Advertisement

mRNA and tRNA modification states influence ribosome speed and frame maintenance during poly(lysine) peptide synthesis

Open AccessPublished:May 17, 2022DOI:https://doi.org/10.1016/j.jbc.2022.102039
      Ribosome speed is dictated by multiple factors including substrate availability, cellular conditions, and product (peptide) formation. Translation slows during the synthesis of cationic peptide sequences, potentially influencing the expression of thousands of proteins. Available evidence suggests that ionic interactions between positively charged nascent peptides and the negatively charged ribosome exit tunnel impede translation. However, this hypothesis was difficult to test directly because of inability to decouple the contributions of amino acid charge from mRNA sequence and tRNA identity/abundance in cells. Furthermore, it is unclear if other components of the translation system central to ribosome function (e.g., RNA modification) influence the speed and accuracy of positively charged peptide synthesis. In this study, we used a fully reconstituted Escherichia coli translation system to evaluate the effects of peptide charge, mRNA sequence, and RNA modification status on the translation of lysine-rich peptides. Comparison of translation reactions on poly(lysine)-encoding mRNAs conducted with either Lys-tRNALys or Val-tRNALys reveals that that amino acid charge, while important, only partially accounts for slowed translation on these transcripts. We further find that in addition to peptide charge, mRNA sequence and both tRNA and mRNA modification status influence the rates of amino acid addition and the ribosome’s ability to maintain frame (instead of entering the −2, −1, and +1 frames) during poly(lysine) peptide synthesis. Our observations lead us to expand the model for explaining how the ribosome slows during poly(lysine) peptide synthesis and suggest that posttranscriptional RNA modifications can provide cells a mechanism to precisely control ribosome movements along an mRNA.

      Keywords

      Abbreviations:

      aatRNAs (aminoacyl-tRNA), eTLC (electrophoretic TLC), IC (initiation complex), m6A (N6-methyladenosine), TC (ternary complex)
      Translation of the nucleic acid code into protein is catalyzed by the ribosome. During this process, ribosomes use mRNAs as molecular maps to direct the programmed assembly of amino acids into polypeptides. In Escherichia coli growing polypeptide chains are extended by an average of 4 to 22 amino acids per second, though the rate of individual amino acid incorporation by the ribosome is not always uniform (
      • Wohlgemuth I.
      • Pohl C.
      • Mittelstaet J.
      • Konevega A.L.
      • Rodnina M.V.
      Evolutionary optimization of speed and accuracy of decoding on the ribosome.
      ,
      • Pedersen S.
      Escherichia coli ribosomes translate in vivo with variable rate.
      ,
      • Siller E.
      • DeZwaan D.C.
      • Anderson J.F.
      • Freeman B.C.
      • Barral J.M.
      Slowing bacterial translation speed enhances eukaryotic protein folding efficiency.
      ). Heterogeneity in peptide elongation rates is caused by a number of factors, including substrate (e.g., mRNA, aminoacyl-tRNAs [aatRNAs], translation factors) availability and modification status, and the formation of stable interactions between the growing polypeptide chain and the ribosome machinery (
      • Qian W.
      • Yang J.-R.
      • Pearson N.M.
      • Maclean C.
      • Zhang J.
      Balanced codon usage optimizes eukaryotic translational efficiency.
      ,
      • Ban N.
      • Nissen P.
      • Hansen J.
      • Moore P.B.
      • Steitz T.A.
      The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution.
      ). Although translation initiation rates are responsible for controlling the rate of protein expression in many circumstances, situations that alter polypeptide elongation rates can change protein levels, protein folding, and mRNA stability to ultimately impact cellular health and fitness (
      • Jacobs W.M.
      • Shakhnovich E.I.
      Evidence of evolutionary selection for cotranslational folding.
      ,
      • Hanson G.
      • Alhusaini N.
      • Morris N.
      • Sweet T.
      • Coller J.
      Translation elongation and mRNA stability are coupled through the ribosomal A-site.
      ,
      • Presnyak V.
      • Alhusaini N.
      • Chen Y.-H.
      • Martin S.
      • Morris N.
      • Kline N.
      • et al.
      Codon optimality is a major determinant of mRNA stability.
      ).
      Contacts between the ribosome and its nascent peptide products are receiving growing recognition for their role in translationally controlling protein expression (
      • Choi J.
      • Grosely R.
      • Prabhakar A.
      • Lapointe C.P.
      • Wang J.
      • Puglisi J.D.
      How mRNA and nascent chain sequences regulate translation elongation.
      ). The interactions involving the ribosome and positively charged peptides present a classic example of this phenomenon. There is overwhelming evidence that translating the cationic peptide sequences commonly present in proteins slows the ribosome (
      • Lu J.
      • Deutsch C.
      Electrostatics in the ribosomal tunnel modulate chain elongation rates.
      ,
      • Brandman O.
      • Stewart-Ornstein J.
      • Wong D.
      • Larson A.
      • Williams C.C.
      • Li G.-W.
      • et al.
      A ribosome-bound quality control complex triggers degradation of nascent peptides and signals translation stress.
      ). In humans, there are over 60,000 examples of proteins containing four or more consecutive basic amino acids, suggesting that the synthesis of positively charged peptides contributes to the posttranscriptional control of a significant fraction of the proteome (Tables S1 and S2). The observation that ribosomes slow while linking iterated positively charged amino acids has long been attributed to the formation of strong ionic interactions between cationic peptides and the anionic ribosome peptide exit channel (
      • Charneski C.A.
      • Hurst L.D.
      Positively charged residues are the major determinants of ribosomal velocity.
      ). However, several recent reports demonstrate that the ribosome produces different amounts of protein from mRNAs possessing synonymous codons that encode identical positively charged poly(lysine) and poly(arginine) peptides (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ,
      • Arthur L.L.
      • Pavlovic-Djuranovic S.
      • Koutmou K.S.
      • Green R.
      • Szczesny P.
      • Djuranovic S.
      Translational control by lysine-encoding A-rich sequences.
      ,
      • Rosenberg A.H.
      • Goldman E.
      • Dunn J.J.
      • Studier F.W.
      • Zubay G.
      Effects of consecutive AGG codons on translation in Escherichia coli, demonstrated with a versatile codon test system.
      ). Ionic interactions alone cannot explain these findings, suggesting that additional factors in the translation system also contribute to modulating ribosome speed during the synthesis of cationic peptides.
      Multiple codons instruct the ribosome to add the positively charged amino acids lysine (AAA and AAG) and arginine (AGA, AGG, and CGN [N=U, C, A,G]). The ability of individual arginine codons to differentially impact protein expression largely depends on the availability of tRNA isoacceptors possessing appropriate anticodon sequences. Some isoacceptors are less abundant, and translation along mRNA sequences containing multiple codons corresponding to these rare tRNAs can slow sufficiently to trigger cellular mechanisms that rescue stalled ribosomes (
      • Sørensen M.A.
      • Kurland C.G.
      • Pedersen S.
      Codon usage determines translation rate in Escherichia coli.
      ,
      • Liu Y.
      • Yang Q.
      • Zhao F.
      Synonymous but not silent: the codon usage code for gene expression and protein folding.
      ). The cause of differential protein expression from the two lysine codons appears to differ from that of arginine codons. In the case of lysine, less protein is produced from mRNAs containing consecutive AAA codons than those with consecutive AAG codons in both eukaryotic and bacterial cells (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ,
      • Arthur L.L.
      • Pavlovic-Djuranovic S.
      • Koutmou K.S.
      • Green R.
      • Szczesny P.
      • Djuranovic S.
      Translational control by lysine-encoding A-rich sequences.
      ,
      • Ito-Harashima S.
      • Kuroha K.
      • Tatematsu T.
      • Inada T.
      Translation of the poly(A) tail plays crucial roles in nonstop mRNA surveillance via translation repression and protein destabilization by proteasome in yeast.
      ,
      • Chiabudini M.
      • Tais A.
      • Zhang Y.
      • Hayashi S.
      • Wölfle T.
      • Fitzke E.
      • et al.
      Release factor eRF3 mediates premature translation termination on polylysine-stalled ribosomes in Saccharomyces cerevisiae.
      ). However, differences in substrate tRNA levels are unlikely to account for these codon-specific observations because AAA and AAG are decoded by a single tRNALys in at least one of the species (E. coli) where codon-dependent differences in poly(lysine) protein output have been observed. Furthermore, in addition to reducing the rate of protein synthesis, under some conditions the presence of two or more AAA codons in a row can promote an unusual ribosome movement termed “ribosome sliding” (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ). During sliding, the ribosome loses reading frame and shifts along an mRNA. The ribosome has been captured moving backward by 1 to 3 nucleotides while translating iterated AAA codons (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ,
      • Arthur L.L.
      • Pavlovic-Djuranovic S.
      • Koutmou K.S.
      • Green R.
      • Szczesny P.
      • Djuranovic S.
      Translational control by lysine-encoding A-rich sequences.
      ), changing the identity of the peptide being made. These movements activate cotranslational surveillance mechanisms that target the translated mRNA and resulting peptide products for degradation (
      • Ito-Harashima S.
      • Kuroha K.
      • Tatematsu T.
      • Inada T.
      Translation of the poly(A) tail plays crucial roles in nonstop mRNA surveillance via translation repression and protein destabilization by proteasome in yeast.
      ). Ribosome sliding differs from other noncanonical ribosome movements, which place the ribosome at a single, discrete location on an mRNA and can produce stable products (
      • Agirrezabala X.
      • Samatova E.
      • Klimova M.
      • Zamora M.
      • Gil-Carton D.
      • Rodnina M.V.
      • et al.
      Ribosome rearrangements at the onset of translational bypassing.
      ,
      • Dinman J.D.
      Mechanisms and implications of programmed translational frameshifting.
      ).These data suggest that the influence of mRNA and tRNA sequences on the translation of poly(lysine) peptide regions warrant further examination.
      Here, we use a reconstituted E. coli translation system to deconvolute the contributions of peptide, mRNA sequence, and RNA (mRNA and tRNA) modification to both the speed of amino acid addition and ribosome frame maintenance during the translation of iterated lysine codons. We chose to investigate the role of RNA modifications in addition to peptide and mRNA sequence because these common chemical changes to nucleosides can alter the hydrogen-bonding interactions between tRNAs and mRNAs used by the ribosome to ensure the faithful and rapid translation of the genetic code into protein (
      • Jones J.D.
      • Monroe J.
      • Koutmou K.S.
      A molecular-level perspective on the frequency, distribution, and consequences of messenger RNA modifications.
      ). Our findings expand the biochemical framework for understanding the contributions of individual components of the translation system to ribosome stalling during cationic peptide synthesis. We demonstrate that in addition to peptide charge, mRNA sequence, along with mRNA and tRNA modification status, are important determinants of ribosome speed during poly(lysine) translation. Additionally, we developed a minimal kinetic mechanism for ribosome sliding on iterated AAA codons, in which the ribosome moves along an mRNA in the 3′ direction one nucleotide at a time, until it can bind an available cognate aatRNA and resume “normal” translation in a different frame. Much like ribosome speed, this series of one nucleotide ribosome movements is controlled not only by peptide charge but also by posttranscriptional modifications to tRNALys and mRNA (N6-methyladenosine [m6A]). While it has been known for decades that tRNA anticodon stem-loop modifications can influence ribosome movements, these data provide the first evidence that mRNA modifications also have the power to impact ribosome reading frame maintenance (
      • Ranjan N.
      • Rodnina M.V.
      tRNA wobble modifications and protein homeostasis.
      ). Our work presents a molecular level rationalization for how seemingly small changes in the translational machinery (e.g., synonymous codon substitution and single posttranscriptional modifications) can result in different protein production outcomes.

