Translational stop codons, UAA, UAG, and UGA, form an integral part of the universal genetic code. They are of significant interest today for their underlying fundamental role in terminating protein synthesis, but also for their potential utilisation for programmed alternative translation events. In diverse organisms, UAA has wide usage, but it is puzzling that the high fidelity UAG is selected against and yet UGA, vulnerable to suppression, is widely used, particularly in those archaeal and bacterial genomes with a high GC content. In canonical protein synthesis, stop codons are interpreted by protein release factors that structurally and functionally mimic decoding tRNAs and occupy the decoding site on the ribosome. The release factors make close contact with the decoding complex through multiple interactions. Correct interactions cause conformational changes resulting in new and enhanced contacts with the ribosome, particularly between specific bases in the mRNA and rRNA. The base following the stop codon (fourth or +4 base) may strongly influence decoding efficiency, facilitating alternative non-canonical events like frameshifting or selenocysteine incorporation. The fourth base is drawn into the decoding site with a compacted stop codon in the eukaryotic termination complex. Surprisingly, mRNA sequences upstream and downstream of this core tetranucleotide signal have a significant influence on the strength of the signal. Since nine bases downstream of the stop codon are within the mRNA channel, their interactions with rRNA, and r-proteins may affect efficiency. With this understanding, it is now possible to design stop signals of desired strength for specific applied purposes.

Background

The solving of the triplet genetic code in the 1960s was an exciting time for molecular biology with most of the triplet code words specifying an amino acid, but with specific signals for stopping protein synthesis predicted. Studies of nonsense mutations that terminated synthesis and were associated with production of amino-terminal fragments of the protein under study [1] led to the discovery that three remaining codons in the triplet genetic code were ‘reserved’ for natural translational stop signals, the amber codon UAG, the ochre codon UAA, and the opal codon UGA [2]. While the triplet base signals for the amino acids were matched by bases pairing from specific tRNAs, initially a similar mechanism for recognition of the stop codons seemed possible.

A cautionary flag was raised, however, when it was discovered that these codons were not decoded by an RNA but instead by proteins [3,4]. The discovery of two specific decoding release factors (RFs) in bacteria, each with specificity for one of the three stop codons [5,6] moreover, suggested that there might be direct recognition between the decoding factor and the RNA signal, and then the signal would not necessarily be restricted to simply three bases. Subsequent studies have found organisms or organelles with variations in this universal code, and even without a simple dedicated stop codon [7,8]; however, in this review, we focus on organisms using the standard cellular codes (archaea, bacteria, and eukaryotes)

The evolution and expansion of these ideas have inspired many subsequent investigations and have contributed to many areas of biology (see recent reviews, [814]). In our focused minireview on the ‘stop signal’, we have concentrated on the evolution of ideas and investigations that have led to current knowledge, recent insights, and the future prospects for this very versatile translational signal.

Decoding release factors recognise the stop codon directly

Although postulated, close physical contact between the decoding factor and the stop codon was not demonstrated until the first U base of the stop codon was substituted with 4-thioU. This created a zero length cross-linking agent, and a covalent bond was formed between the bacterial decoding release factor-2 (RF2) and the stop codon [15]. A tRNA analogue model was proposed [16], essentially as shown in Figure 1A, that the bacterial decoding RF squeezed into the tRNA decoding ‘A’ site, spanning the small subunit decoding and large subunit catalytic centres of the ribosome — just like a tRNA. Later, hydroxyl radical footprinting that utilised production of radicals at specific engineered sites on RF2 implicated biochemically the decoding and peptidyl-tRNA hydrolysis functions and demonstrated that they were in close contact with the respective ribosomal centres [17]. Subsequent cryo-electron microscopy (EM) and X-ray structures of the decoding factors on the ribosome confirmed this concept [1824] (see Figure 2). These structures show that decoding factors occupy the ribosomal A-site and orientate to bind directly to the stop codons through conserved RF protein motifs. The interactions at the A-site are, in turn, stabilised by base stacking between the rRNA and the mRNA specifically through 16S helix 44 in bacteria and 18S h44 in eukaryotes (Figure 2). Notably, the rRNA (16S G530) is involved in stabilisation of the third base in bacteria and of the fourth in eukaryotes (18S G626).

Translation termination at strong and weak termination contexts.

Figure 1.
Translation termination at strong and weak termination contexts.

(A) An efficient or strong stop signal is shown as ‘STOP’ this would include the stop codon and optimal mRNA context that interacts with the RF, ribosome mRNA (green) channel (dotted), or P-site tRNA (P). (B) Poor or weak stop signals. Some potential competing or influencing events are indicated. Suppressor or near cognate tRNAs, or other protein factors, could complete with the termination process, resulting in a failure to terminate (e.g. recoding, frameshifting, readthrough, and stalling).

Figure 1.
Translation termination at strong and weak termination contexts.

(A) An efficient or strong stop signal is shown as ‘STOP’ this would include the stop codon and optimal mRNA context that interacts with the RF, ribosome mRNA (green) channel (dotted), or P-site tRNA (P). (B) Poor or weak stop signals. Some potential competing or influencing events are indicated. Suppressor or near cognate tRNAs, or other protein factors, could complete with the termination process, resulting in a failure to terminate (e.g. recoding, frameshifting, readthrough, and stalling).

Decoding of stop codons by bacterial and eukaryotic release factors.

Figure 2.
Decoding of stop codons by bacterial and eukaryotic release factors.

(A) In bacteria, RF1 (grey side chains) recognises an extended stop codon configuration (UAAAA), where H197 intercalates and the +3 base stacks on G530 of the 16S rRNA. H bonds are shown as black dotted lines (PDB: 4V63). The +4 and +5 A's could form interactions (green) with channel proteins (S3, S5) or 16S rRNA C1397 [72]. These A-site codon and mRNA interactions with rRNA and proteins are different from initiation or elongation complexes [75]. (B) RF1 interaction with UAG, adapted from ref. [25]. (C) The core termination signal recognised by eRF1 is formed by four mRNA bases (+1 to +4, slate) that compact into the A-site. Bases +2 and+3 stack on A1825, which is flipped out of helix 44 (h44), and base +4 on G626 of 18S rRNA (yellow) and +5 on C1698, adapted from ref. [72], (D) RF2 UAA, and (E) RF2 UGA adapted from ref. [25], (F) The core stop codon recognition rules for the bacterial release factors RF1 and RF2. (G) The core recognition sequences in eukaryotic or archaeal RFs.

Figure 2.
Decoding of stop codons by bacterial and eukaryotic release factors.

