Nucleic Acids Research, 2002, Vol. 30, No. 3 759-765
© 2002 Oxford University Press
Imbalance of tRNAPro isoacceptors induces +1 frameshifting at near-cognate codons
J. W. Wilson Laboratory, Department of Molecular and Cellular Biology and Biochemistry, 69 Brown Street, Brown University, Providence, RI 02912, USA
Received September 13, 2001; Revised and Accepted November 29, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Increased expression of the CCU/CCA/CCG-decoding tRNAPro3 on a multicopy plasmid leads to suppression of several +1 frameshift mutations in Salmonella enterica serovar Typhimurium. Systematic analysis of the site of frameshifting indicates that excess tRNAPro3 promotes near-cognate decoding at CCC codons. Re-phasing of the reading frame can be achieved by a subsequent slippage of the tRNA onto a cognate codon in the +1 reading frame. Frameshifting appears to be due to an imbalance of CCC-cognate and near-cognate tRNAs, as the effect of excess tRNAPro3 on reading frame maintenance can be reversed by increasing simultaneously the concentration of the cognate tRNAPro2. Finally, the cmo5U modification present at position 34 of tRNAPro3, which allows this tRNA to decode CCU in addition to CCG and CCA, also affects frameshifting, indicating that the ability of the near-cognate tRNA to decode a cognate codon efficiently in the alternative reading frame is important for re-phasing of the reading frame.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
tRNA levels in bacteria are tightly controlled and subject to growth rate regulation. In Escherichia coli, the relative abundance of the differing isoacceptor tRNAs correlates well with the usage of their cognate codons at varying growth rates (1,2). This ensures that, during rapid growth, an adequate supply of charged tRNAs is available for translation of abundant transcripts that typically are rich in common codons. Thus, tRNA levels can have profound effects on gene expression, and high-level expression of heterologous genes in E.coli may require adjustment of tRNA gene dosage or alteration of the codon composition of the heterologous message (3,4). tRNA levels also influence the accuracy of decoding; tRNA limitation causes frameshifting in vivo (5) while over-abundance of particular tRNAs leads to miscoding and frameshift errors in vivo and in vitro (6–8). Previous work from this laboratory indicated that increased expression of the GGG-decoding tRNA1Gly led to –1 frameshifting at the near-cognate GGA codon (7). In this paper, the effects of increased gene dosage of each of the tRNAPro isoacceptors on +1 frameshifting at CCCN quadruplets has been investigated. Increased levels of
, which decodes CCA, CCG and CCU, led to frameshifting at CCCA, CCCG and CCCU quadruplets, while increased levels of
, which decodes CCG, induced frameshifting at the CCCG quadruplet only. In contrast, increased levels of the CCC/U-decoding
decreased frameshifting at all CCCN quadruplets. These data, together with previous results, indicate that tRNA abundance influences both +1 and –1 frameshifting and that the balance of cognate and near-cognate tRNAs is critical for accurate decoding.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bacterial strains and plasmids
Salmonella strains TR767 (hisD3018), TR3243 (hisD6610), TR2706 (hisO1242 C6581 tyr545 sufA6 sufG70) and TR1034 (hisO1242 D3749) were originally obtained from the laboratory of Dr John Roth, University of Utah, Salt Lake City (9–11). Isogenic aroD::Tn10 derivatives of these strains were constructed by transducing each hisD mutant to tetracycline resistance with phage P22 grown on TT1454 (aroD::Tn10). The E.coli strain MC150 F–
(lac-pro) thi-1 (trpE-C)8 trpE91 recA1 srl– was used as a host for tRNA and lacZ plasmids. Suppression of hisD mutants was tested on minimal E medium plates, containing glucose at 30°C (12). Aro– strains were assayed on minimal medium containing tryptophan, tyrosine and phenylalanine, p-aminobenzoic acid, p-hydroxybenzoic acid and 2,3-dihydroxybenzoic acid (13). Plasmid pAC1203 carrying the Salmonella hisR operon has been described previously (14). pBR1203 was constructed by insertion of the 972 bp EcoRI fragment from pAC1203 into the EcoRI site of pBR322. pBK1203 was derived from pBR1203 by insertion of a kanamycin resistance cassette from pUC4K (Amersham) into the PstI site of pBR1203. Each of the wild-type proline tRNA genes was amplified from the E.coli chromosome by PCR and cloned into the high copy number plasmid pTZ18u and also into the low copy number plasmid pHSG575 (15). Plasmids were verified by DNA sequencing with Sequenase (USB Biochemicals) according to the manufacturers’ instructions. LacZ plasmids containing desired frameshift mutations were constructed by ligating pairs of complementary oligonucleotides carrying ApaI and HindIII overhangs into the wild-type LacZ gene of pSG25 cleaved with ApaI and HindIII (7)
ß-Galactosidase assays
Cells to be assayed for ß-galactosidase activity were grown in minimal E medium (12) containing glucose (0.2%), thiamine, proline, tryptophan and casamino acids (0.2%) together with any necessary antibiotics. ß-Galactosidase was assayed as described previously (7,16).