      Results

      Ribosomes move backward one nucleotide at a time on poly(A) sequences

      There are multiple ways to envision how ribosome sliding on consecutive AAA codons could be achieved. For example, the ribosome might hop directly into the −1 and −3 frames, “scan” along an mRNA until it reaches a specific, desired frame, or make series of discrete one nucleotide frameshifts (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ,
      • Arthur L.L.
      • Pavlovic-Djuranovic S.
      • Koutmou K.S.
      • Green R.
      • Szczesny P.
      • Djuranovic S.
      Translational control by lysine-encoding A-rich sequences.
      ,
      • Tsai T.-W.
      • Yang H.
      • Yin H.
      • Xu S.
      • Wang Y.
      High-efficiency “–1” and “–2” ribosomal frameshiftings revealed by force spectroscopy.
      ,
      • Yan S.
      • Wen J.-D.
      • Bustamante C.
      • Tinoco I.
      Ribosome excursions during mRNA translocation mediate broad branching of frameshift pathways.
      ). We developed a kinetic framework to distinguish between these possibilities and describe how the ribosome moves during sliding using a fully reconstituted in vitro translation system (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ). To accomplish this, we first identified the reading frames the ribosome enters during translation along an mRNA with a AUG-AAA-AAA-UUC-UAA (MK2(AAA)FX; X=stop codon) coding sequence (Fig. 1A). In these assays, 70 nM of E. coli 70S ribosome initiation complexes (ICs) containing 35S-labeled formylmethionine-tRNAfMet bound to an AUG in the P site and an AAA codon in the A site were reacted with saturating concentrations of two ternary complexes (TCs, 10–30 μM; aatRNAaa⋅EF-Tu⋅GTP) formed with elongation factor-Tu (EF-Tu) as well as cognate Lys-tRNALys and individual aatRNAaa species capable of reacting in each of the reading frames that the ribosome could inhabit on our MK2(AAA)FX mRNA (−2 frame, Asn-tRNAAsn; −1 frame, Ile-tRNAIle; 0 frame, Phe-tRNAPhe; +1 frame, Ser-tRNASer; and +2 frame, Leu-tRNALeu) (Fig. 1A). These reactions were conducted in the presence of 0 to 12 μM of elongation factor G GTPase bound to GTP (EF-G⋅GTP) at 37 °C. The reactants (fMet), programmed peptides (MK, MK2, MK2F), and peptide products resulting from the ribosome sliding (MK2I, MK2N, MK2S, MK2L, MK3, and MK4+) were visualized by electrophoretic TLC (eTLC; Fig. 1B). In our assays, the ribosome generated products in five different reading frames (0, −1, −2, −3, and +1; Fig. 1, AC). We find that movement of the ribosome into non-0 frames is EF-G dependent, much like canonical-1 frameshifting ribosome movements (Fig. S1) (
      • Tsai T.-W.
      • Yang H.
      • Yin H.
      • Xu S.
      • Wang Y.
      High-efficiency “–1” and “–2” ribosomal frameshiftings revealed by force spectroscopy.
      ,
      • Kim H.-K.
      • Tinoco I.
      EF-G catalyzed translocation dynamics in the presence of ribosomal frameshifting stimulatory signals.
      ,
      • Klimova M.
      • Senyushkina T.
      • Samatova E.
      • Peng B.Z.
      • Pearson M.
      • Peske F.
      • et al.
      EF-G–induced ribosome sliding along the noncoding mRNA.
      ,
      • Chen J.
      • Petrov A.
      • Johansson M.
      • Tsai A.
      • O'Leary S.E.
      • Puglisi J.D.
      Dynamic pathways of -1 translational frameshifting.
      ).
      Figure thumbnail gr1
      Figure 1The ribosome moves into multiple frames on poly(A) in absence of next cognate aatRNAaa. A, possible ribosome-sliding events on MK2(AAA)FX mRNA result in new discrete codons positioned in the A-site, allowing for decoding and accommodation of noncoded aatRNAaa. B, phosphorimage of a eTLC displaying the products of 20 min end-point reactions investigating frameshifting during ribosome sliding on MK2(AAA)FX or MK2(AAG)FX mRNA incubated with Lys-tRNALys and various aatRNAaa TCs. Lanes indicate the aatRNAaa TCs used in each reaction, with the numbers correlating to the amino acid identity indicated in panel (A) (e.g., −2 indicates reaction performed with Lys-tRNALys and Asn-tRNAAsn). Θ indicates the amino acid that would be added upon successful frameshift and incorporation of aatRNAaa as indicated per lane and in panel (A) (e.g., bands for MK2Θ and MK3Θ in −2 lane correspond to the peptides MK2N and MK3N, respectively). C, percent of peptide product formed due to sliding/frameshifting compared to total peptide synthesized during translation assays, with frame and third amino acid added as signified. Error bars represent SD. D, rate constants for frameshift events during ribosome sliding on MK2(AAA)FX mRNA using Lys-tRNALys and aatRNAaa TCs as specified with frame (panel A), as defined by the proposed mechanism in . Error bars represent SD. There is no difference in the rate constants for amino acid addition in the −1 (Ile; k7I), −2 (Asn; k7N), and −3 (Lys; k3) frames. For +1 (Ser; k7S) and +2 (Leu; k7L) rate constants, a ‘ ∗ ’ represented a significant alteration with a p-value <0.05 using a unpaired student t test when compared to the −3 rate constant (Lys; k3). aatRNAs, aminoacyl-tRNA; eTLC, electrophoretic TLC.
      After establishing which frames the ribosome inhabits during sliding, we measured the extent of amino acid incorporation in each of these frames at discrete time points (0–1200 s, Figs. 1C and S2). These data were used to develop a minimal kinetic mechanism for ribosome sliding by globally fitting our experimental observations with KinTek Explorer (KinTek Corporation; https://kintekexplorer.com/) (Fig. S2). We examined a series of possible mechanisms (Fig. S3) and selected the model that best fit our data to ascertain the rate constants for each step in the mechanism. Our fits indicate that the ribosome undergoes a series of progressive −1 nucleotide movements from the 0-frame into the −1, −2, and −3 frames during ribosome sliding (Figs. 2 and S4). The rate constants for amino acid addition in the −1, −2, and −3 frames are relatively uniform (k7 values range between 0.01–0.02 s−1) (Fig. 1D and Table 1). Furthermore, a subset of ribosomes (∼10%) appear unable to extend the growing polypeptide following each progressive −1 nucleotide ribosome movement (Figs. 1B and 2 (k5, k4:6,obs), Figs. S3 and S5B). While we do observe a small amount of product formation in the +1 frame, both the endpoint and rate constant for this reaction are diminished relative to the same values for the ribosome reacting in the −1, −2, and −3 frames (Fig. 1, C and D). These observations lead us to propose a model for ribosome sliding in which a small (<5% of ribosomes) can undergo a +1 frameshift, while most ribosomes move in the 3′ direction by one nucleotide a time until they enter a reading frame that can react with an available aatRNA species (Fig. 2).
      Figure thumbnail gr2
      Figure 2Proposed general scheme for frameshift events during ribosome sliding on poly(A). In the case where available aatRNAs are incorporated via frameshift, denoted by Θ, this scheme describes subsequent amino acid additions by a ribosome translating on a poly(A) containing mRNA—as displayed in . The scheme contains parameters obtainable from the experiments presented here: dipeptide formation (k1), tripeptide formation (k2), ribosome sliding, and frameshift events (k3, first −3 sliding/frameshift event generating tetrapeptide [MK3]; k5, first sliding/frameshift event resulting in unproductive ribosome(s); k7Θ, first sliding/frameshift of ribosome moving into new coding frame [MK2Θ]), and secondary/tertiary sliding events capable of occurring following first sliding/frameshift event (k4:6,obs). aatRNAs, aminoacyl-tRNA.
      Table 1Rate constants for frameshift and amino acid addition during ribosome sliding
      Frameshift position0 frame−2 frame−1 frame+1 frame+2 frame
      tRNA TCsLys + PheLys + AsnLys + IleLys + SerLys + Leu
      Rate constants (s−1)
      k111.7 ± 0.0113.1 ± 0.0115.3 ± 0.812.7 ± 0.713.3 ± 0.7
      k21.5 ± 0.0021.6 ± 0.0021.4 ± 0.21.5 ± 0.151.2 ± 0.3
      k3-0.005 ± 0.00020.008 ± 0.00150.008 ± 0.0010.02 ± 0.004
      k4:6,obs-0.006 ± 0.0020.003 ± 0.0010.007 ± 0.0030.002 ± 0.009
      k5-0.0008 ± 0.000040.0007 ± 0.00030.002 ± 0.00050.017 ± 0.009
      k70.9 ± 0.00070.01 ± 0.00090.02 ± 0.0060.002 ± 0.00070.0002 ± 0.0001

