DDBJ/EMBL/GenBank accession nos
Y15414-Y15416
ABSTRACT
Many RNA viruses express part of their genomic information by read-through over internal termination codons. We have recently characterized tobacco cytoplasmic (cyt) and chloroplast (chl) tRNACmCATrp and tRNAGCACys as natural suppressor tRNAs that are able to read the leaky UGA codon in RNA-1 of tobacco rattle virus, albeit with different efficiencies. Here we have identified a third natural UGA suppressor in plants. We have purified and sequenced four cyt tRNAArg isoacceptors with ICG, CCG, U*CG and CCU anticodons from wheat germ. With the exception of tRNAICGArg, these are the first sequences of plant tRNAsArg. In order to study the potential suppressor activity of wheat tRNAsArg we have usedin vitro synthesized mRNA transcripts in which different viral read-through regions had been placed. In vitro translation was carried out in a homologous wheat germ extract. We found that tRNAU*CGArg is an efficient UGA suppressorin vitro, whereas the other three tRNAArg isoacceptors exhibit no or very low suppressor activity. Interaction of tRNAU*CGArg with the UGA codon requires a G:U base pair at the third anticodon position. This is the first time that an arginine-accepting tRNA has been characterized as a natural UGA suppressor. A remarkable feature of cyt tRNAU*CGArg is its ability to misread the UGA at the end of the coat protein cistron in RNA-1 of pea enation mosaic virus, which is not accomplished by cyt tRNACmCATrp or cyt tRNAGCACys, due to an unfavourable codon context.
Protein synthesis is normally terminated by either of three termination codons UAG, UAA or UGA. Yet many RNA viruses express part of their genomic information by read-through over internal termination codons. These `leaky' stop codons can be read as sense codons by specific tRNAs, called natural suppressor tRNAs (1,2). Until now only UAG and UGA and not UAA have been identified as leaky stop codons in viral RNAs (1). The best known examples among animal viruses which employ read-through of termination codons are moloney murine leukemia virus (MoMuLV), a type C retrovirus, and sindbis virus (SINV), belonging to the alphavirus genus (1). A great deal of information is available about suppressible stop codons in plant viral RNAs, among them members of the luteovirus, enamovirus, tymovirus, carmovirus, tobamovirus, tobravirus and furovirus groups (1,2), a trivial reason being that single-stranded RNA viruses are represented by many more families in plants than in animals (3).
The best-studied plant viruses that utilize stop codon suppression are tobacco mosaic virus (TMV) and tobacco rattle virus (TRV). Read-through over the UAG codon at the end of the 126 kDa cistron in TMV RNA generates a 183 kDa polypeptide in vivo and in vitro (4). Similarly, suppression of the UGA codon at the end of the 134 kDa cistron in RNA-1 of TRV yields a polypeptide of 194 kDa in vivo and in vitro (5,6). Both read-through products are thought to be involved in RNA replication (7,8), implying that they are essential for virus multiplication.
We have previously isolated and sequenced a number of natural suppressor tRNAs in plants: cytoplasmic tRNAG[psi]ATyr, which promotes UAG suppression in TMV RNA (4,9), cytoplasmic (cyt) as well as chloroplast (chl) tRNACmCATrp and, to a lesser extent, cyt and chl tRNAGCACys, which stimulate read-through over the leaky UGA stop codon in RNA-1 of TRV in vitro (6,10). Pea enation mosaic virus (PEMV), the type member of the genus enamovirus, belongs to a group of plant viruses that generate an extended coat protein by translational read-through. Thus, by suppression of the UGA at the end of the 21 kDa coat protein cistron a minor structural protein of 54 kDa is synthesized which is essential for aphid transmission (11). We have found that the UGA in the PEMV context is suppressed to a low level by tobacco chl tRNACmCATrp, but not by cyt tRNACmCATrp or tRNAGCACys (10,12). Since accumulation of PEMV particles occurs mainly in the cytoplasm of all cell types and not in the chloroplasts (11), it can be ruled out that chl tRNATrp participates in suppression of the UGA codon in RNA-1 of PEMV. Consequently, it is reasonable to assume that another UGA suppressor tRNA exists in plants.