Isolation of leaky His– mutants
Strain MS34 (hisO1242 D6610 zec2::Tn10 trpE91 tufA8 tufB103) is phenotypically Trp+ due to suppression of the trpE91 frameshift by mutant elongation factor EF-Tu but His–, because the hisD6610 frameshift is not suppressed (17). Upon prolonged incubation at 30°C, rare His+ revertants arose. Many of these had the growth characteristic of wild-type hisD+ strains and were not retained. Two mutants, designated hisD6610m2 and hisD6610m5, were slow-growing on histidine-free medium and were studied further. Phage P22 was grown on each of these mutants and used to transduce either a tufA+ tufB+ trpE91 strain or a tufA8 tufB103 trpE91 strain to tetracycline resistance. The tetracycline resistant transductants were examined for co-inheritance of the linked His phenotype. Surprisingly, the hisD6610m2 mutation was leaky irrespective of the presence of tufA/tufB mutations, whereas the His+ phenotype of the hisD6610m5 mutation was dependent on tufA/tufB mutations. Subsequent sequence analysis showed that the hisD6610m5 mutation had a four base deletion following the in-frame UGA codon, thereby converting a +1 frameshift into a EF-Tu-suppressible UGA mutation. The sequence of the hisD6610m2 mutation is presented in Figure 4.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Increased levels of tRNA cause suppression of +1 frameshift mutations
Genetic analysis of reading frame maintenance has typically been approached by the isolation and characterization of secondary (frameshift suppressor) mutations that suppress base pair insertion or deletion mutations in amino acid biosynthetic genes (18). More recently, in an effort to create specialized tRNAs for incorporation of non-natural amino acids into proteins, Schultz and co-workers (19) have constructed an array of tRNAs containing insertions in the anticodon loop that decode four base codons efficiently, thus expanding enormously the number of examples of quadruplet-decoding tRNAs. A range of well-characterized frameshift mutations that has been used for the isolation of frameshift suppressors exists in the histidine biosynthetic genes of Salmonella enterica serovar Typhimurium. In the Salmonella hisD3749 +1 frameshift mutation, the in-frame UGA stop codon generated by the insertion of a C residue is preceded by a CCC proline codon (Table 1). In a preliminary experiment to examine the effects of near-cognate tRNA levels on frameshifting, this hisD mutant was transformed with pRB322 or with a derivative, pBR1203, carrying the hisR operon encoding tRNAHis,
,
and
(20). When the transformants were streaked on minimal medium lacking histidine at 30°C to test for suppression of the hisD3749 frameshift, strains carrying pBR1203 grew within 72 h, while strains carrying the plasmid vector only showed no growth, even after prolonged incubation. Similar results were obtained with pAC1203, a pACYC184-derived plasmid carrying the same tRNA insert. However, in this case, suppression was weaker, with pAC1203 transformants taking up to 4 days to grow on minimal medium. These data indicate that increased expression of one of the tRNAs encoded by the hisR operon leads to frameshifting within the hisD3749 frameshift window. Moreover, suppression appeared to be gene dosage-dependent, as weaker suppression was observed with the lower copy number plasmid pAC1203 than with the pBR322-derived pBR1203.
|
In further suppression tests, the related +1 frameshift mutations hisD6610 and hisD3018, as well as hisC6581, were similarly shown to be suppressed by multicopy expression of the hisR tRNA operon. However, hisD3018 was suppressed far more efficiently than hisD6610, even though these frameshifts are very similar in primary sequence (Table 1). Comparison of the sequences of hisD3018, hisD6610 and hisD3749 (Table 1) showed that, in all cases, the in-frame UGA codons generated by the base insertion mutations were preceded by CCC proline codons. The hisC6581 mutant contains two CCCG and one CCCA quadruplet in its frameshift window.