      tRNALys modifications moderate ribosome sliding during poly-lysine synthesis

      Native tRNAs possess posttranscriptional chemical modifications essential to their stability, structure, and function (
      • Lorenz C.
      • Lünse C.E.
      • Mörl M.
      tRNA modifications: impact on structure and thermal adaptation.
      ,
      • Agris P.F.
      • Narendran A.
      • Sarachan K.
      • Väre V.Y.P.
      • Eruysal E.
      The role of RNA modifications in translational fidelity.
      ). Modifications located in tRNA anticodon stem loops have the capacity to modulate −1 and +1 ribosomal frameshifts and enhance ribosome reading frame maintenance (
      • Bekaert M.
      • Rousset J.-P.
      An extended signal involved in eukaryotic -1 frameshifting operates through modification of the E site tRNA.
      ,
      • Maehigashi T.
      • Dunkle J.A.
      • Miles S.J.
      • Dunham C.M.
      Structural insights into +1 frameshifting promoted by expanded or modification-deficient anticodon stem loops.
      ). Since the tRNALys N6-threonylcarbamoyladenosine (t6A37) and 5-methoxycarbonylmethyl2-thiouridine [mcm5S2U34] modifications in yeast tRNALys,UUU influence tRNA decoding, we speculated that analogous E. coli tRNALys modifications, such as 5-methylaminocarbonylmethyluridine (mnm5s2U34), might suppress ribosome sliding and enhance poly(lysine) translation (
      • Rezgui V.A.N.
      • Tyagi K.
      • Ranjan N.
      • Konevega A.L.
      • Mittelstaet J.
      • Rodnina M.V.
      • et al.
      tRNA tKUUU, tQUUG, and tEUUC wobble position modifications fine-tune protein translation by promoting ribosome A-site binding.
      ,
      • Ranjan N.
      • Rodnina M.V.
      Thio-modification of tRNA at the wobble position as regulator of the kinetics of decoding and translocation on the ribosome.
      ). To test this idea, we compared the rate constants for lysine addition during the translation of AUG-AAA-AAA-UUC-UAA [MK2(AAA)FX] and AUG-AAG-AAG-UUC-UAA [MK2(AAG)FX] messages using saturating levels (20–30 μM) of unmodified T7 transcribed Lys-tRNALys and natively modified Lys-tRNANLys purified from E. coli cells (Figs. 3A and S6). We find that the rate constants for programmed MK and MK2 peptide formation are twofold to fourfold faster when ribosome complexes are reacted with TCs containing modified Lys-tRNANLys than with unmodified Lys-tRNALys. These moderate enhancements in lysine addition rate constants are observed when either AAG or AAA containing mRNAs are translated (Table 2). The modifications have a larger role on frame maintenance than on programmed lysine addition; the rate constants for forming ribosome sliding products (MK3 and MK4+) on AAA codons are decreased by up to 25-fold when natively modified Lys-tRNANLys is used (Fig. 3B). Despite the slowed formation of these sliding products, the percentage of peptides that are eventually extended and generated sliding products is only slightly reduced by the inclusion of modifications on tRNALys (Lys-tRNALys = 60% ± 3% versus Lys-tRNANLys = 40% ± 6%) (Figs. 3, B and C and S6). Our data suggest that tRNANLys modifications likely limit the extent of ribosome sliding in cells.
      Figure thumbnail gr3
      Figure 3Modifications on tRNALys regulate extent of frame loss on poly(A) mRNA. A, rate constants for ribosome sliding on MK2(AAA)FX mRNA using either Lys-tRNALys or Lys-tRNANLys TCs, as defined by the proposed mechanism in A. Error bars represent SD. B, percent of total peptide formed as a result of ribosome sliding (peptides longer than MK2 tripeptide) on MK2(AAA)FX mRNA after 20 min in translation assays conducted with Lys-tRNALys or Lys-tRNANLys TCs. Error bars represent SD. C, phosphorimage eTLCs of of ribosome sliding over time on MK2(AAA)FX mRNA in translation reactions perfomred with either Lys-tRNALys TCs (left) or Lys-tRNANLys TCs (right). eTLC, electrophoretic TLC; TC, ternary complex.
      Table 2Rate constants for lysine addition during ribosome sliding
      mRNA constructMK2(AAA)FXMK2(AAG)FX
      tRNA TCsLys-tRNALysLys-tRNANLysLys-tRNALysLys-tRNANLys
      Rate constants (s−1)
      k110.2 ± 0.328.9 ± 1.92.7 ± 0.512.7 ± 0.4
      k21.2 ± 0.44.8 ± 0.42.1 ± 0.74.4 ± 0.3
      k30.02 ± 0.0040.0008 ± 0.0000080.006 ± 0.0020.002 ± 0.001
      k4 (or k4,obs)0.007 ± 0.0010.07 ± 0.020.0004 ± 0.00010.00009 ± 0.00002
      k50.01 ± 0.0080.0001 ± 0.000050.001 ± 0.00010.001 ± 0.0002
      k60.006 ± 0.0020.02 ± 0.007--

      Ribosomes slow and slide when synthesizing poly(valine) peptides from lysine-encoding mRNAs

      Regardless of tRNALys modification status, we observed that after the first lysine is added into a peptide, the rate constants for adding subsequent lysines on AAA codons are reduced (Table 2 and Fig. 3A). This is consistent with a large body of evidence from cellular reporter and ribosome profiling studies indicating that the translation of iterated positive charges slows the ribosome (
      • Ahmed N.
      • Friedrich U.A.
      • Sormanni P.
      • Ciryam P.
      • Altman N.S.
      • Bukau B.
      • et al.
      Pairs of amino acids at the P- and A-sites of the ribosome predictably and causally modulate translation-elongation rates.
      ). However, the observation that different poly(lysine) encoding mRNA sequences differentially impact translation leads us to wonder if mRNA sequence, and therefore also structure, contribute to ribosome slowing and during poly(lysine) translation (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ,
      • Arthur L.L.
      • Pavlovic-Djuranovic S.
      • Koutmou K.S.
      • Green R.
      • Szczesny P.
      • Djuranovic S.
      Translational control by lysine-encoding A-rich sequences.
      ). To deconvolute the effects of peptide charge from mRNA sequence, we mischarged unmodified tRNALys and natively modified tRNANLys sequences with the small nonpolar amino acid valine (Val-tRNALys, Val-tRNANLys) (Fig. 4A) . Misacylation was accomplished using a small RNA microhelix (flexizyme) capable of attaching an esterified aminoacid acyl donor to virtually any tRNA of interest (
      • Goto Y.
      • Katoh T.
      • Suga H.
      Flexizymes for genetic code reprogramming.
      ). ICs containing mRNAs encoding consecutive lysines (AUG-AAA-AAA-UUC-UAA and AUG-AAG-AAG-UUC-UAA) were reacted with TCs possessing mischarged tRNAs (Val-tRNALys⋅EF-Tu⋅GTP, Val-tRNANLys⋅EF-Tu⋅GTP). Because the translation factor EF-Tu selects for correct tRNA and aminoacyl donor pairings, we titrated EF-Tu with each aatRNA to ensure saturating conditions for incorporating these species (Fig. S7) (
      • Dale T.
      • Sanderson L.E.
      • Uhlenbeck O.C.
      The affinity of elongation factor Tu for an aminoacyl-tRNA is modulated by the esterified amino acid.
      ,
      • Yikilmaz E.
      • Chapman S.J.
      • Schrader J.M.
      • Uhlenbeck O.C.
      The interface between Escherichia coli elongation factor Tu and aminoacyl-tRNA.
      ). Concurrent control assays with TCs containing Val-tRNAVal and ICs formed on an mRNA encoding consecutive valines (AUG-GUG-GUG-UUC-UAA) were also performed. The rate constants for Met-Val dipeptide and tripeptide, MV and MV2, formation are three-fold and five-fold slower (respectively) on AUG-AAA-AAA-UUC-UAA [MK2(AAA)FX] mRNA than AUG-GUG-GUG-UUC-UAA GUG [MV2FX] mRNA regardless of the modification status of tRNALys (Fig. 4C and Table S3). In contrast, the rate constants for synthesizing MV and MV2 on AUG-AAG-AAG-UUC-UAA [MK2(AAG)FX] mRNA are reduced by less than two-fold relative to (MV2FX) mRNA when modified Val-tRNANLys is included in the translation reaction. However, when unmodified Val-tRNALys is used instead, the ability of the ribosome to add Val to a growing polypeptide on an AAG codon is dramatically slowed, and the rate constants for MV and MV2 synthesis are diminished by >1000-fold (Fig. S8 and Table S3).
      Figure thumbnail gr4
      Figure 4Neutral amino acid and tRNA pairing effect ribosome sliding on poly(A) mRNA. A, lysyl E. coli tRNAUUU can be acylated with positively charged lysine or misacylated with neutral charge valine by the dFx flexizyme. Flexizyme charged tRNAs used in our studies assesing influence of amino acid charge on lysine addition during translation are displayed. B, phosphorimages of eTLCs displaying amino acid addition and ribosome sliding on mRNAs. Translation reactions were performed with either Val-tRNALys (left) or Val-tRNANLys TCs (middle) on MK2(AAA)FX mRNA or Val-tRNAVal (transcribed, right) on MV2FX mRNA. C, rate constants for amino acid addition on MK2(AAA)FX mRNA using Val-tRNALys and MV2FX mRNA using Val-tRNAVal TCs as defined by the proposed mechanism in A. The k3,obs rate constant is presented on the right y-axis, and no rate constant was obtainable for peptide synthesis on MV2FX mRNA (‡) as no sliding is observed in these assays. Error bars represent SD. D, percent of peptide products formed during assays on MK2(AAA)FX or MV2FX mRNAs after 20 min in translation assays conducted with Val-tRNALys, Val-tRNANLys, or Val-tRNAVal TCs as shown in panel (B). Error bars represent SD. eTLC, electrophoretic TLC; TC, ternary complex.
      In addition to impacting ribosome speed, our investigations with Val-tRNALys and Val-tRNANLys revealed that peptide charge also contributes to ribosome frame maintenance on lysine-encoding messages. We find that extended MV3+ peptides, analogous to the MK3+ peptides made during ribosome sliding (Fig. 4B), are generated from the AAA but not AAG or GUG containing messages. While unprogrammed MV3+ peptides can still be generated on consecutive AAA codons, the incorporation of additional valines is ∼10-fold slower than unprogrammed lysine addition on the same message. Our findings suggest that peptide charge and mRNA sequence make independent contributions to ribosome speed and frame maintenance during poly(lysine) peptide synthesis.

