Genetic characterization of frameshift suppressors with new decoding properties

Suppressor mutants that cause ribosomes to shift reading frame at specific and new sequences are described. Suppressors for trpE91, the only known suppressible -1 frameshift mutant, have been isolated in Escherichia coli and in Salmonella typhimurium. E. coli hopR acts on trpE91 within the 9-base-pair sequence GGA GUG UGA, is dominant, and is located at min 52 on the chromosome. Its Salmonella homolog maps at an equivalent position and arises as a rarer class in that organism as compared with E. coli. The Salmonella suppressor, hopE, believed to be in a duplicate copy of the same gene, maps at min 17. The +1 suppressor, sufT, acts at the nonmonotonous sequence CCGU, is dominant, and maps at min 59 on the Salmonella chromosome.     

tRNA hopping: enhancement by an expanded anticodon.

At a low level wild-type tRNA(1Val) inserts a single amino acid (valine) for the five nucleotide sequence GUGUA which has overlapping valine codons. Mutants of tRNA(1Val) with an insertion of A or U between positions 34 and 35 of their anticodons have enhanced reading of the quintuplet sequences. We propose that this decoding occurs by a hopping mechanism rather than by quintuplet pairing. Such hopping involves disengagement of the paired codon and anticodon with the mRNA slipping two (or more) bases along the ribosomal--peptidyl tRNA complex and subsequently re-pairing at a second codon--the landing site. The mutant with the anticodon sequence 3'CAAU5' 'hops' over the stop codon in the mRNA sequence GUG UAA GUU with the insertion of a single amino acid (valine). In contrast, in reading the same sequence, the mutant with the anticodon 3'CAUU5' hops onto the stop with the insertion of two valine residues. It is likely that in some instances of hopping alternate anticodon bases are used for the initial pairing and at the landing site.     

Efficiencies of translation in three reading frames of unusual non-ORF sequences isolated from phage display.

An unusual nucleotide sequence, called H10, was previously isolated by biopanning with a random peptide library on filamentous phage. The sequence encoded a peptide that bound to the growth hormone binding protein. Despite the fact that the H10 sequence can be expressed in Escherichia coli as a fusion to the gene III minor coat protein of the M13 phage, the sequence contained two TGA stop codons in the zero frame. Several mutant derivatives of the H10 sequence carried not only a stop codon, but also showed frameshifts, either +1 or -1 in individual isolates, between the H10 start and the gene III sequences. In this work, we have subcloned the H10 sequence and three of its derivatives (one requiring a +1 reading frameshift for expression, one requiring a -1 reading frameshift, and one open reading frame) in gene fusions to a reporter beta-galactosidase gene. These sequences have been cloned in all three reading frames relative to the reporter. The non-open reading frame constructs gave (surprisingly) high expression of the reporter (10-40% of control vector expression levels) in two out of the three frames. A site-directed mutant of the TGA stop codon (to TTA) in the +1 shifter greatly reduced the frameshift and gave expression primarily in the zero frame. By contrast, a site-directed mutant of the TGA in the -1 shifter had little effect on the pattern of expression, and alteration of the first TGA (of two) in H10 itself paradoxically reduced expression by half. We believe these phenomena to reflect a translational recoding mechanism in which ribosomes switch reading frames or read past stop codons upon encountering a signal encoded in the nucleotide sequence of the mRNA, because both the open reading frame derivative (which has six nucleotide changes from parental H10) and the site-directed mutant of the +1 shifter, primarily expressed the reporter only in the zero frame.     

Frameshifting by eukaryotic ribosomes during expression of Escherichia coli release factor 2.

Normal translation of the gene for E. coli release factor 2 (RF-2) is characterized by a +1 frameshift event that occurs with 30-50% efficiency. Frameshifting on synthetic RF-2 mRNA by eukaryotic ribosomes has also been observed, even though they lack the capability to interact with the frameshift signal in the same manner as prokaryotic ribosomes. We have mutagenized the sequence of the RF-2 gene to eliminate the need for a frameshift, thereby allowing frameshifting efficiency to be measured by direct comparison of RF-2 production from the mutant with production from the wild-type. Measurements using this approach confirm that frameshifting by rabbit reticulocyte lysate ribosomes occurs at the frameshift region, but with a limited efficiency of approximately 0.4%.     