, the proline tRNA encoded on the hisR operon, decodes CCA, CCG and CCU proline codons but not the CCC codon. These findings raised the possibility that increased expression of the near-cognate tRNA induced frameshifting at the CCC codon.
Increased expression of tRNAPro induces frameshifting at near-cognate codons
In S.enterica, there are three proline isoacceptor tRNAs.
decodes CCG only,
decodes both CCC and CCU while
, which contains the V base modification (cmo5U) at the wobble position (position 34) of the anticodon, decodes CCG, CCA and CCU (21–23; Fig. 1). Escherichia coli contains the same complement of tRNAPro genes, and these are identical in primary sequence to those found in Salmonella (23). The genes for each of these tRNAs (proK, proL and proM) were amplified by PCR and cloned into the high copy number plasmid, pTZ18u. Each of these plasmids was then introduced into hisD6610-, hisD3018- and hisD3749-containing strains and suppression was assayed on minimal medium. Growth on histidine-free minimal medium was observed only in strains transformed with the proM-containing plasmid expressing
, indicating that over-expression of this tRNA alone was responsible for suppression of the hisD frameshift mutants. Moreover, when
was expressed from the low copy number pSC101-derived plasmid pHSG575, suppression of hisD3749 was still evident but considerably weaker than observed with the pTZ18u-derived plasmid, confirming the influence of tRNA gene dosage on frameshifting.
|
The influence of over-expression of each proline tRNA on frameshifting at CCCN quadruplets was examined systematically using a series of lacZ constructs containing each of the target CCCN and NCCC frameshift sequences (Table 2). Substantial levels of ß-galactosidase activity were observed with most lacZ constructs in the absence of any manipulations of tRNA levels. The data presented in Table 2 showed that increased gene dosage of
enhanced frameshifting at CCCA, CCCU and CCCG, but not at CCCC quadruplets. Similarly, increased expression of proK encoding
induced frameshifting at CCCG quadruplets only. Frameshifting was not dependent on the presence of an adjacent stop codon and the highest level of frameshifting was seen at the CCC ACA sequence. Neither
nor
can decode the in-frame CCC codon, but where frameshifting is observed, each tRNA has the capacity to decode the overlapping CCN codon in the +1 reading frame. A model to explain frameshifting caused by near-cognate decoding (24,25) that is consistent with the observations made here, suggests that frameshifting occurs first by binding of a near-cognate tRNA to the ribosomal A site and subsequent slippage onto the overlapping, cognate codon in the P site. When the CCC codon was altered to ACC or GCC, the basal level of frameshifting decreased considerably and introduction of any of the tRNA plasmids failed to stimulate frameshifting. This last result also rendered unlikely the possibility that frameshifting was due to out-of-frame binding of proline tRNAs to the CCU codon in +1 frame. In contrast to the results observed with proK and proM plasmids, increased gene dosage of proL encoding
did not stimulate frameshifting at any CCCN sequence and in fact decreased the basal level of frameshifting at all CCCN quadruplets. This suggested that the basal level of frameshifting observed in the absence of any excess tRNA genes was due to ribosome slippage at the CCCN sequences and could be reversed by increasing the levels of the cognate, CCC-decoding
.
|
Frameshifting is caused by tRNA imbalance
The CCC codon is normally decoded only by the proL-encoded
. Increased gene dosage of the CCU/CCA/CCG-decoding
is predicted to lead to an imbalance of CCC-cognate and near-cognate tRNAs. To address the question of whether frameshifting was caused merely by increased levels of the near-cognate tRNAs per se or whether an imbalance of near-cognate and cognate tRNAs was necessary for frameshifting, hisD3018 and hisD3749 strains were constructed that (i) over-expressed the near-cognate tRNA alone or (ii) over-expressed the cognate and near-cognate tRNAs simultaneously, and were then assayed for their continued capacity to frameshift at CCCU codons. Strains containing these hisD mutations were first transformed with the proM-containing plasmid, pAC1203. The plasmid-containing strains were co-transformed with each of the pHSG575-derived plasmids, pProK4, pProL5 and pProM11, expressing
,
and
, respectively, as well as with the vector, pHSG575. The double transformants were purified and tested for suppression of the hisD frameshifts by their ability to grow on histidine-free medium. Growth patterns on minimal medium indicated that the presence of the vector plasmid pHSG575 or plasmids pProK4 or pProM11 did not affect multicopy proM-mediated frameshifting at the hisD3018 or hisD3749 frameshift sites. However, the presence of plasmid pProL5, expressing the CCC-cognate
, decreased frameshift suppression of both hisD mutants dramatically (Fig. 2). These results indicate that the effects of increased levels of
on frameshifting at near-cognate codons can be reversed by simultaneous over-expression of the cognate
, and it is concluded that tRNA imbalance is necessary to elicit frameshifting.