      m6A mRNA modifications suppress sliding on consecutive AAA codons in a position dependent manner

      Our data indicate that iterated AAA, but not AAG, lysine-encoding codons promote a series of consecutive −1 movements by the ribosome (ribosome sliding). We hypothesized that poly(A) regions might form a unique structure within the ribosome mRNA channel that promotes these loss of frame events. Recent cryo-EM structures of the yeast ribosome translating an mRNA sequence with six consecutive A nucleosides support this idea, revealing that stacked A’s adopt a single-stranded helix in the ribosome decoding center (
      • Tesina P.
      • Lessen L.N.
      • Buschauer R.
      • Cheng J.
      • Wu C.C.-C.
      • Berninghausen O.
      • et al.
      Molecular mechanism of translational stalling by inhibitory codon combinations and poly(A) tracts.
      ,
      • Chandrasekaran V.
      • Juszkiewicz S.
      • Choi J.
      • Puglisi J.D.
      • Brown A.
      • Shao S.
      • et al.
      Mechanism of ribosome stalling during translation of a poly(A) tail.
      ). In these structures, three A’s are positioned in the mRNA A site, where they form a helical stack with residues in the 18S rRNA (
      • Chandrasekaran V.
      • Juszkiewicz S.
      • Choi J.
      • Puglisi J.D.
      • Brown A.
      • Shao S.
      • et al.
      Mechanism of ribosome stalling during translation of a poly(A) tail.
      ). To test the possibility that such a helical structure might enhance frame loss on poly(A) sequences, we performed translation assays on AUG-AAA-AAA-UUC-UAA [MK2(AAA)FX] messages with various A-nucleosides substituted with m6A (Fig. 5A) to perturb the structure of this poly(A) helix. We selected m6A as a probe because it has been shown to change RNA structure and dynamics (
      • Liu B.
      • Shi H.
      • Rangadurai A.
      • Nussbaumer F.
      • Chu C.-C.
      • Erharter K.A.
      • et al.
      A quantitative model predicts how m6A reshapes the kinetic landscape of nucleic acid hybridization and conformational transitions.
      ,
      • Shi H.
      • Liu B.
      • Nussbaumer F.
      • Rangadurai A.
      • Kreutz C.
      • Al-Hashimi H.M.
      NMR chemical exchange measurements reveal that N6-methyladenosine slows RNA annealing.
      ,
      • Roost C.
      • Lynch S.R.
      • Batista P.J.
      • Qu K.
      • Chang H.Y.
      • Kool E.T.
      Structure and thermodynamics of N6-methyladenosine in RNA: a spring-loaded base modification.
      ). We find that when m6A is positioned in the middle of a six consecutive A nucleosides (at the third and fourth adenosine in the message), where they presumably could disrupt helix formation, very little extended peptide product is formed (Fig. 5, B and C and Table S4). In contrast, when the m6A is positioned at the fifth or sixth adenosine in the poly(A) sequence, sliding levels are comparable to those on an unmodified message (Fig. 5C). We examined the location of m6A in two available datasets that mapped m6A transcriptome wide to begin evaluating if our observation that m6A can promote frame maintenance could have relevance in endogenous A-rich mRNA coding sequences (
      • Linder B.
      • Grozhik A.V.
      • Olarerin-George A.O.
      • Meydan C.
      • Mason C.E.
      • Jaffrey S.R.
      Single-nucleotide resolution mapping of m6A and m6Am throughout the transcriptome.
      ,
      • Zhang Z.
      • Chen L.-Q.
      • Zhao Y.-L.
      • Yang C.-G.
      • Roundtree I.A.
      • Zhang Z.
      • et al.
      Single-base mapping of m6A by an antibody-independent method.
      ). Our bioinformatic analyses reveal that m6A exists both in AAA codons and in iterated A (five or more A’s) stretches found in the coding region of over 80 mRNAs (Table S5). These findings together raise the possibility that one consequence of m6A might be to prevent the ribosome from losing frame on consecutive AAA codons.
      Figure thumbnail gr5
      Figure 5m6A modification to single nucleotides in poly(A) modulate ribosome frame loss. A, MK2(AAA)FX mRNA was chemically modified with a single m6A on one of the six consecutive A’s. Specifically at positions A3, A4, A5, and A6 with position 3, m6A-3, shown. B, phosphorimage eTLC displaying the products of 20 min translation on MK2(AAA)FX transcripts harboring a single m6A modification at nucleotide specified. C, percent of total peptide formed as a result of ribosome sliding (longer than MK2 tri-peptide) on MK2(AAA)FX mRNAs that were either unmodified or harboring a single m6A modification after 20 min of translation using Lys-tRNANLys TCs. Error bars represent SD. There is no significant difference observed for peptide product formed on messages containing m6A at either position A3 or A4 using Lys-tRNANLys TCs when compared to peptide product formed on MK2(AAG)FX using Lys-tRNANLys TCs when using an unpaired student t test. eTLC, electrophoretic TLC; TC, ternary complex.

