binding proteins with high affinity to U-rich sequences.
Ribosomal scanning in chloroplasts?
Elements required for atpB mRNA translation are (i) an AUG start codon, (ii) an unstructured RNA sequence surrounding the AUG and (iii) the context of the initiation codon, which apparently corresponds to the important elements involved in cytoplasmic translational initiation (1,46,47). Translation usually starts at the first AUG in cytoplasmic mRNAs. The 40S ribosomal subunit binds to the 5'-end of mRNA and then scans the 5'-UTR, searching for the initiation codon. An AUG triplet inserted upstream of the authentic initiation codon replaces it as the initiation codon.
Translation from the authentic initiation codon of atpB mRNA was impaired neither by the presence of uORF (see Figure 1) nor by an AUG triplet inserted (–27 to –25) upstream of wt initiation region (see Figure 5), suggesting that the scanning model cannot simply be applied to explain translational initiation in chloroplasts. However, insertion of an SD sequence plus AUG derived from tobacco rbcL mRNA into the same site completely arrested translation from the authentic initiation site; instead, translation initiated from the inserted site though low in efficiency (see Figure 5). On the other hand, the insertion of the same sequence into the downstream region (+10 to +22) did not affect translation from the authentic site. This could exclude the possibility that the inserted SD-AUG and the authentic initiation region compete to capture ribosomes. A plausible explanation is that ribosomal binding to the SD-AUG sequence blocks sliding of the 30S ribosomal subunit from the upper side. As translation activity of tobacco rbcL mRNA is slow in vitro (our unpublished data), the mRNA–ribosome complex via the SD-AUG sequence is likely to be relatively stable, which may stall ribosomal subunits at this point. Taken together, we favor the hypothesis that the 30S subunit enters a 5' distal region, slides to the AUG initiation codon and forms an initiation complex with the aid of p50 and other components. As chloroplast mRNAs have no cap structure, the ribosome entry site, if any, can either be at the 5' end, within the 5'-UTR, or both. One intriguing possibility is that ribosomes primarily bind to a sequence between –53 and –25, because a deletion up to –53 still possesses near full activity but insertion of the SD-AUG at –25 blocks translation. However, other possibilities cannot be ruled out and we need further analyses.
Our in vitro translation experiments revealed the existence of a novel mechanism of translational initiation in chloroplasts. In E.coli, translation of most mRNAs is thought to be initiated via the SD sequence. On the other hand, sequencing of a cyanobacterial ribosomal gene cluster showed that the SD sequence is not always a common sequence element in the prokaryotic genes (13). The mechanism reported here might operate both in chloroplasts and some eubacteria.
ACKNOWLEDGEMENTS
We thank Dr Takahiro Nakamura for his critical reading of this manuscript. This work was supported in part by a Grant-in-Aid for Scientific Research (B) (No. 12440226) and Exploratory Research (No. 13874111) from the Ministry of Education, Science, Sports and Culture.
REFERENCES
Kozak,M. ( (1999) ) Initiation of translation in prokaryotes and eukaryotes. Gene, , 234, , 187–208.
McCarthy,J.E.G. and Gualerzi,C. ( (1990) ) Translational control of prokaryotic gene expression. Trends Genet., , 6, , 78–85.
Sugiura,M. ( (1992) ) The chloroplast genome. Plant Mol. Biol., , 19, , 149–168.
Wakasugi,T., Tsudzuki,T. and Sugiura,M. ( (2001) ) The genomics of land plant chloroplasts: gene content and alternation of genomic information by RNA editing. Photosynthesis Res., , 70, , 107–118.
Stern,D.B., Higgs,D.C. and Yang,J.J. ( (1997) ) Transcription and translation in chloroplasts. Trends Plant Sci., , 2, , 308–315.
Bruick,R.K. and Mayfield,S.P. ( (1999) ) Light-activated translation of chloroplast mRNAs. Trends Plant Sci., , 4, , 190–195.
Zerges,W. ( (2000) ) Translation in chloroplasts. Biochimie, , 82, , 583–601.
Choquet,Y. and Wollmann,F.-A. ( (2002) ) Translational regulations as specific traits of chloroplast gene expression. FEBS Lett., , 529, , 39–42.
Yamaguchi,K., von Knoblauch,K. and Subramanian,A.R. ( (2000) ) The plastid ribosomal proteins: identification of all the proteins in the 30 S subunit of an organelle ribosome (chloroplast). J. Biol. Chem., , 275, , 28455–28465.
Yamaguchi,K., Prieto,S., Beligni,M.V., Haynes,P.A., McDonald,W.H., Yates,J.R. and Mayfield,S.P. ( (2002) ) Proteomic characterization of the small subunit of Chlamydomonas reinhardtii chloroplast ribosome: identification of a novel S1 domain-containing protein and unusually large orthologs of bacterial S2, S3, and S5. Plant Cell, , 14, , 2957–2974.