|
tRNA modification influences near-cognate decoding
The extended decoding capacity of
is due to the cmo5U modification at position 34 in the anticodon. In E.coli and S.enterica, this same modification is found in tRNAs decoding, alanine, valine, serine, proline and threonine codons (26,27). Genetic analysis has shown that the cmo5U34 modification is derived from the aromatic amino acid biosynthetic pathway (13,28). Thus, aroA, B, C, D and E mutants lack this modification and, consequently, the decoding abilities of the (unmodified)
are expected to be restricted to CCA and CCG in such Aro– strains. The analysis of frameshifting at CCCN quadruplets presented above strongly suggested that frameshifting occurred when the near-cognate tRNA slipped forward onto the overlapping, cognate CCA, CCG or CCU codon in the +1 reading frame. Lack of the cmo5U34 modification is thus predicted to inhibit frameshifting at CCCU sequences. This prediction was tested by constructing aroD– derivatives of the hisD3018 and hisD3749 strains and monitoring suppression of the hisD frameshift mutations in the wild-type and Aro– strains. Isogenic pairs of Aro+ and aroD– hisD3018 and hisD3749 strains were transformed with the proM-containing plasmid pBK1203 and suppression of the hisD frameshifts was monitored on minimal, histidine-free medium. Growth patterns on this medium showed that the aroD mutation decreased hisD3749 suppression substantially (Fig. 3), whereas suppression of hisD3108 (which is very efficiently suppressed by a multicopy proM-containing plasmid) was only slightly affected (not shown). These experiments showed that tRNA-mediated frameshifting at the CCCU quadruplet can be affected by the ability of the
to decode the CCU codon in the +1 reading frame. However, the continued suppression of the hisD3018 frameshift in the absence of the cmo5U modification indicates that even the unmodified
can, in a favorable sequence context, decode the CCU codon in the +1 reading frame at an appreciable frequency.
|
Spontaneous frameshifting at CCCU quadruplets
The hisD mutants examined here that are suppressed by increased proM expression are non-leaky and do not exhibit any detectable frameshifting in wild-type cells. External suppressor mutations in both tRNA and EF-Tu genes have been isolated that promote frameshifting at each of these hisD frameshift sites (17,18). Curiously, while mutant EF-Tu causes frameshifting of the hisD3018 frameshift mutant, it fails to suppress the closely related frameshift hisD6610 (Table 1; 17). In the course of an attempt to obtain derivatives of hisD6610 that were suppressed by mutant EF-Tu, His+ revertants of a hisD6610 tufA8 tufB103 strain were isolated. Among the potentially EF-Tu-suppressible hisD6610 derivatives that grew slowly on minimal medium, one mutant, hisD6610m2, was shown to be leaky even in the absence of any mutant translational components, suggesting that the nucleotide alteration(s) in this mutant stimulated frameshifting by wild-type ribosomes. Sequencing of the hisD6610m2 mutant indicated that it differed from the parental hisD6610 sequence only by the deletion of 3 bp 5' to the CCC CCC UGA sequence (where UGA is the in-frame stop codon; Fig. 4). When this hisD mutant was transformed with plasmid pProL5 encoding
, the leaky phenotype was abolished, indicating that frameshifting was occurring at the CCC UGA sequence. Several other instances of frameshifting at CCC UGA have been described in the literature; expression of the human catechol O-methyltransferase in E.coli leads to the synthesis of two proteins, one of which contains a C-terminal extension caused by frameshifting at CCC UGA (29). Expression in E.coli of a construct containing the human transferrin gene placed downstream of a fragment derived from the replicase gene of phage MS2 leads to the synthesis of a fusion protein produced by frameshifting at CCC UGA, in addition to transferrin protein alone (30). Compared with the original hisD6610 mutant, the nucleotide alteration in hisD6610m2 has the effect of changing the E site codon from threonine (ACA) to valine (GUA) during near-cognate decoding of CCC by
in the A site. Numerous studies have reported effects of 5' and 3' codons and tRNAs on nonsense and frameshift suppression (31–33). A more recent study has identified the effects of the amino acid encoded by the E site codon (the penultimate amino acid) on readthrough of UGA codons in the A site (34). In particular, valine as the penultimate amino acid promoted a 3-fold higher rate of UGA readthrough than threonine. Both UGA readthrough and the frameshifting event studied here are brought about by near-cognate decoding, raising the possibility that the effect of the penultimate amino acid on UGA readthrough (34) may also influence frameshifting by near-cognate tRNAs. In keeping with this proposal, examination of the two published examples of frameshifting at CCC UGA cited above reveals the presence of alanine and glycine triplets two codons upstream of the shifty CCC codon. These penultimate amino acids had also been found to promote high levels of UGA readthrough (31). However, other explanations for the leakiness of the hisD6610m2 frameshift are possible: the identity of the E site tRNA may also influence A site decoding, as it is believed that both A and E sites are coupled allosterically (35). In addition, alteration of the amino acid sequence in the hisD6610m2 mutant may influence the residual activity of the histidinol dehydrogenase enzyme.