The function of a ribosomal frameshifting signal from human immunodeficiency virus-1 in Escherichia coli

A 15-17 nucleotide sequence from the gag-pol ribosome frameshift site of HIV-1 directs analogous ribosomal frameshifting in Escherichia coli. Limitation for leucine, which is encoded precisely at the frameshift site, dramatically increased the frequency of leftward frameshifting. Limitation for phenylaianine or arginine, which are encoded just before and just after the frameshift, did not significantly affect frameshifting. Protein sequence analysis demonstrated the occurrence of two closeiy related frameshift mechanisms. In the first, ribosomes appear to bind leucyl-tRNA at the frameshift site and then slip leftward. This is the 'simultaneous slippage’mechanism. In the second, ribosomes appear to slip before binding amlnoacyl-tRNA, and then bind phenylaianyl-tRNA, which is encoded in the left-shifted reading frame. This mechanism is identicai to the‘overlapping reading’we have demonstrated at other bacterial frameshift sites. The HIV-1 sequence is prone to frame-shifting by both mechanisms in E. coli.     

Ribosomal Pausing at a Frameshifter RNA Pseudoknot Is Sensitive to Reading Phase but Shows Little Correlation with Frameshift Efficiency

Here we investigated ribosomal pausing at sites of programmed -1 ribosomal frameshifting, using translational elongation and ribosome heelprint assays. The site of pausing at the frameshift signal of infectious bronchitis virus (IBV) was determined and was consistent with an RNA pseudoknot-induced pause that placed the ribosomal P- and A-sites over the slippery sequence. Similarly, pausing at the simian retrovirus 1 gag/pol signal, which contains a different kind of frameshifter pseudoknot, also placed the ribosome over the slippery sequence, supporting a role for pausing in frameshifting. However, a simple correlation between pausing and frameshifting was lacking. Firstly, a stem-loop structure closely related to the IBV pseudoknot, although unable to stimulate efficient frameshifting, paused ribosomes to a similar extent and at the same place on the mRNA as a parental pseudoknot. Secondly, an identical pausing pattern was induced by two pseudoknots differing only by a single loop 2 nucleotide yet with different functionalities in frameshifting. The final observation arose from an assessment of the impact of reading phase on pausing. Given that ribosomes advance in triplet fashion, we tested whether the reading frame in which ribosomes encounter an RNA structure (the reading phase) would influence pausing. We found that the reading phase did influence pausing but unexpectedly, the mRNA with the pseudoknot in the phase which gave the least pausing was found to promote frameshifting more efficiently than the other variants. Overall, these experiments support the view that pausing alone is insufficient to mediate frameshifting and additional events are required. The phase dependence of pausing may be indicative of an activity in the ribosome that requires an optimal contact with mRNA secondary structures for efficient unwinding.     

E. coli ribosomes re-phase on retroviral frameshift signals at rates ranging from 2 to 50 percent

Many retroviruses express gag-pol or gag-pro-pol polypeptides by coupling their translation from overlapping reading frames with -1 ribosomal frameshifts. Here, we show that the well-known ribosomal frameshift signals found in retroviral mRNA will provoke Escherichia coli ribosomes to shift frame in the same manner as their eukaryotic counterparts. Ribosomes of E. coli respond in vivo to both the tandem slippery codons present at the retroviral frameshift site and the 3' flanking sequence. Slight alteration of the mouse mammary tumor virus gag-pro frameshift site from A-AAA-AAC to A-AAA-AAG boosts the level of frameshifting in E. coli to over 50%. This suggests that A-AAA-AAG, and its slippery relatives, may be utilized by E. coli genes as sites of high-level ribosomal frameshifting. This observed conservation of response to retroviral frameshift signals affords new avenues to dissect the mechanism of ribosomal frameshifting evoked by these mRNA sequences.     