A good candidate for a potential UGA suppressor is arginine tRNA. Feng et al. (13) have demonstrated that mutation of the leaky UAG codon at the gag-pol junction in MoMuLV RNA to a UGA codon resulted in incorporation of tryptophan, cysteine and arginine at the corresponding position in the read-through product upon translation in reticulocyte lysate, indicating that unknown cysteine and arginine tRNA species might be UGA suppressors. The tRNAArg family in higher eukaryotes consists of five isoacceptors, of which tRNAArg with a U*CG anticodon is the most likely candidate for a UGA suppressor, since recognition of the UGA stop codon only involves an unconventional G:U base pairing at the third anticodon position.
In order to examine the suppressor activity of plant tRNAsArg we have isolated and sequenced four tRNAArg isoacceptors from wheat germ and ascertained that tRNAU*CGArg is a potential suppressor which is capable of misreading the UGA in the TRV as well as in the PEMV context.
SP6 RNA polymerase, T4 polynucleotide kinase, RNase inhibitor from human placenta, T4 RNA ligase, nuclease P1 and benzoylated DEAE-cellulose were from Boehringer (Mannheim, Germany). 14C-Methylated proteins used as molecular weight markers for gel electrophoresis and [3H]tryptophan, [3H]arginine and [35S]-methionine with specific activities of 1.15, 2.15 and 37 Tbq/mmol respectively were obtained from Amersham (Braunschweig, Germany). Untreated wheat germs were a gift from SynPharma GmbH (Augsburg, Germany).
The recombinant plasmid pSP65-ML1 carries a zein gene from maize seedlings cloned into the BamHI and PstI sites of pSP65 vector DNA (6,14). The construction of the expression vectors pSP65-TRV, pSP65-SINV, pSP65-PEMV and pSP65-globin are described elsewhere (6,12).
Preparation of unfractionated tRNA from wheat germ was performed essentially as described by Beier et al. (4). Total wheat germ tRNA was loaded onto a BD-cellulose column and fractionated by elution with a linear gradient of 0.35-1 M NaCl in 10 mM MgCl2, 12 mM NaOAc, pH 4.5. The tRNAArg-specific fractions were further purified by successive PAGE in a native 10% polyacrylamide gel, pH 8.3 (containing 9% v/v glycerol and 0.3* TBE), followed by a 10% denaturing polyacrylamide gel, pH 3.5 (containing 8 M urea and 25 mM citric acid) and a 12.5% denaturing polyacrylamide gel, pH 8.3.
Assay of amino acid acceptance of wheat tRNAs was performed in a reaction mixture containing 100 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 55 mM KCl, 3 mM dithiothreitol, 5 mM ATP, 0.1 mM CTP, 150 Bq/µl 3H-labelled tryptophan or arginine and appropriate amounts of tRNA and aminoacyl-tRNA synthetase, which was prepared from wheat germ as described (6).
Nucleotide sequences were determined essentially according to Stanley and Vassilenko (15), with some modifications as described by Beier et al. (4). 5'-Labelling with T4 polynucleotide kinase, 3'-pCp-labelling with T4 RNA ligase and controlled enzymatic digestion of end-labelled tRNA were performed as described by Beier and Gross (16).
For tRNAArg cDNA synthesis aliquots of gel fractions containing either of the purified tRNAArg isoacceptors were incubated with 10 pmol oligonucleotide primer (complementary to the 3'-end of the specific tRNAArg) and denatured for 5 min at 90°C in a total volume of 12 µl distilled water. Reverse transcription with Superscripttm II was performed as recommended by the manufacturer (Gibco BRL). One tenth of the reaction mixture was taken for PCR amplification using appropriate primers, 200 µM each dNTP and 2.5 U Pfu DNA polymerase (Stratagene) in a 50 µl final volume. PCR products were cloned into the SmaI site of pUC19 DNA.
Template DNA was prepared by PstI digestion of the plasmid DNAs. Run-off transcription with SP6 RNA polymerase was carried out as described by Zerfa[beta] and Beier (6).