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The availability of cognate tRNA is essential for faithful decoding of the genetic code. Lack of cognate tRNA can clearly have a serious impact on the level of misincorporation and nonsense errors (36). However, evidence from several sources has accumulated to suggest that a deficiency of the appropriate tRNA species can also influence maintenance of the translational reading frame. tRNA pools are easily depleted by amino acid starvation or by addition of analogs and inhibitors of aminoacyl tRNA synthetases. Over-production of proteins rich in rare codons can also lead to a depletion of the cognate, rare tRNA species. Both of these treatments have been shown to cause frameshifting in E.coli at levels approaching 50% (5,37). tRNA imbalance also can be achieved by over-expression of particular tRNAs; manipulation of tRNA levels by these means has been shown to effect nonsense suppression in yeast (8) and missense suppression and –1 frameshifting in E.coli (7). All of these latter events ensue from the binding of a near-cognate tRNA to the ribosomal A site. Decoding during missense incorporation of this sort involves only two codon–anticodon pairs with continuation of translation in frame. Frameshifting promoted by near-cognate tRNAs requires in addition that the reading frame be re-phased. A model to explain the involvement of near-cognate decoding in frameshifting has been proposed by Bjork, Farabaugh and co-workers (24,25). According to this ‘dual error’ model, such mismatched codon–anticodon complexes may be particularly prone to slippage into alternate reading frames and the presence of a cognate codon in the alternate reading frame may stabilize binding of the shifted tRNA as it involves three codon–anticodon pairs. Frameshifting errors of this sort are thus proposed to derive from two sequential events. Cognate tRNAs do not display the same propensity to frameshift as, in general, their binding to the ribosome involves three codon–anticodon pairs. Conditions that inhibit near-cognate decoding or interfere with binding of the shifted tRNA in the alternate reading frame are expected to decrease frameshifting. The influence of near-cognate proline tRNAs on frameshifting described here is fully consistent with the dual error model. Increases in the level of the cognate
out-compete the near-cognate tRNAs for A site binding and decrease frameshifting, while the absence of a cognate codon in the +1 reading frame or the lack of cmo5U modification, which is expected to decrease CCU-decoding by tRNA, was also found to decrease frameshifting.
The dual error model was originally invoked to explain the decoding properties of sufA and sufB mutant proline tRNAs (derived from
and
, respectively) that permitted quadruplet decoding (24). These mutant tRNAs contained insertions in their anticodons and it had been assumed that a simple expansion of the anticodon loop permitted decoding of a correspondingly expanded, complementary codon. However, analysis of the base modification content indicated that the mutant tRNAs were incapable of engaging in a four base codon–anticodon interaction. Moreover, the dependence of suppression on the Aro+ status of the cell suggested that suppression at CCCN codons was effected instead by the near-cognate
. Thus, in the case of the sufB mutants, a situation akin to tRNA imbalance is created through debilitation of the cognate tRNAs by base insertions in the anticodon loop. This permits binding of the near-cognate tRNA to the ribosome and a subsequent slippage of the tRNA re-phases the reading frame. The dual error model was subsequently proposed to explain frameshifting by most, if not all, quadruplet-decoding suppressor tRNAs (25). This, however, appears to be an overstatement: the ability of some tRNAs with expanded anticodons to decode complementary, quadruplet anticodons is perhaps best illustrated by the many suppressor tRNAs derived from tRNASer constructed by Schultz and co-workers (19). The most efficient of these tRNAs invariably have perfect Watson–Crick complementarity to the suppressed quadruplet anticodon. Moreover, suppression at one particular site in the ß-lactamase reporter gene requires the insertion of serine, indicating the direct action of the mutant tRNA. Thus, while the dual error model neatly explains suppression by some mutant tRNAs, its applicability to all instances of frameshift suppression by mutant tRNAs needs to approached with caution.