(A) In bacteria, RF1 (grey side chains) recognises an extended stop codon configuration (UAAAA), where H197 intercalates and the +3 base stacks on G530 of the 16S rRNA. H bonds are shown as black dotted lines (PDB: 4V63). The +4 and +5 A's could form interactions (green) with channel proteins (S3, S5) or 16S rRNA C1397 [72]. These A-site codon and mRNA interactions with rRNA and proteins are different from initiation or elongation complexes [75]. (B) RF1 interaction with UAG, adapted from ref. [25]. (C) The core termination signal recognised by eRF1 is formed by four mRNA bases (+1 to +4, slate) that compact into the A-site. Bases +2 and+3 stack on A1825, which is flipped out of helix 44 (h44), and base +4 on G626 of 18S rRNA (yellow) and +5 on C1698, adapted from ref. [72], (D) RF2 UAA, and (E) RF2 UGA adapted from ref. [25], (F) The core stop codon recognition rules for the bacterial release factors RF1 and RF2. (G) The core recognition sequences in eukaryotic or archaeal RFs.

The complex array of interactions that form between the different residues within the factors and rRNA determines codon recognition and, in the case of the two bacterial factors discriminate between UAG and UGA codons, and excludes recognition of UGG [25]. Hydroxyl radical footprinting has also been used sensitively to distinguish how the RF distinguishes a stop codon from a near cognate codon at the decoding centre and communicates this distinction through its structure to the catalytic centre [26] by showing that there is a tighter interaction between the factor and stop codon than the near cognate codon. Other regions of the RF then undergo conformational changes that alter the interaction of the factor with the catalytic centre [26]. It can be concluded that the RF adopts a very distinct orientation to achieve maximal catalytic activity for facilitating the release of the completed polypeptide.

Molecular dynamic simulations of cognate and near cognate codon RF–ribosomal complexes aimed to understand the energetics of the high accuracy of RF binding. They proposed new interactions and recognition switches [27] that could not be explained by the three amino acid ‘peptide anticodon models’ [28]. The codon recognition regions of the bacterial factors, bacterial decoding release factor-1 RF1 and RF2, include an external loop of 13 amino acids (Figure 2). While three key ‘anticodon-like’ residues in the loop were part of the recognition motif (PAT in RF1 and SPF in RF2, Figure 2F), systematic site-directed mutagenesis to shift an RF1 loop to an RF2 showed that recognition was more complex [29]. Most of the amino acids in the loop influence the specificities of recognition of UAG and UGA by the two factors (Figure 2B,E). Critically, differences in amino acids within this recognition loop between the two factors allow either stringent or more flexible recognition at the second or the third positions of the codon. For example, RF1 recognises only a second base A, but third base A and G, while the converse occurs with RF2 where second base recognition is flexible allowing A or G, but third base recognition is stringent, only allowing A [29,30]. Having deciphered stop codon recognition rules from this study (Figure 2F), we tested them by creating two novel recognition loops that were predicted to stringently recognise only UAA, or be flexible in both positions, recognising all three stop codons and UGG [2931]. Indeed, the recombinant RFs showed the predicted codon recognition properties [31]. An alternative route, changing residues outside SPF, towards an omnipotent RF2 (R213I, Figure 2F), was recently used [32].

The two bacterial and the one omnipotent eukaryotic or archaeal decoding factors have evolved independently, being built on different structural principles with different codon recognition motifs [6,33,34]. In the eukaryotic release factor-1 (eRF1), two loops 15 Å apart within the N-domain [35] containing YxCxxxF and NIKS motifs are important for codon recognition [36,37] (Figure 2G). These observations, together with an in silico analysis of eRF1 sequences [38], support a more complex pattern of stop recognition (Figure 2C), compared with bacteria. Cross-linking moieties on the A's of UAA cross-linked to residues on the N-domain of eRF1 near the YxCxxxF motif, and these were different when G's were present at +2 and +3 [39]. But, cross-linking from U1 modifies a residue near the NIKS loop [39]. As well there are conformational changes in the rRNA of both ribosomal subunits on eRF1 binding [40], and the eRF1 M domain has a specific orientation to the 3′−terminus of the P-site tRNA carrying the polypeptide to be released, analogous to the specific orientation of the N-domain to the stop signal [41].

The X-ray and cryo-EM structural studies of the bacterial and eukaryotic release factors supported the premise that they have evolved different solutions for discriminating against UGG as a stop codon (Figure 2). The two bacterial factors RF1 and RF2 exist as a ‘closed form’ off the ribosome [42,43] unlike a tRNA, but domain 3 unfolds when on the ribosome to structurally mimic a tRNA [19,20]. In contrast, the eRF1 structure exhibits the tRNA structural mimicry both off [35] and on the ribosome [24], although the conformation also differs significantly in the ribosome-bound state [39].

The translational stop signal at length — three, four or more

Inefficient stop signals

The expectation of universally high fidelity of recognition by an RF of a triplet stop codon, if the codon was the sole mRNA determinant, proved not to be true. Indeed, such failure or infidelity can be programmed, and it is now termed a recoding event [9,44]. For example, the efficiency of readthrough of the stop codon and continuing protein synthesis varies enormously according to the stop codon and its context [4547]. This is essentially a competition between two mechanisms, a reading of the stop codon by a near cognate tRNA (two out of three base recognition) and the faithful decoding of the signal by the RF (Figure 1B). Either the stop infidelity could arise from enhancement of the competition from near cognate tRNAs misreading codon — having a pseudouracil in the first position of the stop codon is an example of this [48] — or from low efficiency decoding by the RF itself. However, the discovery of stop codons within coding regions where some events occurred outside of the constraints of the classical genetic code, such as translational frameshifting [49], and incorporation of selenocysteine, the 21st amino acid [50,51], suggested to us that there might be ‘weak’ and ‘strong’ translational stop signals; ‘weak’ used as a ‘pause’ in translation allowing ‘kinetic time’ for an alternative genetic recoding event [52,53]. ‘Strong’ stop signals, by contrast, exhibited the high decoding fidelity expected [54] and had evolved for highly efficient termination. The enigma was how a particular stop codon (e.g. UGA) could be decoded as both a strong and a weak signal depending on its context.

Weak and strong stop signals

The rare examples of naturally occurring programmed translational suppression, frameshifting, selenocysteine incorporation, or translational ‘hops’ at specific mRNA stop signals [44] suggested that there were either hidden features of the mRNA or stop signal itself that permitted such infidelities. What was not initially clear was whether the ribosome and release factor were simply ‘fooled’ by ‘recoding signals’ in the mRNA facilitating specific recoding events, or whether an efficient stop signal was longer than the triplet codon, and thereby the stop signal itself could be ‘weak’ and contribute to this.

Indeed, initially, it seemed that the recoding signals might be the sole explanation. However, while a bacterial in vitro ribosomal assay for termination could be mediated by three nucleotide RNAs, UAA, UAG, and UGA alone [55]; in contrast, in the equivalent eukaryotic assay, these triplet codons were insufficient and four nucleotide RNAs (UAAN, UGAN, or UAGN) were required for activity [56]. This suggested that a fourth base may be integral to the stop.