Ruf,M. and K?ssel,H. ( (1988) ) Occurrence and spacing of ribosome recognition sites in mRNAs of chloroplasts from higher plants. FEBS Lett., , 240, , 41–44.
Bonham-Smith,P.C. and Bourque,D.P. ( (1989) ) Translation of chloroplast-encoded mRNA: potential initiation and termination signals. Nucleic Acids Res., , 17, , 2057–2080.
Sugiura,M., Hirose,T. and Sugita,M. ( (1998) ) Evolution and mechanism of translation in chloroplasts. Annu. Rev. Genet., , 32, , 437–459.
Mayfield,S.P., Cohen,A., Danon,A. and Yohn,C.B. ( (1994) ) Translation of the psbA mRNA of Chlamydomonas reinhardtii requires a structured RNA element contained within the 5' untranslated region. J. Cell Biol., , 127, , 1537–1545.
Sakamoto,W., Chen,X., Kindle,K.L. and Stern,D.B. ( (1994) ) Function of the Chlamydomonas reinhardtii petD 5' untranslated region in regulating the accumulation of subunit IV of the cytochrome b6/f complex. Plant J., , 6, , 503–512.
Fargo,D.C., Zhang,M., Gillham,N.W. and Boynton,J.E. ( (1998) ) Shine-Dalgarno-like sequences are not required for translation of chloroplast mRNAs in Chlamydomonas reinhardtii chloroplasts or in Escherichia coli. Mol. Gen. Genet., , 257, , 271–282.
Fargo,D.C., Boynton,J.E. and Gillham,N.W. ( (1999) ) Mutations altering the predicted secondary structure of a chloroplast 5' untranslated region affect its physical and biochemical properties as well as its ability to promote translation of reporter mRNAs in both the Chlamydomonas reinhardtii chloroplast and in Escherichia coli. Mol. Cell. Biol., , 19, , 6980–6990.
Fargo,D.C., Boynton,J.E. and Gillham,N.W. ( (2001) ) Chloroplast ribosomal protein S7 of Chlamydomonas binds to chloroplast mRNA leader sequences and may be involved in translation initiation. Plant Cell, , 13, , 207–218.
Staub,J.M. and Maliga,P. ( (1993) ) Accumulation of D1 polypeptide in tobacco plastids is regulated via the untranslated region of the psbA mRNA. EMBO J., , 12, , 601–606.
Staub,J.M. and Maliga,P. ( (1994) ) Translation of psbA mRNA is regulated by light via the 5'-untranslated region in tobacco plastids. Plant J., , 6, , 547–553.
Eibl,C., Zou,Z., Beck,A., Kim,M., Mullet,J. and Koop,H.-U. ( (1999) ) In vivo analysis of plastid psbA, rbcL and rpl32 UTR elements by chloroplast transformation: tobacco plastid gene expression is controlled by modulation of transcript levels and translation efficiency. Plant J., , 19, , 333–345.
Zou,Z., Eibl,C. and Koop,H.-U. ( (2003) ) The stem–loop region of the tobacco psbA 5'UTR is an important determinant of mRNA stability and translation efficiency. Mol. Gen. Genomics, , 269, , 340–349.
Kuroda,H. and Maliga,P. ( (2001) ) Sequences downstream of the translation initiation codon are important determinants of translation efficiency in chloroplasts. Plant Physiol., , 125, , 430–436.
Hirose,T. and Sugiura,M. ( (1996) ) Cis-acting elements and trans-acting factors for accurate translation of chloroplast psbA mRNAs: development of an in vitro translation system from tobacco chloroplasts. EMBO J., , 15, , 1687–1695.
Hirose,T., Kusumegi,T. and Sugiura,M. ( (1998) ) Translation of tobacco chloroplast rps14 mRNA depends on a Shine-Dalgarno-like sequence in the 5'-untranslated region but not on internal RNA editing in the coding region. FEBS Lett., , 430, , 257–260.
Hirose,T. and Sugiura,M. ( (2004) ) Functional Shine-Dalgarno-like sequences for translational initiation of chloroplast mRNAs. Plant Cell Physiol., , 45, , 114–117.
Plader,W. and Sugiura,M. ( (2003) ) The Shine-Dalgarno-like sequence is a negative regulatory element for translation of tobacco chloroplast rps2 mRNA: an additional mechanism for translational control in chloroplasts. Plant J., , 34, , 377–382.
Hirose,T. and Sugiura,M. ( (1997) ) Both RNA editing and RNA cleavage are required for translation of tobacco chloroplast ndhD mRNA: a possible regulatory mechanism for the expression of a chloroplast operon consisting of functionally unrelated genes. EMBO J., , 16, , 6804–6811.