In the experiments described here, increased multicopy expression of near-cognate tRNAs is necessary for frameshifting. The dependence of frameshift suppression on increased tRNA gene dosage and its reversal by increased cognate tRNA levels indicates that frameshifting derives from a competition between cognate and near-cognate tRNAs. Increased near-cognate
levels presumably allow it to compete effectively with the cognate
for A site binding. The existence of sequence contexts that apparently promote frameshifting by near-cognate tRNAs in the absence of manipulation of tRNA gene dosage indicates that such tRNA competition can occur in completely wild-type cells and be influenced by the surrounding mRNA and possibly the amino acid sequence contexts.
While the frameshifted protein products have not been characterized here, data in the literature indicates that the CCCU sequence can be decoded as a single proline by wild-type tRNAs. When human catechol O-methyltransferase was expressed in E.coli, two proteins with enzymic activity were produced: the smaller had the expected molecular weight of the native protein while the larger protein appeared to be a C-terminal extension product. Characterization of the larger protein by mass spectroscopy indicated that it derived from frameshifting at the GCA GGG CCC UGA AGA sequence where a single proline was inserted at the CCCU quadruplet, and translation continued in the +1 reading frame.
While in E.coli there are several tRNAs available to decode the four codons of a codon ‘family box’ (such as the proline codons), in mycoplasmas and mitochondria typically there is a single tRNA that can decode all four codons of a codon family (38). In Mycoplasma capricolum, a single tRNAPro with an unmodified U34 decodes all four proline codons, presumably by reading just the first two letters of the triplet (39). The structure of the M.capricolum tRNAPro anticodon differs from that of E.coli
only in the absence of the cmo5U34 and Um32 modifications. The presence of a C residue at position 32 in M.capricolum tRNAGly enables that tRNA to decode all four glycine codons without discrimination (40). The features of the M.capricolum tRNAPro that allow it to decode all four proline codons have not been analyzed. As described here for proline codons, and previously for glycine codons, such two-out-of-three reading of a codon by a near-cognate tRNA in E.coli is associated with shifts in the reading frame. However, two-out-of-three reading appears to be common in mycoplasmas and mitochondria and it remains unclear how these organisms avoid shifts in the reading frame.
|
ACKNOWLEDGEMENTS |
|---|
I am indebted to Dr Albert Dahlberg for providing the support to carry out these experiments. Thanks are due to Drs John Roth and John Atkins for supplying numerous bacterial strains and Hugo Strait for erudite legalese. This work was supported by grant GMS19756 to Albert E. Dahlberg from the National Institutes of Health.
|
FOOTNOTES |
|---|
* Tel: +1 401 863 3652; Fax: +1 401 863 1182; Email: michael_o'connor{at}brown.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-
1 Berg,O.G. and Kurland,C.G. (1997) Growth rate-optimised tRNA abundance and codon usage. J. Mol. Biol., 270, 544–550.[Web of Science][Medline]
2 Dong,H., Nilsson,L. and Kurland,C.G. (1996) Co-variation of tRNA abundance and codon usage in Escherichia coli at different growth rates. J. Mol. Biol., 260, 649–663.[Web of Science][Medline]
3 Kane,J.F. (1995) Effects of rare codon clusters on high-level expression of heterologous proteins in Escherichia coli. Curr. Opin. Biotechnol., 6, 494–500.[Web of Science][Medline]
4 Del Tito,B.J., Ward,J.M., Hodgson,J., Gershater,C.J., Edwards,H., Wysocki,L.A., Watson,F.A., Sathe,G. and Kane,J.F. (1995) Effects of a minor isoleucyl tRNA on heterologous protein translation in Escherichia coli. J. Bacteriol., 177, 7086–7091.