Resolution of this apparent paradox was helped by the discovery of translational frameshifting at a UGA stop codon within a bacterial mRNA for the prfB gene [49]. This serendipitous discovery in the coding region for the RF that decoded UGA stop codons (RF2) — a protein we were intensely studying — was suggestive of an exquisite mechanism of autoregulation, as the RF2 gene product of prfB fed back onto the internal UGA in the coding region of its mRNA during translation. With our strong interest in stop signals, we carefully evaluated the possibility that the prfB stop codon context itself was a significant contributor to the high frameshifting efficiency. We discovered that of the 12 possible [STOP N] sequences, UGAC was the weakest [57], and of the 64 possible UGANNN sequences (Table 1), the UGACUA sequence at the frameshift site was the least competitive against frameshifting [58]. Indeed, this UGA context is avoided elsewhere at natural stop signals in the Escherichia coli genome [59]. Here, was an example then of a ‘weak’ stop signal that was allowing the alternative recoding event to occur, along with other facilitating elements in the mRNA [9,52].

Table 1
Extended translation termination signals
Species Kingdom Simple signal1 Extended2 Weak3 References4 
E. coli UAAU/G AAG UAA UCU UGAC, UAGC [34,52,69,71,87
GCG UAG GAG 
UUC UGA UUU 
Bacillus subtilis UAAA UAAaaaa NNNC,UAG [34,71
Thermus thermophilus UAAG/A UA/GAgg NNNY  
S. cerevisiae UAAA/G UAAac7 NNNC, UGAC, UAG [67,69,70,76,88,89
Caenorhabditis elegans UAA UAA UGAC [34
Arabidopsis thaliana UG/AAA/G UAAa NNNC, UGACARYYA6 [46,90
D. melanogaster UAAA/G UAAgc NNNC, UGAU/G [76,91
Homo sapiens UAAG/A UAA GUAG UGAC, UAG [76,92,93
Methanocaldococcus jannaschii5 UG/AAA UAAa NNNc/u, UAG [94
Aeropyrum pernix UAG/A UAG/A nd [33
Species Kingdom Simple signal1 Extended2 Weak3 References4 
E. coli UAAU/G AAG UAA UCU UGAC, UAGC [34,52,69,71,87
GCG UAG GAG 
UUC UGA UUU 
Bacillus subtilis UAAA UAAaaaa NNNC,UAG [34,71
Thermus thermophilus UAAG/A UA/GAgg NNNY  
S. cerevisiae UAAA/G UAAac7 NNNC, UGAC, UAG [67,69,70,76,88,89
Caenorhabditis elegans UAA UAA UGAC [34
Arabidopsis thaliana UG/AAA/G UAAa NNNC, UGACARYYA6 [46,90
D. melanogaster UAAA/G UAAgc NNNC, UGAU/G [76,91
Homo sapiens UAAG/A UAA GUAG UGAC, UAG [76,92,93
Methanocaldococcus jannaschii5 UG/AAA UAAa NNNc/u, UAG [94
Aeropyrum pernix UAG/A UAG/A nd [33

Abbreviations: nd: not determined/not tested.

1

If an efficient stop signal has been identified by experiment or bioinformatics, it is listed. Bold are experimentally supported.

2

Overrepresented bases are shown in lowercase [64]. Bold are experimentally supported.

3

In all the species listed, UGA also encodes Sec, but these are not Sec insertion sites.

4

Includes reviews or recent research papers and references therein, all species are found in the TransTerm database [64].

5

UAG also encodes Pyl the 22 amino acid in this species.

6

CARYYA following stops has promoted readthrough in all eukaryotes tested.

7

The last codons and amino acids may have a role (see [74]).

Is the fourth base important for the stop signal?

UGA stop codons that allow incorporation of selenocysteine at UGA codons intriguingly also most often contain ‘C’ as the base following the UGA at the selenocysteine (Sec) insertion site. Examples are the bacterial formate dehydrogenase F (fdhF) and the eukaryotic type I iodothyronine 5′-deiodinase [14,50,60]. There is a UAGC at the site in the bacteriophage T4 gene 60 where ribosome detachment on the mRNA occurs at UAGC followed by ratcheting and then re-engagement at some distance downstream on the mRNA (hopping) [9]. These examples suggested that the base following the stop codon might be an important determinant of termination efficiency. Indeed, when the fdhF UGAC was changed to UGAU, Sec incorporation decreased significantly, and the termination event became much more competitive over Sec insertion. Similarly, if UGAC in the human deiodinase incorporation site was changed to UGAA/G, Sec incorporation decreased markedly, and the termination product increased concordantly [53] (Table 1).

The fourth base following the stop codon is a core part of an efficient signal

Once gene sequences were available in sufficient numbers from the prokaryote, E. coli, and a variety of eukaryotic species like Saccharomyces cerevisiae, Drosophila melanogaster, and a range of mammals, the contexts of the natural stop codons could be explored. An early initial study of 73 eukaryotic sequences by Kohli and Grosjean [61] indicated that there might be a preferred context for efficient termination. We established and continue to maintain a database TransTerm that includes stop contexts [6264] (Table 1). Remarkably, the first analyses of the TransTerm data base showed for bacterial, archaeal, and eukaryotic organisms that there was a strong bias in the bases around the stop codon for most organisms examined. There was a marked bias in the fourth base with lesser bias in other positions upstream and downstream of the stop codons [65] (Table 1). Highly expressed genes were a subset that were selected from the TransTerm database — of the E. coli genes the bias in the fourth base was more extreme and approximately two-thirds had U in this position [52,57]. Interestingly, in the similar subset of mammalian highly expressed genes, 90% had A or G in the fourth base position [65] (Table 1).

These data suggested that the fourth base was a key determinant of how rapidly the stop signal was decoded and we tested this hypothesis experimentally in both pro- and eukaryotic systems at recoding sites where there was competition for the mechanistic outcome (Table 1). In bacteria, there was wide ranging gradation (50-fold) of UA/GA/GN signal strengths as determined from an in vivo bacterial termination assay. This hierarchy of strengths matched their frequency of occurrence at natural sites — a fourth base, U or G, gave the strongest signals (the predominant signal in highly expressed genes is UAAU), while a fourth base, A or C, provided the weaker signals [59,65,66]. In eukaryotes, although a gradation of strengths was confirmed by an in vitro mammalian termination assay (70-fold range) in this case, it was the fourth base A and G that were the strongest signals and U and C were the weakest [53] (Table 1). A range of strengths was also seen in vivo in S. cerevisiae with a good correlation between usage and inefficient readthrough, with UGAG and UAAG and having the least readthrough (<0.5%) [67]. Furthermore, an extended six base sequence selected from a degenerate pool (CARNBA) promoted 5% readthrough, the authors speculated by interaction with the rRNA around position 500 [68]. Upstream sequences, possibly through the tRNA or protein sequence, can also affect readthrough in S. cerevisiae [69,70] and bacteria [71].