Sasaki,T., Yukawa,Y., Miyamoto,T., Obokata,J. and Sugiura,M. ( (2003) ) Identification of RNA editing sites in chloroplast transcripts from the maternal and paternal progenitors of tobacco (Nicotiana tabacum): comparative analysis shows the involvement of distinct trans-factors for ndhB editing. Mol. Biol. Evol., , 20, , 1028–1035.
Shinozaki,K., Ohme,M., Tanaka,M., Wakasugi,T., Hayashida,N., Matsubayashi,T., Zaita,N., Chungwongse,J., Obokata,J., Yamaguchi-Shinozaki,K. et al. ( (1986) ) The complete nucleotide sequence of the tobacco chloroplast genome: its gene organization and expression. EMBO J., , 5, , 2043–2049.
Sugiura,M., Shinozaki,K., Zaita,N., Kusuda,M. and Kumano,M. ( (1986) ) Clone bank of the tobacco (Nicotiana tabacum) chloroplast genome as a set of overlapping restriction endonuclease fragments: mapping of eleven ribosomal protein genes. Plant Sci., , 44, , 211–216.
Hirose,T. and Sugiura,M. ( (2001) ) Involvement of a site-specific trans-acting factor and a common RNA-binding protein in the editing of chloroplast mRNAs: development of a chloroplast in vitro RNA editing system. EMBO J., , 20, , 1144–1152.
Shinozaki,K., Deno,H., Kato,A. and Sugiura,M. ( (1983) ) Overlap and cotranscription of the genes for the beta and epsilon subunits of tobacco chloroplast ATPase. Gene, , 24, , 147–155.
Kapoor,S., Wakasugi,T., Deno,H. and Sugiura,M. ( (1994) ) An atpE-specific promoter within the coding region of the atpB gene in tobacco chloroplast DNA. Curr. Genet., , 26, , 263–268.
Kapoor,S., Suzuki,J.Y. and Sugiura,M. ( (1997) ) Identification and functional significance of a new class of non-consensus-type plastid promoters. Plant J., , 11, , 327–337.
Kim,J. and Mullet,J.E. ( (1994) ) Ribosome-binding sites on chloroplast rbcL and psbA mRNAs and light-induced initiation of D1 translation. Plant Mol. Biol., , 25, , 437–448.
Li,Y. and Sugiura,M. ( (1990) ) Three distinct ribonucleoproteins from tobacco chloroplast: each contains a unique amino terminal acidic domain and two ribonucleoprotein consensus motifs. EMBO J., , 9, , 3059–3066.
Koo,J.S. and Spremulli,L.L. ( (1994) ) Effect of the secondary structure in the Euglena gracilis chloroplast ribulose-bisphosphate carboxylase/oxygenase messenger RNA on translational initiation. J. Biol. Chem., , 269, , 7501–7508.
Esposito,D., Fey,J.P., Eberhard,S., Hicks,A.J. and Stern,D.B. ( (2003) ) In vivo evidence for the prokaryotic model of extended codon-anticodon interaction in translation initiation. EMBO J., , 22, , 651–656.
Alexander,C., Faber,N. and Klaff,P. ( (1998) ) Characterization of protein-binding to the spinach chloroplast psbA mRNA 5' untranslated region. Nucleic Acids Res., , 26, , 2265–2272.
McCormac,D.J. and Barkan,A. ( (1999) ) A nuclear gene in maize required for the translation of the chloroplast atpB/E mRNA. Plant Cell, , 11, , 1709–1716.
Nickelsen,J. and Link,G. ( (1991) ) RNA–protein interactions at transcript 3' ends and evidence for trnK-psbA cotranscription in mustard chloroplasts. Mol. Gen. Genet., , 228, , 89–96.
Chen,H. and Stern,D.B. ( (1991) ) Specific binding of chloroplast proteins in vitro to 3' untranslated region of spinach chloroplast petD mRNA. Mol. Cell. Biol., , 11, , 4380–4388.
Hayes,R., Kudla,J., Schuster,G., Gabay,L., Maliga,P. and Gruissem,W. ( (1996) ) Chloroplast mRNA 3'-end processing by a high molecular weight protein complex is regulated by nuclear encoded RNA binding proteins. EMBO J., , 15, , 1132–1141.
Stern,D.B., Jones,H. and Gruissem,W. ( (1989) ) Function of plastid mRNA 3' inverted repeats: RNA stabilization and gene-specific protein binding. J. Biol. Chem., , 264, , 18742–18750.
Pestova,T.V. and Hellen,C.U.T. ( (1999) ) Ribosome recruitment and scanning: what's new? Trends Biochem. Sci., , 24, , 85–87.
Gallie,D.R. ( (2002) ) Protein–protein interactions required during translation. Plant Mol. Biol., , 50, , 949–970.(Tetsuro Hirose1 and Masahiro Sugiura1,2,)