      Discussion

      The elongation of cationic peptides slows the ribosome and can impact the expression of thousands of proteins (Tables S1 and S2) (
      • Ahmed N.
      • Friedrich U.A.
      • Sormanni P.
      • Ciryam P.
      • Altman N.S.
      • Bukau B.
      • et al.
      Pairs of amino acids at the P- and A-sites of the ribosome predictably and causally modulate translation-elongation rates.
      ,
      • Leininger S.E.
      • Rodriguez J.
      • Vu Q.V.
      • Jiang Y.
      • Li M.S.
      • Deutsch C.
      • et al.
      Ribosome elongation kinetics of consecutively charged residues are coupled to electrostatic force.
      ). Available evidence suggests that ionic interactions between positively charged peptides and the ribosome account for the reduced translation speeds observed on these sequences (
      • Lu J.
      • Deutsch C.
      Electrostatics in the ribosomal tunnel modulate chain elongation rates.
      ,
      • Tesina P.
      • Lessen L.N.
      • Buschauer R.
      • Cheng J.
      • Wu C.C.-C.
      • Berninghausen O.
      • et al.
      Molecular mechanism of translational stalling by inhibitory codon combinations and poly(A) tracts.
      ,
      • Dimitrova L.N.
      • Kuroha K.
      • Tatematsu T.
      • Inada T.
      Nascent peptide-dependent translation arrest leads to Not4p-mediated protein degradation by the proteasome.
      ). We directly tested this model by comparing the rate constants for amino acid incorporation on MK2 encoding mRNAs using tRNALys aminoacylated with either a positively charged (Lys-tRNALys) or neutral (Val-tRNALys) amino acid (Fig. 4). Our investigations reveal that ribosomes translating consecutive AAA codons add a second amino acid (V or K) more slowly. The extent of slowing amino is not dependent on amino acid identity (Val-tRNALys k1/k2 = 6, Lys-tRNALys k1/k2 = 8). In contrast, on consecutive AAG lysine codons, the rate constants for adding the first and second lysine are equivalent (Lys-tRNALys k1/k2 = 1), while a second valine is added substantially more slowly (Val-tRNALys k1/k2 = 20) (Table S3). The inclusion of modifications in tRNANLys diminished the differences we observed in the rate constants for adding consecutive lysines and valines on AAG and AAA codons (on AAG: Val-tRNANLys k1/k2 = 4, Lys-tRNANLys k1/k2 = 3; on AAA: Val-tRNANLys k1/k2 = 2, Lys-tRNANLys k1/k2 = 6) (Tables 2 and S3). Our findings indicate that both peptide charge and codon:tRNALys interactions have significant roles in controlling poly(lysine) peptide synthesis rates. These factors play similarly important roles in ribosome frame maintenance during poly(lysine) peptide translation. While we observe ribosome sliding when Val-tRNALys/Val-tRNANLys are used (Fig. 4), the rate and extent of sliding on AAA codons are both modestly enhanced by Lys-tRNALys/Lys-tRNANLys. mRNA sequence (AAA versus AAG) and the posttranscriptional modification status of tRNALys have larger impacts on frame maintenance than peptide charge. As previously reported, we find that ribosome sliding is only prevalent on lysine encoding mRNAs with consecutive AAA codons (Figs. 3 and S1) (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ). We also noticed that the presence of tRNALys modifications significantly reduces the rate constant (25-fold) for sliding associated frame loss events (Fig. 3 and Table 2). The strong influence of tRNANLys modifications on frame maintenance is consistent with previous observations that both bacterial and yeast tRNA with these modifications also promote mRNA–tRNA interactions (
      • Rezgui V.A.N.
      • Tyagi K.
      • Ranjan N.
      • Konevega A.L.
      • Mittelstaet J.
      • Rodnina M.V.
      • et al.
      tRNA tKUUU, tQUUG, and tEUUC wobble position modifications fine-tune protein translation by promoting ribosome A-site binding.
      ,
      • Licznar P.
      • Mejlhede N.
      • Prère M.-F.
      • Wills N.
      • Gesteland R.F.
      • Atkins J.F.
      • et al.
      Programmed translational –1 frameshifting on hexanucleotide motifs and the wobble properties of tRNAs.
      ,
      • Johansson M.J.O.
      • Esberg A.
      • Huang B.
      • Björk G.R.
      • Byström A.S.
      Eukaryotic wobble uridine modifications promote a functionally redundant decoding system.
      ,
      • Rozov A.
      • Demeshkina N.
      • Khusainov I.
      • Westhof E.
      • Yusupov M.
      • Yusupova G.
      Novel base-pairing interactions at the tRNA wobble position crucial for accurate reading of the genetic code.
      ). In addition to enhancing frame maintenance, tRNANLys modifications appear to be especially important for the addition of valine on AAG, but not AAA, codons by mischarged Val-tRNALys. The rate constant for Val insertion increases by 70-fold to >1000-fold when fully modified Val-tRNANLys is reacted on AAG codons (Fig. S8 and Table S3). These findings collectively help to rationalize why loss of tRNANLys modifications is lethal in yeast and the observation that mutations in the tRNANLys modification machinery are linked to disease (
      • Ranjan N.
      • Rodnina M.V.
      tRNA wobble modifications and protein homeostasis.
      ,
      • Karlsborn T.
      • Tükenmez H.
      • Chen C.
      • Byström A.S.
      Familial dysautonomia (FD) patients have reduced levels of the modified wobble nucleoside mcm5s2U in tRNA.
      ,
      • Björk G.R.
      • Huang B.
      • Persson O.P.
      • Byström A.S.
      A conserved modified wobble nucleoside (mcm5s2U) in lysyl-tRNA is required for viability in yeast.
      ,
      • Suzuki T.
      The expanding world of tRNA modifications and their disease relevance.
      ,
      • Torres A.G.
      • Batlle E.
      • Ribas de Pouplana L.
      Role of tRNA modifications in human diseases.
      ,
      • Yasukawa T.
      • Suzuki T.
      • Ishii N.
      • Ohta S.
      • Watanabe K.
      Wobble modification defect in tRNA disturbs codon–anticodon interaction in a mitochondrial disease.
      ).
      mRNA–tRNA interactions can be controlled not only by tRNA modifications but also by the posttranscriptional modification of mRNAs. Emerging evidence suggests that mRNA modifications can slow the ribosome and influence the extent of amino acid misincorporation into peptides (
      • Jones J.D.
      • Monroe J.
      • Koutmou K.S.
      A molecular-level perspective on the frequency, distribution, and consequences of messenger RNA modifications.
      • Hoernes T.P.
      • Heimdörfer D.
      • Köstner D.
      • Faserl K.
      • Nußbaumer F.
      • Plangger R.
      • et al.
      Eukaryotic translation elongation is modulated by single natural nucleotide derivatives in the coding sequences of mRNAs.
      ,
      • Hoernes T.P.
      • Faserl K.
      • Juen M.A.
      • Kremser J.
      • Gasser C.
      • Fuchs E.
      • et al.
      Translation of non-standard codon nucleotides reveals minimal requirements for codon-anticodon interactions.
      ,
      • You C.
      • Dai X.
      • Wang Y.
      Position-dependent effects of regioisomeric methylated adenine and guanine ribonucleosides on translation.