The structural basis for eRF1 stop codon recognition in eukaryotes has provided explanation of why the fourth base is a key part of the stop signal. Structural studies [24,72,73] showed the stop codon in the mRNA is compacted during recognition at the ribosomal A-site, so the fourth base is drawn into this site as well — four bases now fit into space that is normally would accommodate only three codon bases (Figure 2C). This fourth base flips out, and purines stack with G626 of the 18S rRNA. Hence, while a specific identity of the fourth base is not required for stop codon recognition as for the first three bases, this fourth base is important for modulating the strength of the signal and is important in regulating translation.

Is the termination signal even longer than four bases?

Diverse bacterial and eukaryotic genomes were compared to deduce whether the protein synthesis termination signal has more of the features of a sequence element than simply a tetrameric signal and whether there were common determinants within and across both bacterial and eukaryotic kingdoms. The genomes of four independently living bacteria and many eukaryotes were investigated in detail and demonstrated a similar pattern of bias at both 5′ and 3′ of the stop codon [74], suggesting a stop signal of the form NNN [STOPN]NNN. The preferred core signal of four bases was evident, encompassing the stop codon and the base following. Codons decoded by hyper-modified tRNAs were over-represented in the region 5′ to the stop codon in genes from both kingdoms, causing a bias. The nature of the 3′ bias was more variable, particularly among the bacterial organisms [64]. In both kingdoms, genes with the highest expression index exhibited a strong bias, but genes with the lowest expression showed none. In E. coli, signal abundance correlated with termination efficiency for UAA and UGA stop codons, but not in mammalian cells. Termination signals that were inefficient could be made more efficient by enhancing concentrations of the cognate decoding release factor [74].

How did the observed bias on both sides of the stop codon affect the termination efficiency? The bias in usage in E coli of NNN UAANNN, NNN UAGNNN, and NNN UGANNN was used as an initial predictor of the strength of the stop signals, and so the most and least abundant were tested for their efficiency to compete in vivo with a readthrough recoding event. The frequency at natural termination sites indeed predicted signal strength, such that ‘designer signals’ of the model (5′ NNN [STOPN] NN 3′) where each 3′ or 5′ context could be either weak or strong showed the predicted strengths when tested experimentally. When all downstream contexts of the format UAANNN, UAGNNN, and UGANNN were tested, the fifth and sixth bases had additional effects above that of the core four base signals, and there were even small effects further downstream for the seventh to ninth positions.

How could such distant bases downstream from the stop codon affect the strength of the signal? When a stop codon is at the A-site of the ribosome, there are six nucleotides (+4 to +9) downstream that are deduced to be associated with the ribosome occupying the mRNA channel [75,76]. They are in close proximity to ribosomal RNA nucleotides and ribosomal proteins [72,75] (Figure 2A,C). The interactions discernible at +4 and +5 in the RF1 UAA AA termination structure (Figure 2A) differ from that of an elongation or initiation complex [72,75]. A termination complex with a +4 U has not yet been published, but in elongation a +5 U (+8 from the A-site) interacts with 16S helix 34 (U1196), C1397 and S3, whereas a +4 or +5 A in a termination complex does not interact with these (Figure 2A).

In the eukaryotic complex, the +4 G stacks with G626 and a +5 C with C1698 (Figure 2C). We showed that downstream nucleotides affected translation termination success or failure in mammalian cells at each of the three stop codons [76]. With highly substantial effects on termination readthrough facilitated by nucleotides at the fourth position as expected, but also at the eighth position, and a more complex pattern of influence from the nucleotides at positions +5 and +6 in combination with contexts at other positions. When termination efficiency was significantly reduced artificially by cognate suppressor tRNAs, this observed pattern was maintained. We propose that termination efficiency is influenced not only by interrogation of the stop signal directly by the eRF1 (first–fourth base) but also by downstream ribosomal interactions with the mRNA nucleotides in the entry channel [75,76].

Significance and future perspectives

This minireview has focused on efficient stop signals in well-studied organisms, while highlighting there are natural instances where these stops can be programmed to be recoded in particular genes (e.g. selenocysteine) [9,14,77,78].

To use stop codons to advantage in some biological applications, it is important to optimise and preserve the highest fidelity, for example, for high level expression of proteins [11]. In other applications, however, the aim may be to prevent termination and allow incorporation of novel amino acids. Indeed, this has been a major area of recent research [79]. One goal is to reverse genetic disorders caused by nonsense codons, and several classes of drugs to suppress these mutations are now either in use or in development [10]. Early studies indicated that aminoglycosides like preparations of gentamicin could promote infidelity and readthrough [80]. It has recently been shown that the minor component, gentamicin B1, is the readthrough active component of gentamicin [81]. The drug ataluren may also target the ribosome to actively promote readthrough of premature stop codons [82].

Other research aims to extend the range of genetic coding by suppressing stop codons and allowing incorporation of novel amino acids. In these studies, de-optimising the stop context promotes incorporation [11]. It uses the principles that weak stop codon contexts enhance the competitiveness of the competing translational event just like classic suppressor tRNAs.

Under the diverse pressures of mutation and selection throughout biology, the use of stop codons has evolved, and some have been reassigned (notably in mitochondria and ciliates). Possibly the strangest situation is that recently discovered in several eukaryotic species, where UAA means stop very near the 3′-end of the mRNA, but sense within the coding region [7,8385]. This is a special case of context, where the 3′-end proximity, by some unknown mechanism, promotes termination [7].

A long-standing mystery is why UAG is the least commonly used stop codon, particularly in bacteria (Table 1) [64]. UAA is more common in AT-rich genomes and UGA in GC-rich organisms [59], although there are some differences between bacteria, with two RFs, and archaea and eukarya, with one RF [34]. Functionally UAG works well in termination assays, and where tested is not more suppressible than UAA, unlike the recodeable and suppressible UGA [86]. This led to questions of the relative role of selection or mutation in UAA/UAG abundance [66]. Recent studies support the idea that UAA is retained by selection, but UAG selected against, switching to other stop codons [34].

The stop codon seemed initially to be the most ‘one-dimensional’ and least interesting of all the translational signals. Remarkably, it has proved to be multidimensional and perhaps the most interesting in how it is used and interpreted for fine tuning the molecular biology of the cell and gene expression. It is providing opportunities as a specific target for increasing the diversity of proteins and for improving outcomes from a range of genetic diseases.