Suppression of a nuclear frameshift mutation by a mitochondrial tRNA in the yeast Kluyveromyces lactis

A fragment of mitochondrial DNA containing the Kluyveromyces lactis gene for valine-tRNA (tRNAVAL) was isolated as a multicopy suppressor of a respiratory-deficient mutant of this yeast. The mutant produced a truncated Cox14p because of a +1 frameshift mutation in COX14, a nuclear gene encoding a protein imported into mitochondria which is necessary for respiration (Fiori et al. 2000 Yeast 16: 307-314). We report here that the mitochondrial tRNAVAL gene, when transformed into K. lactis cells, is transcribed outside mitochondria and suppresses the frameshift mutation in COX14 restoring the correct reading frame during translation of its mRNA. In fact, using histidine tagging, the existence of a suppressed Cox14p of normal length was demonstrated in mutants expressing the mt-tRNAVAL from the nucleus. Suppression could occur through a non-canonical four base pairing between the tRNAVAL and the mutated mRNA or through slippage of ribosomes during translation. This is a new case of informational suppression in that the suppression of a chromosomal mutation is not caused by a second mutation but to a mislocalization/expression of a mt-tRNA.     

Recognition of translational termination signals

Ribosomes can specifically shift at certain codons so that the mRNA is read in two different reading frames. To determine if frameshifting occurs at the level of termination, polymers of defined sequence containing AUG, a coding sequence and an in- or out-of-phase nonsense codon were used to bind a termination substrate or to program synthesis and release of dipeptides in a highly purified in vitro translation system. fMet-tRNA bound to ribosomes with AUGUAA, AUGUAAn, AUGUUU, AUGUUA or AUGUAn was not a substrate for release factor RF-1. In contrast, AUGU1UAA, AUGU3UAAn, AUGU4UAAn, AUGU5UAAn effected RF-1-dependent release of fMet from ribosomes. This suggests that nonsense codons can stimulate release whether they occur in- or out-of-phase. Addition of exogenous UAA and RF-1 stimulated release with all oligonucleotides tested. Propagation restricted the RF-1-dependent recognition of out-of-phase nonsense codons but did not restrict recognition of in-phase UAA in AUGU3UAAn. Release of dipeptides from ribosomes programmed with AUGU4UAAn occurred without EF-G and with a mutant lacking EF-G activity, suggesting that out-of-phase termination can occur prior to translocation outside the ribosomal A-site. We propose that the ribosome X RF complex is required to complete proteins, but is also able to frameshift at a nonsense codon resulting in occasional out-of-phase termination of protein synthesis.     

Translation of the F protein of hepatitis C virus is initiated at a non-AUG codon in a +1 reading frame relative to the polyprotein

The hepatitis C virus (HCV) genome contains an internal ribosome entry site (IRES) followed by a large open reading frame coding for a polyprotein that is cleaved into 10 proteins. An additional HCV protein, the F protein, was recently suggested to result from a +1 frameshift by a minority of ribosomes that initiated translation at the HCV AUG initiator codon of the polyprotein. In the present study, we reassessed the mechanism accounting for the synthesis of the F protein by measuring the expression in cultured cells of a luciferase reporter gene with an insertion encompassing the IRES plus the beginning of the HCV-coding region preceding the luciferase-coding sequence. The insertion was such that luciferase expression was either in the +1 reading frame relative to the HCV AUG initiator codon, mimicking the expression of the F protein, or in-frame with this AUG, mimicking the expression of the polyprotein. Introduction of a stop codon at various positions in-frame with the AUG initiator codon and substitution of this AUG with UAC inhibited luciferase expression in the 0 reading frame but not in the +1 reading frame, ruling out that the synthesis of the F protein results from a +1 frameshift. Introduction of a stop codon at various positions in the +1 reading frame identified the codon overlapping codon 26 of the polyprotein in the +1 reading frame as the translation start site for the F protein. This codon 26(+1) is either GUG or GCG in the viral variants. Expression of the F protein strongly increased when codon 26(+1) was replaced with AUG, or when its context was mutated into an optimal Kozak context, but was severely decreased in the presence of low concentrations of edeine. These observations are consistent with a Met-tRNA_i-dependent initiation of translation at a non-AUG codon for the synthesis of the F protein.     

Positions +5 and +6 can be major determinants of the efficiency of non-AUG initiation codons for protein synthesis.