A wheat germ cell-free extract partially depleted of endogenous RNAs was prepared essentially as described by Pfitzinger et al. (17). The extract we use is strictly mRNA-dependent, as recently shown by Zerfa[beta] and Beier (6,9), however, it contains some residual tRNAs which are needed for basic translation in studies where only specific tRNA-enriched column fractions or highly purified tRNAs are added to the wheat germ extract. In vitro translation was performed for 1 h at 30°C in the presence of 11.9 Mbq/ml [35S]methionine in a total volume of 10 µl containing 25% (v/v) wheat germ extract and 10% (v/v) wheat germ initiation factor solution as described recently (10).
Proteins were analysed by electrophoresis in 15% polyacrylamide slab gels containing 0.1% SDS. Gels were fixed overnight, fluorographed and exposed to Dupont Cronex 4 Medical X-ray film at -80°C. The ratio of the protein products derived from termination and read-through events was determined by densitometric quantification of incorporated radioactive methionine using an Elscript 400 scanner.
In order to study the capability of plant tRNAsArg to serve as potential natural UGA suppressors we used zein mRNA transcripts into which three different viral read-through regions as well as the region surrounding the UGA codon at the end of [beta]-globin mRNA had been inserted (Fig. 1). The three read-through regions all contained a leaky UGA and six codons flanking this stop codon in tobacco rattle virus (TRV) and pea enation virus (PEMV) RNA-1, as well as in sindbis virus RNA. Expression of the full-length zein protein of 26 kDa was investigated in a wheat germ extract partially depleted of tRNAs in the presence of tRNAArg-enriched fractions.
The four purified tRNAArg isoacceptors were examined with respect to their UGA suppressor activity. The transcript derived from pSP65-TRV was translated in the tRNA-depleted wheat germ extract in the presence of either of the four tRNAsArg and in the absence of exogenous tRNA. A significant synthesis of the 26 kDa read-through protein (>4%) was observed only if tRNAArg with a U*CG anticodon had been added to the translation mixture. The two tRNAArg isoacceptors with CCG and CCU anticodons were unable to bypass the UGA codon in the TRV context, whereas tRNAICGArg expressed low read-through activities of ~2% (not shown). Thus it is only one out of four tRNAArg isoacceptors present in the cytoplasm of wheat germ that promotes read-through over the UGA stop codon efficiently.
In a recent study we have shown that tobacco chl and cyt tRNAsTrp with a CmCA anticodon stimulate read-through of the leaky UGA stop codon in plant and animal viral RNAs, albeit to a different extent. The calculated read-through activities in the presence of chl tRNACmCATrp were 24, 18 and 6% using the transcripts containing a UGA codon in the TRV, SINV and PEMV contexts respectively and 8, 10, and 1.5% if cyt tRNACmCATrp had been added to the translation mixture (12). In the following experiments we used the same constructs, i.e. pSP65-TRV, pSP65-SINV and pSP65-PEMV DNA, for in vitro transcription and subsequent translation in wheat germ extract in the presence of tRNAU*CGArg. The UGA codon in all three viral contexts was suppressed, as seen in Figure 5. The estimated absolute read-through activities were 4.5, 8 and 7% for the transcripts containing a UGA codon in the TRV, SINV and PEMV contexts respectively. No suppression by tRNAU*CGArg of the UGA in the globin context was observed (Fig. 5).
The six arginine codons CGG, CGA, CGC, CGU, AGG and AGA are read by four isoacceptors in Escherichia coli (24) and by five isoacceptors in the cytosol of higher eukaryotes (21). We have purified and sequenced the four cyt tRNAArg isoacceptors with ICG, CCG, U*CG and CCU anticodons from wheat germ (Fig. 4). Of these, only the major isoacceptor (tRNAICGArg) has already been characterized in wheat germ (18); the other three sequences are the first identified in plants at either the tRNA or DNA level. We have not detected tRNAArg with a U*CU anticodon in wheat germ. This isoacceptor reads AGA, which is a codon rarely used in monocots (25). In fact, Hatfield and Rice (19) were unable to clearly identify this isoacceptor by RPC-5 chromatography of tRNAs from wheat germ and by studying codon recognition properties of isolated fractions. Thus it is very likely that tRNAArg with a U*CU anticodon is a very minor isoacceptor.