Abbreviations

     
  • EM

    electron microscopy

  •  
  • eRF1

    eukaryotic release factor-1

  •  
  • RF2

    release factor-2

  •  
  • RFs

    release factors

Acknowledgments

W.T., A.C., and C.B thank the University of Otago, the Marsden Fund of New Zealand, and the Health Research Council of New Zealand for grants supporting our work. The authors thank Bronwyn Carlisle for contributions to the production of figures.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Garen
,
A.
(
1968
)
Sense and nonsense in the genetic code. Three exceptional triplets can serve as both chain-terminating signals and amino acid codons
.
Science
160
,
149
159
2
Brenner
,
S.
,
Barnett
,
L.
,
Katz
,
E.R.
and
Crick
,
F.H.
(
1967
)
UGA: a third nonsense triplet in the genetic code
.
Nature
213
,
449
450
3
Capecchi
,
M.R.
(
1967
)
Polypeptide chain termination in vitro: isolation of a release factor
.
Proc. Natl Acad. Sci. U.S.A.
58
,
1144
1151
4
Caskey
,
C.T.
,
Tompkins
,
R.
,
Scolnick
,
E.
,
Caryk
,
T.
and
Nirenberg
,
M.
(
1968
)
Sequential translation of trinucleotide codons for the initiation and termination of protein synthesis
.
Science
162
,
135
138
5
Scolnick
,
E.
,
Tompkins
,
R.
,
Caskey
,
T.
and
Nirenberg
,
M.
(
1968
)
Release factors differing in specificity for terminator codons
.
Proc. Natl Acad. Sci. U.S.A.
61
,
768
774
6
Lind
,
C.
,
Oliveira
,
A.
and
Aqvist
,
J.
(
2017
)
Origin of the omnipotence of eukaryotic release factor 1
.
Nat. Commun.
8
,
1425
7
Lobanov
,
A.V.
,
Heaphy
,
S.M.
,
Turanov
,
A.A.
,
Gerashchenko
,
M.V.
,
Pucciarelli
,
S.
,
Devaraj
,
R.R.
et al. 
(
2017
)
Position-dependent termination and widespread obligatory frameshifting in Euplotes translation
.
Nat. Struct. Mol. Biol.
24
,
61
68
8
Hellen
,
C.U.T.
(
2018
)
Translation termination and ribosome recycling in eukaryotes
.
Cold Spring Harb. Perspect. Biol.
10
,
a032656
9
Atkins
,
J.F.
,
Loughran
,
G.
,
Bhatt
,
P.R.
,
Firth
,
A.E.
and
Baranov
,
P.V.
(
2016
)
Ribosomal frameshifting and transcriptional slippage: from genetic steganography and cryptography to adventitious use
.
Nucleic Acids Res.
44
,
7007
7078
10
Bidou
,
L.
,
Bugaud
,
O.
,
Belakhov
,
V.
,
Baasov
,
T.
and
Namy
,
O.
(
2017
)
Characterization of new-generation aminoglycoside promoting premature termination codon readthrough in cancer cells
.
RNA Biol.
14
,
378
388
11
Chemla
,
Y.
,
Ozer
,
E.
,
Algov
,
I.
and
Alfonta
,
L.
(
2018
)
Context effects of genetic code expansion by stop codon suppression
.
Curr. Opin. Chem. Biol.
46
,
146
155
12
Schuller
,
A.P.
and
Green
,
R.
(
2018
)
Roadblocks and resolutions in eukaryotic translation
.
Nat. Rev. Mol. Cell Biol.
19
,
526
541
13
Huter
,
P.
,
Muller
,
C.
,
Arenz
,
S.
,
Beckert
,
B.
and
Wilson
,
D.N.
(
2017
)
Structural basis for ribosome rescue in bacteria
.
Trends Biochem. Sci.
42
,
669
680
14
Vindry
,
C.
,
Ohlmann
,
T.
and
Chavatte
,
L.
(
2018
)
Translation regulation of mammalian selenoproteins
.
Biochim. Biophys. Acta, Gen. Subj.
1862
,
2480
2492
15
Tate
,
W.
,
Greuer
,
B.
and
Brimacombe
,
R.
(
1990
)
Codon recognition in polypeptide chain termination: site directed crosslinking of termination codon to Escherichia coli release factor 2
.
Nucleic Acids Res.
18
,
6537
6544
16
Moffat
,
J.G.
and
Tate
,
W.P.
(
1994
)
A single proteolytic cleavage in release factor 2 stabilizes ribosome binding and abolishes peptidyl-tRNA hydrolysis activity
.
J. Biol. Chem.
269
,
18899
18903
PMID:
[PubMed]
17
Scarlett
,
D.J.
,
McCaughan
,
K.K.
,
Wilson
,
D.N.
and
Tate
,
W.P.
(
2003
)
Mapping functionally important motifs SPF and GGQ of the decoding release factor RF2 to the Escherichia coli ribosome by hydroxyl radical footprinting. Implications for macromolecular mimicry and structural changes in RF2
.
J. Biol. Chem.
278
,
15095
15104
18
Petry
,
S.
,
Brodersen
,
D.E.
,
Murphy
,
F.
,
Dunham
,
C.M.
,
Selmer
,
M.
,
Tarry
,
M.J.
et al. 
(
2005
)
Crystal structures of the ribosome in complex with release factors RF1 and RF2 bound to a cognate stop codon
.
Cell
123
,
1255
1266
19
Rawat
,
U.B.
,
Zavialov
,
A.V.
,
Sengupta
,
J.
,
Valle
,
M.
,
Grassucci
,
R.A.
,
Linde
,
J.
et al. 
(
2003
)
A cryo-electron microscopic study of ribosome-bound termination factor RF2
.
Nature
421
,
87
90
20
Klaholz
,
B.P.
,
Pape
,
T.
,
Zavialov
,
A.V.
,
Myasnikov
,
A.G.
,
Orlova
,
E.V.
,
Vestergaard
,
B.
et al. 
(
2003
)
Structure of the Escherichia coli ribosomal termination complex with release factor 2
.
Nature
421
,
90
94
21
Laurberg
,
M.
,
Asahara
,
H.
,
Korostelev
,
A.
,
Zhu
,
J.
,
Trakhanov
,
S.
and
Noller
,
H.F.
(
2008
)
Structural basis for translation termination on the 70S ribosome
.
Nature
454
,
852
857
22
Korostelev
,
A.
,
Asahara
,
H.
,
Lancaster
,
L.
,
Laurberg
,
M.
,
Hirschi
,
A.
,
Zhu
,
J.
et al. 
(
2008
)
Crystal structure of a translation termination complex formed with release factor RF2
.
Proc. Natl Acad. Sci. U.S.A.
105
,
19684
19689
23
Weixlbaumer
,
A.
,
Jin
,
H.
,
Neubauer
,
C.
,
Voorhees
,
R.M.
,
Petry
,
S.
,
Kelley
,
A.C.
et al. 
(
2008
)