Only rarely do GUG (or CUG or ACG) codons which precede the 5'-proximal AUG function as initiators of protein synthesis, even when they are within a context that contains a purine at position -3 and a G at +4. For example, the upstream GUG of the human parainfluenza virus type 1 (hPIV1) P gene is initiated by ribosomes at high frequency, whereas a seemingly similar GUG codon in the hPIV3 P gene is not used at all. We have examined the reasons for this by expressing chimeric hPIV3/hPIV1 mRNAs, both in vivo and in vitro. A major determinant for efficient GUG utilization was located downstream of the GUG, but this did not appear to be involved in the formation of secondary structure. Rather, the sequence immediately downstream was found to be critical; this determinant was mapped to positions +5 and +6. GUG could be used efficiently for ribosomal initiation when the second codon was GAU but not when it was GUA. Similar results were found when other non-AUG start sites, the Sendai virus P gene ACG and the c-myc-1 CUG, were examined. These results suggest that positions +5 and +6 are important determinants for initiation at non-AUG start sites, and that they are recognized independently of the overall secondary structure of the mRNA.     

The nucleotide sequence of the first externally suppressible--1 frameshift mutant, and of some nearby leaky frameshift mutants

Nine mutants within a 23 nucleotide sequence of the trpE gene of Salmonella typhimurium have been characterized. trpE91, a mutant which is externally suppressible has a single base deletion. Eight (or nine) nucleotides upstream of this deletion, two independently isolated mutations have the same transversion. In combination with trpE91 these mutations lead to partial restoration of synthesis of anthranilate synthetase in the absence of external suppressors. In the transversion the sequence A CA is changed to A AA and this new sequence may be the site where frameshifting occurs to allow leakiness. Leakiness is displayed by two further mutants of the same sign as trpE91, and one of the opposite sign, in the absence of any base substitution or external suppressors. Specific sequences, e.g., UUUC, may be especially prone to frameshifting and this sequence is created at the site of the +1 frameshift mutant which displays leakiness. In the new reading frame generated by the two -1 frame leaky mutants, a tryptophan codon is encountered. Leakiness is necessarily detected in the absence of tryptophan and under these conditions there will be a shortage of charged tryptophan tRNA. The possibility of such functional imbalance leading to frameshifting in these mutants is discussed.     

Quantitative analysis of in vivo ribosomal events at UGA and UAG stop codons.

An in vivo translation assay system has been designed to measure, in one and the same assay, the three alternatives for a ribosome poised at a stop codon (termination, read-through and frameshift). A quantitative analysis of the competition has been done in the presence and absence of release factor (RF) mutants, nonsense suppressors and an upstream Shine-Dalgarno-like sequence. The ribosomal +1 frameshift product is measurable when the stop codon is decoded by wild-type or mutant RF (prf A1 or prf B2) and also in the presence of competing suppressor tRNAs. Frameshift frequency appears to be influenced by RF activity. The amount of frameshift product decreases in the presence of competing suppressor tRNAs, however, this decrease is not in proportion to the corresponding increase in the suppression product. Instead, there is an increase in the total amount of protein expressed from the gene, perhaps due to the purging of queued ribosomes. Mutated RFs reduce the total output of the reporter gene by reducing the amount of all three protein products. The nascent peptide has earlier been shown to influence the translation termination process by interacting with the RFs. At 42 degrees C in a temperature-sensitive RF mutant strain, protein measurements indicate that the nascent peptide seems to influence the binding efficiencies of the RFs.     

Ribosomal frameshifting requires a pseudoknot in the Saccharomyces cerevisiae double-stranded RNA virus.

The large double-stranded RNA of the Saccharomyces cerevisiae (yeast) virus has two large overlapping open reading frames on the plus strand, one of which is translated via a -1 ribosomal frameshift. Sequences including the overlapping region, placed in novel contexts, can direct ribosomes to make a -1 frameshift in wheat germ extract, Escherichia coli and S. cerevisiae. This sequence includes a consensus slippery sequence, GGGUUUA, and has the potential to form a pseudoknot 3' to the putative frameshift site. Based on deletion analysis, a region of 71 nucleotides including the potential pseudoknot and the putative slippery sequence is sufficient for frameshifting. Site-directed mutagenesis demonstrates that the pseudoknot is essential for frameshifting.     

Efficient stimulation of site-specific ribosome frameshifting by antisense oligonucleotides

Evidence is presented that morpholino, 2'-O-methyl, phosphorothioate, and RNA antisense oligonucleotides can direct site-specific -1 translational frameshifting when annealed to mRNA downstream from sequences where the P- and A-site tRNAs are both capable of repairing with -1 frame codons. The efficiency of ribosomes shifting into the new frame can be as high as 40%, determined by the sequence of the frameshift site, as well as the location, sequence composition, and modification of the antisense oligonucleotide. These results demonstrate that a perfect duplex formed by complementary oligonucleotides is sufficient to induce high level -1 frameshifting. The implications for the mechanism of action of natural programmed translational frameshift stimulators are discussed.     