We have fractionated total tRNA from wheat germ and identified tRNAsArg by aminoacylation with a synthetase preparation from wheat germ. During the subsequent purification procedure we have not detected any tRNAArg isoacceptor from organelles, a result that we had anticipated for two reasons. First, it is known that the genomes of plant mitochondria do not contain native or functional chloroplast-like tRNAArg genes (26,27). The tRNAsArg are of nuclear origin and the corresponding cyt tRNAsArg are imported into all plant mitochondria studied so far (28). Second, chl tRNAs and chl aminoacyl synthetases are presumably under-represented in wheat germ, since these contain only proplastids. Moreover, cyt arginyl-tRNA synthetase is very likely unable to charge chl tRNAsArg efficiently. In this regard it should be noted that the discriminator base of tRNAsArg encoded by chloroplast genomes is an A (29-31), whereas it is a G in plant cyt tRNAsArg (Fig. 4). Furthermore, the two chl isoacceptors with ICG and UCU anticodons differ in a number of other nucleotides from their cyt counterparts, so that we can definitely say that all tRNAsArg identified by us in wheat germ are of nuclear origin.
It is obvious that the four cyt tRNAArg isoacceptors differ considerably in their nucleotide sequences and base modifications (Fig. 4), which raises the question of how arginyl-tRNA synthetase recognizes its cognate substrates. Sissler et al. (32) have studied identity elements employed by the yeast enzyme and found that arginine aminoacylation is basically linked to the C35 residue, present in all tRNAArg isoacceptors, and that distinction from the non-cognate tRNAAsp is achieved by a negative discrimination mechanism involving the single methyl group in m1G at position 37 of tRNAAsp.
The four tRNAArg isoacceptors from wheat germ were examined for their ability to read the UGA stop codon in vitro. For these studies we employed a mRNA- and tRNA-depleted wheat germ extract (17). Thus the in vitro translation experiments were performed in an entirely homologous system. In a first series of studies we used the pSP65-TRV transcript as template, which contains in-frame the leaky UGA codon of TRV RNA-1 (Fig. 1). We found that tRNAArg with a U*CG anticodon stimulated read-through over this UGA codon to ~5% (Fig. 5), whereas the two tRNAArg isoacceptors with CCG and CCU anticodons did not suppress this stop codon at all. A low but reproducible UGA suppressor activity of ~2% was observed in the presence of tRNAICGArg. It should be mentioned in this context that chl tRNAICGArg is able to read all four arginine codons of the CGN type in wheat germ extract (33).
The tRNAU*CGArg isoacceptor normally decodes the CGA codon. Interaction of this tRNA with a UGA stop codon necessitates a G:U base pairing at the third anticodon position, which is not in accordance with the `wobble' rules. Crick (34) postulated that G:U and U:G as well as I:U, I:C and I:A interactions at the first anticodon position (i.e. wobble position) would be tolerated and would not affect fidelity of protein synthesis due to degeneracy of the genetic code. In a recent investigation using three Tetrahymena tRNAGln isoacceptors as tools for studying unorthodox codon reading we have observed that normal cyt tRNAUmUGGln suppresses UAG and UAA stop codons in the TMV- and TRV-specific contexts in wheat germ extract (35), indicating that G:U base pairing has in fact occurred at the third anticodon position. It has been discussed that the nature of the nucleoside 3' of the anticodon at position 37 restricts or enhances unconventional base interactions (36-38), but we do not know to what extent m1G, present in tRNAU*CGArg at this position, contributes to the observed unorthodox codon reading.