5 Weiss,R.B. and Gallant,J.A. (1986) Frameshift suppression in aminoacyl-tRNA limited cells. Genetics, 112, 727–739.
6 Atkins,J.F., Gesteland,R.F., Reid,B.R. and Anderson,C.W. (1979) Normal tRNAs promote ribosomal frameshifting. Cell, 18, 1119–1131.[Web of Science][Medline]
7 O’Connor,M. (1998) tRNA imbalance promotes –1 frameshifting via near-cognate decoding. J. Mol. Biol., 279, 727–736.[Web of Science][Medline]
8 Navarro,F. and Thuriaux,P. (2000) In vivo misreading by tRNA overdose. RNA, 6, 103–110.[Abstract]
9 Bossi,L. and Roth,J.R. (1981) Four-base codons ACCA, ACCU and ACCC are recognized by frameshift suppressor sufJ. Cell, 25, 489–496.[Web of Science][Medline]
10 Riddle,D.L. and Roth,J.R. (1970) Suppressors of frameshift mutations in Salmonella typhimurium. J. Mol. Biol., 54, 131–144.[Web of Science][Medline]
11 Hartman,P.E., Ames,B.N., Roth,J.R., Barnes,W.M. and Levin,D.E. (1986) Target sequences for mutagenesis in Salmonella histidine requiring mutants. Environ. Mutagen., 8, 631–641.[Web of Science][Medline]
12 Vogel,H.J. and Bonner,D.M. (1956) Acetylornithinase of Escherichia coli: partial purification and some properties. J. Biol. Chem., 218, 97–106.
13 Bjork,G.R. (1980) A novel link between the biosynthesis of aromatic amino acids and transfer RNA modification in Escherichia coli. J. Mol. Biol., 140, 391–410.[Web of Science][Medline]
14 O’Connor,M., Willis,N.M., Bossi,L., Gesteland,R.F. and Atkins,J.F. (1993) Functional tRNAs with altered 3' ends. EMBO J., 12, 2559–2566.[Web of Science][Medline]
15 Takeshita,S., Sato,M., Toba,M., Masahashi,W. and Hashimoto-Gotoh,T. (1987) High-copy-number and low-copy-number plasmid vectors for lacZ 
-complementation and chloramphenicol or kanamycin resistance selection. Gene, 61, 63–74.[Web of Science][Medline]
16 Miller,J.H. (1991). A Short Course in Bacterial Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
17 Hughes,D., Atkins,J.F. and Thompson,S. (1987) Mutants of elongation factor Tu promote ribosomal frameshifting and nonsense readthrough. EMBO J., 6, 4235–4239.[Web of Science][Medline]
18 Atkins,J.F., Weiss,R.B., Thompson,S. and Gesteland,R.F. (1991) Towards a genetic dissection of the basis of triplet decoding and its natural subversion: programmed reading frame shifts and hops. Annu. Rev. Genet., 25, 201–228.[Web of Science][Medline]
19 Magliery,T.J., Anderson,J.C. and Schultz,P.G. (2001) Expanding the genetic code: selection of efficient four-base codons and identification of ‘shifty’ four-base codons with a library approach in Escherichia coli. J. Mol. Biol., 307, 755–769.[Web of Science][Medline]
20 Bossi,L. (1983) The hisR locus of Salmonella: nucleotide sequence and expression. Mol. Gen. Genet., 192, 163–170.[Web of Science][Medline]
21 Komine,Y., Adach,T., Inokuchi,H. and Ozeki,H. (1990) Genomic organization and physical mapping of the transfer RNA genes in Escherichia coli K12. J. Mol. Biol., 212, 579–598.[Web of Science][Medline]
22 Qiang,Q. and Bjork,G.R. (1997) Structural alterations far from the anticodon of the tRNAProGGG of Salmonella typhimurium induce +1 frameshifting at the peptidyl site. J. Mol. Biol., 273, 978–992.[Web of Science][Medline]
23 Kuchino,Y., Mori,F. and Nishimura,S. (1985) Structure and transcription of the tRNAPro1 gene from Escherichia coli. Nucleic Acids Res., 13, 3213–3220.
24 Qian,Q., Li, J.N., Zhao,H., Hagervall,T.G., Farabaugh,P.J. and Bjork,G.R. (1998) A new model for phenotypic suppression of frameshift mutations by mutant tRNAs. Mol. Cell, 1, 471–482.[Web of Science][Medline]
25 Farabaugh,P.J. and Bjork,G.R. (1999) How translational accuracy influences reading frame maintenance. EMBO J., 18, 1427–1434.[Web of Science][Medline]
26 Hagervall,T.G., Jonsson,Y.H., Edmonds,C.G., McCloskey,J.A. and Bjork,G.R. (1990) Chorismic acid, a key metabolite in modification of tRNA. J. Bacteriol., 172, 252–259.
27 Yokoyama,S., Watanabe,T., Murao,K., Ishikura,H., Yamaizumi,Z., Nishimura,S. and Miyazawa,T. (1985) Molecular mechanism of codon recognition by tRNA species with modified uridine in the first position of the anticodon. Proc. Natl Acad. Sci. USA, 82, 4905–4909.