IF3-mediated suppression of a GUA initiation codon mutation in the recJ gene of Escherichia coli.

A mutational change of the initiation codon to GUA was found to reduce, but not abolish, expression of the recJ gene of Escherichia coli. Specific mutations in translational initiation factor IF3 have been isolated as second-site suppressors of this GUA initiation codon mutation. One of these, infC135, with an arginine-to-proline change at amino acid 131, completely restores a wild-type phenotype to recJ GUA initiation codon mutants and acts in a semidominant fashion. The infC135 mutation increased expression of RecJ from the GUA mutant but had no effect on the normal GUG start. The infC135 mutation also abolished autoregulation of IF3 in cis and in trans. The behavior of this IF3 mutant suggests that it has specifically lost its ability to abort initiation from poor initiation codons such as GUA of recJ and the AUU of infC. Because of the impact of IF3 on recJ, a recombination and repair gene, this role of IF3 must be general and not restricted to translation genes. The dominance of infC135 suggests that the other functions of IF3, for instance its ability to bind to 30S ribosomes, must remain intact. Although the ability to discriminate among initiation codons has been lost in the infC135 mutant, translational initiation was still restricted to the normal initiation site in recJ, even in the presence of a closely juxtaposed alternative initiation codon. Because the recJ gene lacks a canonical Shine-Dalgarno sequence, other unknown features of the mRNA must serve to specify the initiation site.     

Near-cognate peptidyl-tRNAs promote +1 programmed translational frameshifting in yeast

Translational frameshifting is a ubiquitous, if rare, form of alternative decoding in which ribosomes spontaneously shift reading frames during translation elongation. In studying +1 frameshifting in Ty retrotransposons of the yeast S. cerevisiae, we previously showed that unusual P site tRNAs induce frameshifting. The frameshift-inducing tRNAs we show here are near-cognates for the P site codon. Their abnormal decoding induces frameshifting in either of two ways: weak codon-anticodon pairing allows the tRNA to disengage from the mRNA and slip +1, or an unusual codon-anticodon structure interferes with cognate in-frame decoding allowing out-of-frame decoding in the A site. We draw parallels between this mechanism and a proposed mechanism of frameshift suppression by mutant tRNAs.     

Influence of the stacking potential of the base 3' of tandem shift codons on -1 ribosomal frameshifting used for gene expression

Translating ribosomes can shift reading frame at specific sites with high efficiency for gene expression purposes. The most common type of shift to the -1 frame involves a tandem realignment of two anticodons from pairing with mRNA sequence of the form X XXY YYZ to XXX YYY Z where the spaces indicate the reading frame. The predominant -1 shift site of this type in eubacteria is A AAA AAG. The present work shows that in Escherichia coli the identity of the 6 nt 3' of this sequence can be responsible for a 14-fold variation in frameshift frequency. The first 3' nucleotide has the primary effect, with, in order of decreasing efficiency, U > C > A > G. This effect is independent of other stimulators of frameshifting. It is detected with other X XXA AAG sequences, but not with several other heptameric -1 shift sites. Pairing of E. coli tRNALYS with AAG is especially weak at the third codon position. We propose that strong stacking of purines 3' of AAG stabilizes pairing of tRNALys, diminishing the chance of codon:anticodon dissociation that is a prerequisite for the realignment involved in frameshifting.     

Role of premature translational termination in the regulation of expression of the phi X174 lysis gene

Expression of the phi X174 lysis (E) gene, a member of an overlapping gene pair, appears to depend on a frameshift-induced chain termination by ribosomes translating the upstream D gene. A -1 reading frameshift, possibly induced by misreading of an alanine codon as a doublet, causes ribosomes to terminate translation at two different sites, suggesting two modes of regulating expression of the E gene. One frameshift can cause translational termination at a stop codon(s) near the E gene ribosome binding site (RBS), resulting in reinitiation by ribosomes at the E gene RBS. Termination at a second site some 70 bases upstream from the E gene RBS, while too far away to allow ribosomal re-initiation at the E gene RBS, probably results in an unmasking of the message, allowing entry of a new ribosome at the E gene RBS.