Insights into translational termination from the structure of RF2 bound to the ribosome
.
Science
322
,
953
956
24
Brown
,
A.
,
Shao
,
S.
,
Murray
,
J.
,
Hegde
,
R.S.
and
Ramakrishnan
,
V.
(
2015
)
Structural basis for stop codon recognition in eukaryotes
.
Nature
524
,
493
496
25
Hoernes
,
T.P.
,
Clementi
,
N.
,
Juen
,
M.A.
,
Shi
,
X.
,
Faserl
,
K.
,
Willi
,
J.
et al. 
(
2018
)
Atomic mutagenesis of stop codon nucleotides reveals the chemical prerequisites for release factor-mediated peptide release
.
Proc. Natl Acad. Sci. U.S.A.
115
,
E382
E389
26
He
,
S.L.
and
Green
,
R.
(
2010
)
Visualization of codon-dependent conformational rearrangements during translation termination
.
Nat. Struct. Mol. Biol.
17
,
465
470
27
Sund
,
J.
,
Ander
,
M.
and
Aqvist
,
J.
(
2010
)
Principles of stop-codon reading on the ribosome
.
Nature
465
,
947
950
28
Ito
,
K.
,
Uno
,
M.
and
Nakamura
,
Y.
(
2000
)
A tripeptide ‘anticodon’ deciphers stop codons in messenger RNA
.
Nature
403
,
680
684
29
Young
,
D.J.
,
Edgar
,
C.D.
,
Murphy
,
J.
,
Fredebohm
,
J.
,
Poole
,
E.S.
and
Tate
,
W.P.
(
2010
)
Bioinformatic, structural, and functional analyses support release factor-like MTRF1 as a protein able to decode nonstandard stop codons beginning with adenine in vertebrate mitochondria
.
RNA
16
,
1146
1155
30
Young
,
D.J.
,
Edgar
,
C.D.
,
Poole
,
E.S.
and
Tate
,
W.P.
(
2010
)
The codon specificity of eubacterial release factors is determined by the sequence and size of the recognition loop
.
RNA
16
,
1623
1633
31
Young
,
D.
(
2009
)
The Recognition of Stop Codons by the Decoding Release Factors
,
University of Otago
,
Dunedin
32
Korkmaz
,
G.
and
Sanyal
,
S.
(
2017
)
R213i mutation in release factor 2 (RF2) is one step forward for engineering an omnipotent release factor in bacteria Escherichia coli
.
J. Biol. Chem.
292
,
15134
15142
33
Kobayashi
,
K.
,
Saito
,
K.
,
Ishitani
,
R.
,
Ito
,
K.
and
Nureki
,
O.
(
2012
)
Structural basis for translation termination by archaeal RF1 and GTP-bound EF1alpha complex
.
Nucleic Acids Res.
40
,
9319
9328
34
Belinky
,
F.
,
Babenko
,
V.N.
,
Rogozin
,
I.B.
and
Koonin
,
E.V.
(
2018
)
Purifying and positive selection in the evolution of stop codons
.
Sci. Rep.
8
,
9260
35
Song
,
H.
,
Mugnier
,
P.
,
Das
,
A.K.
,
Webb
,
H.M.
,
Evans
,
D.R.
,
Tuite
,
M.F.
et al. 
(
2000
)
The crystal structure of human eukaryotic release factor eRF1 — mechanism of stop codon recognition and peptidyl-tRNA hydrolysis
.
Cell
100
,
311
321
36
Seit-Nebi
,
A.
,
Frolova
,
L.
and
Kisselev
,
L.
(
2002
)
Conversion of omnipotent translation termination factor eRF1 into ciliate-like UGA-only unipotent eRF1
.
EMBO Rep.
3
,
881
886
37
Bulygin
,
K.N.
,
Khairulina
,
Y.S.
,
Kolosov
,
P.M.
,
Ven'yaminova
,
A.G.
,
Graifer
,
D.M.
,
Vorobjev
,
Y.N.
et al. 
(
2010
)
Three distinct peptides from the N domain of translation termination factor eRF1 surround stop codon in the ribosome
.
RNA
16
,
1902
1914
38
Inagaki
,
Y.
,
Blouin
,
C.
,
Doolittle
,
W.F.
and
Roger
,
A.J.
(
2002
)
Convergence and constraint in eukaryotic release factor 1 (eRF1) domain 1: the evolution of stop codon specificity
.
Nucleic Acids Res.
30
,
532
544
39
Bulygin
,
K.N.
,
Khairulina
,
Y.S.
,
Kolosov
,
P.M.
,
Ven'yaminova
,
A.G.
,
Graifer
,
D.M.
,
Vorobjev
,
Y.N.
et al. 
(
2011
)
Adenine and guanine recognition of stop codon is mediated by different N domain conformations of translation termination factor eRF1
.
Nucleic Acids Res.
39
,
7134
7146
40
Bulygin
,
K.N.
,
Graifer
,
D.M.
,
Hountondji
,
C.
,
Frolova
,
L.Y.
and
Karpova
,
G.G.
(
2017
)
Exploring contacts of eRF1 with the 3′-terminus of the P site tRNA and mRNA stop signal in the human ribosome at various translation termination steps
.
Biochim. Biophys. Acta, Gene Regul. Mech.
1860
,
782
793
41
Bulygin
,
K.N.
,
Bartuli
,
Y.S.
,
Malygin
,
A.A.
,
Graifer
,
D.M.
,
Frolova
,
L.Y.
and
Karpova
,
G.G.
(
2016
)
Chemical footprinting reveals conformational changes of 18S and 28S rRNAs at different steps of translation termination on the human ribosome
.
RNA
22
,
278
289
42
Vestergaard
,
B.
,
Van
,
L.B.
,
Andersen
,
G.R.
,
Nyborg
,
J.
,
Buckingham
,
R.H.
and
Kjeldgaard
,
M.
(
2001
)
Bacterial polypeptide release factor RF2 is structurally distinct from eukaryotic eRF1
.
Mol. Cell
8
,
1375
1382
43
Trappl
,
K.
and
Joseph
,
S.
(
2016
)
Ribosome induces a closed to open conformational change in release factor 1
.
J. Mol. Biol.
428
,
1333
1344
44
Gesteland
,
R.F.
,
Weiss
,
R.B.
and
Atkins
,
J.F.
(
1992
)
Recoding: reprogrammed genetic decoding
.
Science
257
,
1640
1641
45
Salser
,
W.
(
1969
)
The influence of the reading context upon the suppression of nonsense codons
.
Mol. Gen. Genet.
105
,
125
130
46
Skuzeski
,
J.M.
,
Nichols
,
L.M.
,
Gesteland
,
R.F.
and
Atkins
,
J.F.
(
1991
)
The signal for a leaky UAG stop codon in several plant viruses includes the two downstream codons
.
J. Mol. Biol.
218
,
365
373
47
Loughran
,
G.
,
Chou
,
M.Y.
,
Ivanov
,
I.P.
,
Jungreis
,
I.
,
Kellis
,
M.
,
Kiran
,
A.M.
et al. 
(
2014
)
Evidence of efficient stop codon readthrough in four mammalian genes
.
Nucleic Acids Res.
42
,
8928
8938