      ). However, the impact of mRNA modifications on ribosome frame maintenance has not been explored. This question is especially relevant in the context of ribosome sliding because the most common mRNA modification, m6A, is present in AAA codons in cells (
      • Linder B.
      • Grozhik A.V.
      • Olarerin-George A.O.
      • Meydan C.
      • Mason C.E.
      • Jaffrey S.R.
      Single-nucleotide resolution mapping of m6A and m6Am throughout the transcriptome.
      ,
      • Zhang Z.
      • Chen L.-Q.
      • Zhao Y.-L.
      • Yang C.-G.
      • Roundtree I.A.
      • Zhang Z.
      • et al.
      Single-base mapping of m6A by an antibody-independent method.
      ). Our analysis of available datasets that map the location of m6A transcriptome-wide reveals that m6A is included into > 80 mRNAs containing five or more consecutive A’s (Table S5). While m6A disrupts both RNA base pairing and tRNA selection by the ribosome, the influence of m6A, or any other mRNA modification, on ribosome frame maintenance is not known (
      • Liu B.
      • Shi H.
      • Rangadurai A.
      • Nussbaumer F.
      • Chu C.-C.
      • Erharter K.A.
      • et al.
      A quantitative model predicts how m6A reshapes the kinetic landscape of nucleic acid hybridization and conformational transitions.
      ,
      • Choi J.
      • Ieong K.-W.
      • Demirci H.
      • Chen J.
      • Petrov A.
      • Prabhakar A.
      • et al.
      N 6 -methyladenosine in mRNA disrupts tRNA selection and translation-elongation dynamics.
      ,
      • Ieong K.-W.
      • Indrisiunaite G.
      • Prabhakar A.
      • Puglisi J.D.
      • Ehrenberg M.
      N 6-Methyladenosines in mRNAs reduce the accuracy of codon reading by transfer RNAs and peptide release factors.
      ). Our results demonstrate that m6A can suppress frame loss events. We find that ribosome sliding is limited when m6A is positioned to break up stretches of iterative adenosines (Fig. 5). Together, our biochemical and bioinformatic findings suggest that one consequence of having m6A present in these mRNA sequences could be to prevent ribosome sliding in homopolymeric A-rich stretches (
      • Kladwang W.
      • Topkar V.V.
      • Liu B.
      • Rangan R.
      • Hodges T.L.
      • Keane S.C.
      • et al.
      Anomalous reverse transcription through chemical modifications in polyadenosine stretches.
      ).
      While previous studies revealed that the ribosome moves robustly into multiple frames on poly(A) sequences in vitro, the mechanism by which this occurs was not known (
      • Koutmou K.S.
      • Schuller A.P.
      • Brunelle J.L.
      • Radhakrishnan A.
      • Djuranovic S.
      • Green R.
      Ribosomes slide on lysine-encoding homopolymeric A stretches.
      ,
      • Tsai T.-W.
      • Yang H.
      • Yin H.
      • Xu S.
      • Wang Y.
      High-efficiency “–1” and “–2” ribosomal frameshiftings revealed by force spectroscopy.
      ,
      • Yan S.
      • Wen J.-D.
      • Bustamante C.
      • Tinoco I.
      Ribosome excursions during mRNA translocation mediate broad branching of frameshift pathways.
      ). Here, we measured the rate constants for the ribosome moving into five different reading frames while translating an AUG-AAA-AAA-UUC-UAA [MK2(AAA)FX] mRNA sequence. These data lead us to propose a model for ribosome sliding in which the ribosome moves backward, one nucleotide at a time, along a homopolymeric(A) sequence until it either encounters an aatRNAaa that it can react with, or enters an unproductive state (Figs. 2 and S3). We find that the rate constants for backward movements by the ribosome are reduced by 50-fold relative to 0-frame amino acid addition and that the greatest levels of nonproductive complex formation are observed following the first −1 movement (Fig. 1D and Table 1). The rate constants for these movements, collectively referred to as ribosome sliding, are reduced relative to normal amino acid addition, amino acid misincorporation, and −1 programmed frame shifting (
      • Jones J.D.
      • Monroe J.
      • Koutmou K.S.
      A molecular-level perspective on the frequency, distribution, and consequences of messenger RNA modifications.
      ,
      • Caliskan N.
      • Katunin V.I.
      • Belardinelli R.
      • Peske F.
      • Rodnina M.V.
      Programmed –1 frameshifting by kinetic partitioning during impeded translocation.
      ,
      • Eyler D.E.
      • Franco M.K.
      • Batool Z.
      • Wu M.Z.
      • Dubuke M.L.
      • Dobosz-Bartoszek M.
      • et al.
      Pseudouridinylation of mRNA coding sequences alters translation.
      ). While we do observe some amino acid addition in the +1 frame, these events are less robust than their −1 counterparts, and we believe that the +1 frameshifts are not a product of sliding, but rather the result of an empty A site (
      • Liao P.-Y.
      • Gupta P.
      • Petrov A.N.
      • Dinman J.D.
      • Lee K.H.
      A new kinetic model reveals the synergistic effect of E-, P- and A-sites on +1 ribosomal frameshifting.
      ).
      The slow nature of the progressive −1 movements and formation of nonproductive complexes can help explain why it is possible to capture the ribosome in multiple frames in vitro but only visualize ribosome stalling and the first −1 movement in cells. Following the first −1 movement on poly(A) sequence, our results suggest that the ribosome is capable of adding the amino acid in the −1 frame, likely leading to the creation of a prematurely truncated protein product and triggering nonsense-mediated decay. Indeed, there is evidence that nonsense-mediated decay occurs on following −1 frameshifts on poly(A) containing reporters in human cells (
      • Arthur L.L.
      • Pavlovic-Djuranovic S.
      • Koutmou K.S.
      • Green R.
      • Szczesny P.
      • Djuranovic S.
      Translational control by lysine-encoding A-rich sequences.
      ,
      • Sitron C.S.
      • Brandman O.
      Detection and degradation of stalled nascent chains via ribosome-associated quality control.
      ). Additionally, the −1 movement is quite slow (longer than 1 min), giving time for ribosome collisions to occur and a cotranslational quality control mechanism targeting the stalled ribosome complex to be activated (
      • Meydan S.
      • Guydosh N.R.
      Disome and trisome profiling reveal genome-wide targets of ribosome quality control.
      ,
      • Juszkiewicz S.
      • Hegde R.S.
      Initiation of quality control during poly(A) translation requires site-specific ribosome ubiquitination.
      ,
      • Guydosh N.R.
      • Green R.
      Translation of poly(A) tails leads to precise mRNA cleavage.
      ). Together, this work quantitatively describes how the ribosome translates mRNAs containing poly(A) regions and reveals the important contributions of tRNA–mRNA interactions to ribosome slowing during cationic peptide. Our results suggest that posttranscriptional mRNA modifications may provide cells with a previously unrecognized avenue to ensure that the ribosome remains in the correct mRNA reading frame during the translation of slippery sequences.