An absolute prerequisite for all types of stop codon read-through appears to be a favourable codon context. Thus it was found that the six nucleotides downstream of the leaky UAG in TMV and several other plant viral RNAs, i.e. CAAUUA, confer leakiness to UAG as well as to UAA and UGA (39,40) and that changes at specific positions of this sequence completely abolished read-through by tRNAG[psi]ATyr (9). Similarly, the ability of plant tRNACmCATrp to promote read-through over the leaky UGA in TRV RNA-1 depends mainly on the three downstream nucleotides. However, in contrast to the very well-defined nucleotide sequence required for efficient UAG suppression by tRNAG[psi]ATyr (9), single nucleotide exchanges at either of the three downstream positions had little effect on UGA suppression by tRNACmCATrp (12). The suppressor activity of wheat tRNAU*CGArg also relies on a suitable codon context. Thus we have shown that UGA in the globin context is not recognized by tRNAU*CGArg. On the other hand, the same tRNA is able not only to read the leaky UGA in the TRV, but also in the PEMV and SINV contexts (Figs 1 and 5).
Sindbis virus contains an in-frame UGA termination codon in the non-structural protein coding region separating nsP3 and nsP4. This internal UGA codon can be suppressed in cultured cells of chicken, human and insect origin (41). Interestingly, in the closely related semliki forest virus (SFV), which also belongs to the alphaviruses, there is no UGA, but instead a CGA arginine codon at this position (42), indicating that possibly the leaky UGA in sindbis virus RNA is recognized by tRNAU*CGArg, which routinely reads the CGA codon.
A remarkable feature of wheat tRNAU*CGArg is its ability to suppress the UGA codon in the PEMV context, even better than in the TRV context (Fig. 5). The PEMV context is clearly different from the TRV and SINV nucleotide context (Fig. 1). We have recently found that UGA in the PEMV context is in fact not or only to <2% suppressed by cyt tRNAGCACys and cyt tRNACmCATrp respectively (10,12). Only chl tRNACmCATrp expressed higher suppressor activities in the range 2-6% (12). However, as mentioned above, it is unlikely that PEMV multiplication is linked in any way to chloroplasts and consequently PEMV must rely on an efficient UGA suppressor present in the cytoplasm. As shown in this report, a good candidate for this task could indeed be cyt tRNAU*CGArg.
Together with tRNAU*CGArg, we have now identified three natural UGA suppressor tRNA species in plants, which possibly exert their various actions in vivo depending on a suitable codon context.
This work was supported by a grant from the Deutsche Forschungsgemeinschaft to H.B.
Nucleic Acids Research
Pages
Introduction
Materials And Methods
Enzymes and reagents
Plasmids
Fractionation and isolation of tRNAsArg from wheat germ
Aminoacylation of tRNAs
Sequencing of tRNA by post-labelling techniques
Reverse transcription and PCR amplification
In vitro transcription with SP6 RNA polymerase
In vitro translation in wheat germ extract
Analysis of translation products
Results
Purification and characterization of cytoplasmic tRNAArg isoacceptors from wheat germ
Cytoplasmic tRNAArg with a U*CG anticodon promotes read-through over leaky UGA codons in different viral RNAs
Discussion
Acknowledgement
References
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 27 Feb 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Beier and M. Grimm Misreading of termination codons in eukaryotes by natural nonsense suppressor tRNAs Nucleic Acids Res., December 1, 2001; 29(23): 4767 - 4782. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Bertram, S. Innes, O. Minella, J. P. Richardson, and I. Stansfield Endless possibilities: translation termination and stop codon recognition Microbiology, February 1, 2001; 147(2): 255 - 269. [Full Text]
|
|
|
|
 |
|
 Z. Wang, A. Gaba, and M. S. Sachs A Highly Conserved Mechanism of Regulated Ribosome Stalling Mediated by Fungal Arginine Attenuator Peptides That Appears Independent of the Charging Status of Arginyl-tRNAs J. Biol. Chem., December 31, 1999; 274(53): 37565 - 37574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. E. Glover, D. F. Spencer, and M. W. Gray Identification and Structural Characterization of Nucleus-encoded Transfer RNAs Imported into Wheat Mitochondria J. Biol. Chem., January 5, 2001; 276(1): 639 - 648. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||