48
Svidritskiy
,
E.
,
Madireddy
,
R.
and
Korostelev
,
A.A.
(
2016
)
Structural basis for translation termination on a pseudouridylated stop codon
.
J. Mol. Biol.
428
,
2228
2236
49
Craigen
,
W.J.
,
Cook
,
R.G.
,
Tate
,
W.P.
and
Caskey
,
C.T.
(
1985
)
Bacterial peptide chain release factors: conserved primary structure and possible frameshift regulation of release factor 2
.
Proc. Natl Acad. Sci. U.S.A.
82
,
3616
3620
50
Zinoni
,
F.
,
Birkmann
,
A.
,
Leinfelder
,
W.
and
Bock
,
A.
(
1987
)
Cotranslational insertion of selenocysteine into formate dehydrogenase from Escherichia coli directed by a UGA codon
.
Proc. Natl Acad. Sci. U.S.A.
84
,
3156
3160
51
Berry
,
M.J.
,
Banu
,
L.
,
Chen
,
Y.Y.
,
Mandel
,
S.J.
,
Kieffer
,
J.D.
,
Harney
,
J.W.
et al. 
(
1991
)
Recognition of UGA as a selenocysteine codon in type I deiodinase requires sequences in the 3′ untranslated region
.
Nature
353
,
273
276
52
Tate
,
W.P.
and
Brown
,
C.M.
(
1992
)
Translational termination: ‘stop’ for protein synthesis or ‘pause’ for regulation of gene expression
.
Biochemistry
31
,
2443
2450
53
McCaughan
,
K.K.
,
Brown
,
C.M.
,
Dalphin
,
M.E.
,
Berry
,
M.J.
and
Tate
,
W.P.
(
1995
)
Translational termination efficiency in mammals is influenced by the base following the stop codon
.
Proc. Natl Acad. Sci. U.S.A.
92
,
5431
5435
54
Freistroffer
,
D.V.
,
Kwiatkowski
,
M.
,
Buckingham
,
R.H.
and
Ehrenberg
,
M.
(
2000
)
The accuracy of codon recognition by polypeptide release factors
.
Proc. Natl Acad. Sci. U.S.A.
97
,
2046
2051
55
Caskey
,
C.T.
,
Beaudet
,
A.L.
,
Scolnick
,
E.M.
and
Rosman
,
M.
(
1971
)
Hydrolysis of fMet-tRNA by peptidyl transferase
.
Proc. Natl Acad. Sci. U.S.A.
68
,
3163
3167
56
Caskey
,
C.T.
,
Beaudet
,
A.L.
and
Tate
,
W.P.
(
1974
)
Mammalian release factor; in vitro assay and purification
.
Methods Enzymol.
30
,
293
303
57
Poole
,
E.S.
,
Brown
,
C.M.
and
Tate
,
W.P.
(
1995
)
The identity of the base following the stop codon determines the efficiency of in vivo translational termination in Escherichia coli
.
EMBO J.
14
,
151
158
58
Major
,
L.L.
,
Poole
,
E.S.
,
Dalphin
,
M.E.
,
Mannering
,
S.A.
and
Tate
,
W.P.
(
1996
)
Is the in-frame termination signal of the Escherichia coli release factor-2 frameshift site weakened by a particularly poor context?
Nucleic Acids Res.
24
,
2673
2678
59
Brown
,
C.M.
,
Stockwell
,
P.A.
,
Trotman
,
C.N.
and
Tate
,
W.P.
(
1990
)
The signal for the termination of protein synthesis in procaryotes
.
Nucleic Acids Res.
18
,
2079
2086
60
Berry
,
M.J.
,
Banu
,
L.
and
Larsen
,
P.R.
(
1991
)
Type I iodothyronine deiodinase is a selenocysteine-containing enzyme
.
Nature
349
,
438
440
61
Kohli
,
J.
and
Grosjean
,
H.
(
1981
)
Usage of the three termination codons: compilation and analysis of the known eukaryotic and prokaryotic translation termination sequences
.
Mol. Gen. Genet.
182
,
430
439
62
Dalphin
,
M.E.
,
Brown
,
C.M.
,
Stockwell
,
P.A.
and
Tate
,
W.P.
(
1996
)
Transterm: a database of translational signals
.
Nucleic Acids Res.
24
,
216
218
63
Jacobs
,
G.H.
,
Stockwell
,
P.A.
,
Tate
,
W.P.
and
Brown
,
C.M.
(
2006
)
Transterm--extended search facilities and improved integration with other databases
.
Nucleic Acids Res.
34
,
D37
D40
64
Jacobs
,
G.H.
,
Chen
,
A.
,
Stevens
,
S.G.
,
Stockwell
,
P.A.
,
Black
,
M.A.
,
Tate
,
W.P.
et al. 
(
2009
)
Transterm: a database to aid the analysis of regulatory sequences in mRNAs
.
Nucleic Acids Res.
37
,
D72
D76
65
Brown
,
C.M.
,
Stockwell
,
P.A.
,
Trotman
,
C.N.A.
and
Tate
,
W.P.
(
1990
)
Sequence analysis suggests that tetra-nucleotides signal the termination of protein synthesis in eukaryotes
.
Nucleic Acids Res.
18
,
6339
6345
66
Wei
,
Y.
and
Xia
,
X.
(
2017
)
The role of +4U as an extended translation termination signal in bacteria
.
Genetics
205
,
539
549
67
Bonetti
,
B.
,
Fu
,
L.
,
Moon
,
J.
and
Bedwell
,
D.M.
(
1995
)
The efficiency of translation termination is determined by a synergistic interplay between upstream and downstream sequences in Saccharomyces cerevisiae
.
J. Mol. Biol.
251
,
334
345
68
Namy
,
O.
,
Hatin
,
I.
and
Rousset
,
J.P.
(
2001
)
Impact of the six nucleotides downstream of the stop codon on translation termination
.
EMBO Rep.
2
,
787
793
69
Mottagui-Tabar
,
S.
,
Tuite
,
M.F.
and
Isaksson
,
L.A.
(
1998
)
The influence of 5′ codon context on translation termination in Saccharomyces cerevisiae
.
Eur. J. Biochem.
257
,
249
254
70
Tork
,
S.
,
Hatin
,
I.
,
Rousset
,
J.P.
and
Fabret
,
C.
(
2004
)
The major 5′ determinant in stop codon read-through involves two adjacent adenines
.
Nucleic Acids Res.
32
,
415
421
71
Mottagui-Tabar
,
S.
and
Isaksson
,
L.A.
(
1998
)
The influence of the 5′ codon context on translation termination in Bacillus subtilis and Escherichia coli is similar but different from Salmonella typhimurium
.
Gene
212
,
189
196
72
Shao
,
S.
,
Murray
,
J.
,
Brown
,
A.
,
Taunton
,
J.
,
Ramakrishnan
,
V.
and
Hegde
,
R.S.
(
2016
)
Decoding mammalian ribosome-mRNA states by translational GTPase complexes
.
Cell
167
,
1229
1240.e15
73
Tang
,
X.
,
Zhu
,
Y.
,
Baker
,
S.L.