      Experimental procedures

      In vitro translation assays

      70S ICs were prepared using E. coli ribosomes programmed with various mRNAs and f-[35S]-Met-tRNAMet in the P site (
      • Youngman E.M.
      • Brunelle J.L.
      • Kochaniak A.B.
      • Green R.
      The active site of the ribosome is composed of two layers of conserved nucleotides with distinct roles in peptide bond formation and peptide release.
      ). Translation was initiated by mixing equal volumes of TC (20–60 μM aatRNA(s), 24 μM EF-G, 60 μM EF-Tu) with ICs (140 nM) in 219-Tris buffer (50 mM Tris pH 7.5, 70 mM NH4Cl, 30 mM KCl, 7 mM MgCl2, and 5 mM βME). All initiation factors (IF-1, IF-2, and IF-3) and translation factors (EF-Tu and EF-G) used were His-tag purified from E. coli using plasmids available from AddGene (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ). The reactions were quenched with equal volume of 1 M KOH at discrete time points (0 s–20 min) by hand (5 s–20 min) or using a KinTek RQF-3 quench flow apparatus (0.001–5 s) (
      • Zaher H.S.
      • Green R.
      Quality control by the ribosome following peptide bond formation.
      ). Each sample was diluted 1:10 in nuclease-free water, and the reactants, intermediates, and products were separated by eTLC, visualized by phosphorimaging, and quantified with ImageQuant (Cytiva Life Sciences) as previously described (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ). Depending on the expected peptide products, eTLCs were run in different running buffer conditions to improve separation (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ). eTLCs analyzing peptides containing one or more lysines were run in pyridine acetate buffer, pH 2.8, while eTLCs separating peptides with valine (but no lysine) were run in pyridine acetate buffer, pH 5.2. For m6A studies in this work, all m6A mRNA constructs were purchased from Dharmacon, Horizon Discovery.

      tRNA aminoacylation by synthetases and flexizyme

      tRNAs used in experiments were either transcribed with T7 polymerase or were overexpressed and purified from using a pCWAug vector. tRNAs were then aminoacylated using either purified bacterial lysine aatRNA synthetase (LysRS) or misacylated using the dFx flexizyme as described previously (
      • Goto Y.
      • Katoh T.
      • Suga H.
      Flexizymes for genetic code reprogramming.
      ,
      • Murakami H.
      • Ohta A.
      • Ashigai H.
      • Suga H.
      A highly flexible tRNA acylation method for non-natural polypeptide synthesis.
      ). In the case of native tRNANLys, a pUC57 plasmid containing the E. coli tRNANLysUUU sequence for study was transformed in HB101 cells. These were then grown, purified, and tRNA deacylated for use in synthetase and flexizyme aminoacylation assays, as described previously (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ). All other T7 transcribed tRNAs used in this study were aminoacylated using their cognate aatRNA synthetase (AsnRS, IleRS, PheRS, SerRS, and ValRS) and were His-tag purified from E. coli using plasmids available from AddGene (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ). E. coli tRNANfMet was natively purified from a pCWAug vector and methionine was formylated and installed on tRNANfMet using MTF and MetRS enzymes, which were His-tag purified from E. coli using plasmids available from AddGene (
      • Monroe J.G.
      • Smith T.J.
      • Koutmou K.S.
      Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
      ,
      • Walker S.E.
      • Fredrick K.
      Preparation and evaluation of acylated tRNAs.
      ).