28 Bjork,G.R. (1995) Genetic dissection of synthesis and function of modified nucleosides in bacterial transfer RNA. Prog. Nucleic Acid Res. Mol. Biol., 50, 263–338.[Web of Science][Medline]
29 Vilbois,F., Caspers,P., da Prada,M., Lang,G., Karrer,C., Lahm,H.W. and Cesura,A.M. (1994) Mass spectrometric analysis of human soluble catechol O-methyltransferase expressed in Escherichia coli. Identification of a product of ribosomal frameshifting and of reactive cysteines involved in S-adenosyl-L-methionine binding. Eur. J. Biochem., 222, 377–386.[Web of Science][Medline]
30 de Smit,M.H., van Duin,J., van Knippenberg,P.H. and van Eijk,H.G. (1994) CCC.UGA: a new site of ribosomal frameshifting in Escherichia coli. Gene, 143, 43–47.[Web of Science][Medline]
31 Bossi,L. (1983) Context effects: translation of UAG codon by suppressor tRNA is affected by the sequence following UAG in the message. J. Mol. Biol., 164, 73–87.[Web of Science][Medline]
32 Kopelowitz,J., Hampe,C., Goldman,R., Reches,M. and Engelberg-Kulka,H. (1992) Influence of codon context on UGA suppression and readthrough. J. Mol. Biol., 225, 261–269.[Web of Science][Medline]
33 Curran,J.F., Poole,E.S., Tate,W.P. and Gross,B.L. (1995) Selection of aminoacyl-tRNAs at sense codons: the size of the tRNA variable loop determines whether the immediate 3' nucleotide to the codon has a context effect. Nucleic Acids Res., 23, 4104–4108.
34 Mottagui-Tabar,S., Bjornsson,A. and Isaksson,L.A. (1994) The second to last amino acid in the nascent peptide as a codon context determinant. EMBO J., 13, 249–257.[Web of Science][Medline]
35 Rheinberger,H.J. and Nierhaus,K.H. (1986) Allosteric interactions between the ribosomal transfer RNA-binding sites A and E. J. Biol. Chem., 261, 9133–9139.
36 Calderone,T.L, Stevens,R.D. and Oas,T.G. (1996) High-level misincorporation of lysine for arginine at AGA codons in a fusion protein expressed in Escherichia coli. J. Mol. Biol. 262, 407–412.[Web of Science][Medline]
37 Spanjaard,R.A. and van Duin,J. (1988) Translation of the sequence AGG-AGG yields 50% ribosomal frameshift. Proc. Natl Acad. Sci. USA, 85, 7967–7971.
38 Lagerkvist,U. (1978) ‘Two out of three’: an alternative method for codon reading. Proc. Natl Acad. Sci. USA, 75, 1759–1762.
39 Andachi,Y., Yamao,F., Muto,A. and Osawa,S. (1989) Codon recognition patterns as deduced from sequences of the complete set of transfer RNA species in Mycoplasma capricolum. J. Mol. Biol., 209, 37–54.[Web of Science][Medline]
40 Claesson,C., Lustig,F., Bóren,T., Simonsson,C., Barciszewska,M. and Lagerkvist,U. (1995) Glycine codon discrimination and the nucleotide in position 32 of the anticodon loop. J. Mol. Biol., 247, 191–196.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. F. Atkins and G. R. Bjork A Gripping Tale of Ribosomal Frameshifting: Extragenic Suppressors of Frameshift Mutations Spotlight P-Site Realignment Microbiol. Mol. Biol. Rev., March 1, 2009; 73(1): 178 - 210. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gong, F. Gong, and C. Yanofsky Overexpression of tnaC of Escherichia coli Inhibits Growth by Depleting tRNA2Pro Availability. J. Bacteriol., March 1, 2006; 188(5): 1892 - 1898. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. L. Gurvich, P. V. Baranov, R. F. Gesteland, and J. F. Atkins Expression Levels Influence Ribosomal Frameshifting at the Tandem Rare Arginine Codons AGG_AGG and AGA_AGA in Escherichia coli J. Bacteriol., June 15, 2005; 187(12): 4023 - 4032. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. O'Connor, S. T. Gregory, and A. E. Dahlberg Multiple defects in translation associated with altered ribosomal protein L4 Nucleic Acids Res., October 27, 2004; 32(19): 5750 - 5756. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||