,
Bowler
,
M.W.
,
Chen
,
B.J.
,
Chen
,
C.
et al. 
(
2016
)
Structural basis of suppression of host translation termination by Moloney Murine Leukemia Virus
.
Nat. Commun.
7
,
12070
74
Cridge
,
A.G.
,
Major
,
L.L.
,
Mahagaonkar
,
A.A.
,
Poole
,
E.S.
,
Isaksson
,
L.A.
and
Tate
,
W.P.
(
2006
)
Comparison of characteristics and function of translation termination signals between and within prokaryotic and eukaryotic organisms
.
Nucleic Acids Res.
34
,
1959
1973
75
Jenner
,
L.B.
,
Demeshkina
,
N.
,
Yusupova
,
G.
and
Yusupov
,
M.
(
2010
)
Structural aspects of messenger RNA reading frame maintenance by the ribosome
.
Nat. Struct. Mol. Biol.
17
,
555
560
76
Cridge
,
A.G.
,
Crowe-McAuliffe
,
C.
,
Mathew
,
S.F.
and
Tate
,
W.P.
(
2018
)
Eukaryotic translational termination efficiency is influenced by the 3′ nucleotides within the ribosomal mRNA channel
.
Nucleic Acids Res.
46
,
1927
1944
77
Martitz
,
J.
,
Hofmann
,
P.J.
,
Johannes
,
J.
,
Kohrle
,
J.
,
Schomburg
,
L.
and
Renko
,
K.
(
2016
)
Factors impacting the aminoglycoside-induced UGA stop codon readthrough in selenoprotein translation
.
J. Trace Elem. Med. Biol.
37
,
104
110
78
Dabrowski
,
M.
,
Bukowy-Bieryllo
,
Z.
and
Zietkiewicz
,
E.
(
2015
)
Translational readthrough potential of natural termination codons in Eukaryotes — the impact of RNA sequence
.
RNA Biol.
12
,
950
958
79
Johnson
,
D.B.
,
Xu
,
J.
,
Shen
,
Z.
,
Takimoto
,
J.K.
,
Schultz
,
M.D.
,
Schmitz
,
R.J.
et al. 
(
2011
)
RF1 knockout allows ribosomal incorporation of unnatural amino acids at multiple sites
.
Nat. Chem. Biol.
7
,
779
786
80
Chowdhury
,
H.M.
,
Siddiqui
,
M.A.
,
Kanneganti
,
S.
,
Sharmin
,
N.
,
Chowdhury
,
M.W.
and
Nasim
,
M.T.
(
2018
)
Aminoglycoside-mediated promotion of translation readthrough occurs through a non-stochastic mechanism that competes with translation termination
.
Hum. Mol. Genet.
27
,
373
384
81
Baradaran-Heravi
,
A.
,
Niesser
,
J.
,
Balgi
,
A.D.
,
Choi
,
K.
,
Zimmerman
,
C.
,
South
,
A.P.
et al. 
(
2017
)
Gentamicin B1 is a minor gentamicin component with major nonsense mutation suppression activity
.
Proc. Natl Acad. Sci. U.S.A.
114
,
3479
3484
82
Roy
,
B.
,
Friesen
,
W.J.
,
Tomizawa
,
Y.
,
Leszyk
,
J.D.
,
Zhuo
,
J.
,
Johnson
,
B.
et al. 
(
2016
)
Ataluren stimulates ribosomal selection of near-cognate tRNAs to promote nonsense suppression
.
Proc. Natl Acad. Sci. U.S.A.
113
,
12508
12513
83
Zahonova
,
K.
,
Kostygov
,
A.Y.
,
Sevcikova
,
T.
,
Yurchenko
,
V.
and
Elias
,
M.
(
2016
)
An unprecedented non-canonical nuclear genetic code with all three termination codons reassigned as sense codons
.
Curr. Biol.
26
,
2364
2369
84
Swart
,
E.C.
,
Serra
,
V.
,
Petroni
,
G.
and
Nowacki
,
M.
(
2016
)
Genetic codes with no dedicated stop codon: context-dependent translation termination
.
Cell
166
,
691
702
85
Panek
,
T.
,
Zihala
,
D.
,
Sokol
,
M.
,
Derelle
,
R.
,
Klimes
,
V.
,
Hradilova
,
M.
et al. 
(
2017
)
Nuclear genetic codes with a different meaning of the UAG and the UAA codon
.
BMC Biol.
15
,
8
86
Wang
,
D.
,
Fan
,
F.
,
Li
,
Z.
,
Liu
,
X.
,
Song
,
S.
,
Wei
,
S.
et al. 
(
2017
)
Stop codon mutagenesis for homogenous expression of human papillomavirus L1 protein in Escherichia coli
.
Protein Expr. Purif.
133
,
110
120
87
Wei
,
Y.
,
Wang
,
J.
and
Xia
,
X.
(
2016
)
Coevolution between stop codon usage and release factors in bacterial species
.
Mol. Biol. Evol.
33
,
2357
2367
88
Urakov
,
V.N.
,
Mitkevich
,
O.V.
,
Safenkova
,
I.V.
and
Ter-Avanesyan
,
M.D.
(
2017
)
Ribosome-bound Pub1 modulates stop codon decoding during translation termination in yeast
.
FEBS J.
284
,
1914
1930
89
Merritt
,
G.H.
,
Naemi
,
W.R.
,
Mugnier
,
P.
,
Webb
,
H.M.
,
Tuite
,
M.F.
and
von der Haar
,
T.
(
2010
)
Decoding accuracy in eRF1 mutants and its correlation with pleiotropic quantitative traits in yeast
.
Nucleic Acids Res.
38
,
5479
5492
90
Nyikó
,
T.
,
Auber
,
A.
,
Szabadkai
,
L.
,
Benkovics
,
A.
,
Auth
,
M.
,
Mérai
,
Z.
et al. 
(
2017
)
Expression of the eRF1 translation termination factor is controlled by an autoregulatory circuit involving readthrough and nonsense-mediated decay in plants
.
Nucleic Acids Res.
45
,
4174
4188
91
Jungreis
,
I.
,
Lin
,
M.F.
,
Spokony
,
R.
,
Chan
,
C.S.
,
Negre
,
N.
,
Victorsen
,
A.
et al. 
(
2011
)
Evidence of abundant stop codon readthrough in Drosophila and other metazoa
.
Genome Res.
21
,
2096
2113
92
Hofhuis
,
J.
,
Schueren
,
F.
,
Notzel
,
C.
,
Lingner
,
T.
,
Gartner
,
J.
,
Jahn
,
O.
et al. 
(
2016
)
The functional readthrough extension of malate dehydrogenase reveals a modification of the genetic code
.
Open Biol.
6
,
160246
93
Loughran
,
G.
,
Jungreis
,
I.
,
Tzani
,
I.
,
Power
,
M.
,
Dmitriev
,
R.I.
,
Ivanov
,
I.P.
et al. 
(
2018
)
Stop codon readthrough generates a C-terminally extended variant of the human vitamin D receptor with reduced calcitriol response
.
J. Biol. Chem.
293
,
4434
4444
94
Alkalaeva
,
E.
,
Eliseev
,
B.
,
Ambrogelly
,
A.
,
Vlasov
,
P.
,
Kondrashov
,
F.A.
,
Gundllapalli
,
S.
et al. 
(
2009
)
Translation termination in pyrrolysine-utilizing archaea
.
FEBS Lett.
583
,
3455
3460