      Global analysis simulations of amino acid addition

      The reactions used to fit and model our data are displayed in Fig. S3. The fits used to obtain k1 and k2 were modeled using differential equations in Kaleidagraph. Subsequent rate constants (k3+) were modeled against simulations using KinTek Explorer. Simulations in KinTek Explorer were run against different potential mechanisms of ribosome sliding (Fig. S3) using data from quantified peptide formation.

      Homopolymeric A sequences in human coding sequences

      Homo sapiens genome assembly GRCh37 (hg19 release 75, cds.fa) data were used to identify the consecutive A’s in human coding sequences. Consecutive A’s were counted using in-house R scripts. Then, we analyzed single nucleotide resolution m6A mapping studies in different tissues (
      • Linder B.
      • Grozhik A.V.
      • Olarerin-George A.O.
      • Meydan C.
      • Mason C.E.
      • Jaffrey S.R.
      Single-nucleotide resolution mapping of m6A and m6Am throughout the transcriptome.
      ,
      • Zhang Z.
      • Chen L.-Q.
      • Zhao Y.-L.
      • Yang C.-G.
      • Roundtree I.A.
      • Zhang Z.
      • et al.
      Single-base mapping of m6A by an antibody-independent method.
      ) to find out whether any of these consecutive A’s have at least one installed m6A modification. This analysis yielded m6A frequencies that were reported in Table S5.

      Data availability

      The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information.

      Supporting information

      This article contains supporting information (
      • Linder B.
      • Grozhik A.V.
      • Olarerin-George A.O.
      • Meydan C.
      • Mason C.E.
      • Jaffrey S.R.
      Single-nucleotide resolution mapping of m6A and m6Am throughout the transcriptome.
      ,
      • Zhang Z.
      • Chen L.-Q.
      • Zhao Y.-L.
      • Yang C.-G.
      • Roundtree I.A.
      • Zhang Z.
      • et al.
      Single-base mapping of m6A by an antibody-independent method.
      ).

      Conflict of interest

      The authors declare that they have no conflicts of interest with the contents of this article.

      Acknowledgments

      We are grateful to Prof. Pavel Nagorny (University of Michigan) for discussions and use of laboratory infrastructure to synthesize activated Valine-DNB required for the flexizyme reactions. BioRender software was used to generate Fig. S4.

      Author contributions

      T. J. S. conceptualization; T. J. S. and H. R. K. methodology; M. T. validation; T. J. S. and M. T. formal analysis; T. J. S. investigation; H. R. K. resources; M. T. data curation; T. J. S. writing–original draft; K. S. K. writing–review & editing; T. J. S. visualization; K. S. K. supervision; K. S. K. project administration; K. S. K. funding acquisition.

      Funding and additional information

      National Institutes of Health (NIGMS R35 GM128836) (K. S. K.); Research Corporation for Science Advancement (Cottrell Scholar Award) (K. S. K.); University of Michigan (start-up funds to K. S. K.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

      Supporting information

      References

        • Wohlgemuth I.
        • Pohl C.
        • Mittelstaet J.
        • Konevega A.L.
        • Rodnina M.V.
        Evolutionary optimization of speed and accuracy of decoding on the ribosome.
        Philos. Trans. R. Soc. B Biol. Sci. 2011; 366: 2979-2986
        • Pedersen S.
        Escherichia coli ribosomes translate in vivo with variable rate.
        EMBO J. 1984; 3: 2895-2898
        • Siller E.
        • DeZwaan D.C.
        • Anderson J.F.
        • Freeman B.C.
        • Barral J.M.
        Slowing bacterial translation speed enhances eukaryotic protein folding efficiency.
        J. Mol. Biol. 2010; 396: 1310-1318
        • Qian W.
        • Yang J.-R.
        • Pearson N.M.
        • Maclean C.
        • Zhang J.
        Balanced codon usage optimizes eukaryotic translational efficiency.
        PLoS Genet. 2012; 8e1002603
        • Ban N.
        • Nissen P.
        • Hansen J.
        • Moore P.B.
        • Steitz T.A.
        The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution.
        Science. 2000; 289: 905-920
        • Jacobs W.M.
        • Shakhnovich E.I.
        Evidence of evolutionary selection for cotranslational folding.
        Proc. Natl. Acad. Sci. U. S. A. 2017; 114: 11434-11439
        • Hanson G.
        • Alhusaini N.
        • Morris N.
        • Sweet T.
        • Coller J.
        Translation elongation and mRNA stability are coupled through the ribosomal A-site.
        RNA. 2018; 24: 1377-1389
        • Presnyak V.
        • Alhusaini N.
        • Chen Y.-H.
        • Martin S.
        • Morris N.
        • Kline N.
        • et al.
        Codon optimality is a major determinant of mRNA stability.
        Cell. 2015; 160: 1111-1124
        • Choi J.
        • Grosely R.
        • Prabhakar A.
        • Lapointe C.P.
        • Wang J.
        • Puglisi J.D.
        How mRNA and nascent chain sequences regulate translation elongation.
        Annu. Rev. Biochem. 2018; 87: 421-449
        • Lu J.
        • Deutsch C.
        Electrostatics in the ribosomal tunnel modulate chain elongation rates.
        J. Mol. Biol. 2008; 384: 73-86
        • Brandman O.
        • Stewart-Ornstein J.
        • Wong D.
        • Larson A.
        • Williams C.C.
        • Li G.-W.
        • et al.
        A ribosome-bound quality control complex triggers degradation of nascent peptides and signals translation stress.
        Cell. 2012; 151: 1042-1054
        • Charneski C.A.
        • Hurst L.D.
        Positively charged residues are the major determinants of ribosomal velocity.
        PLoS Biol. 2013; 11e1001508
        • Koutmou K.S.
        • Schuller A.P.
        • Brunelle J.L.
        • Radhakrishnan A.
        • Djuranovic S.
        • Green R.
        Ribosomes slide on lysine-encoding homopolymeric A stretches.
        Elife. 2015; 4e05534
        • Arthur L.L.
        • Pavlovic-Djuranovic S.
        • Koutmou K.S.
        • Green R.
        • Szczesny P.
        • Djuranovic S.
        Translational control by lysine-encoding A-rich sequences.
        Sci. Adv. 2015; 1e1500154
        • Rosenberg A.H.
        • Goldman E.
        • Dunn J.J.
        • Studier F.W.
        • Zubay G.
        Effects of consecutive AGG codons on translation in Escherichia coli, demonstrated with a versatile codon test system.
        J. Bacteriol. 1993; 175: 716-722
        • Sørensen M.A.
        • Kurland C.G.
        • Pedersen S.
        Codon usage determines translation rate in Escherichia coli.
        J. Mol. Biol. 1989; 207: 365-377
        • Liu Y.
        • Yang Q.
        • Zhao F.
        Synonymous but not silent: the codon usage code for gene expression and protein folding.
        Annu. Rev. Biochem. 2021; 90: 375-401
        • Ito-Harashima S.
        • Kuroha K.
        • Tatematsu T.
        • Inada T.
        Translation of the poly(A) tail plays crucial roles in nonstop mRNA surveillance via translation repression and protein destabilization by proteasome in yeast.
        Genes Dev. 2007; 21: 519-524
        • Chiabudini M.
        • Tais A.
        • Zhang Y.
        • Hayashi S.
        • Wölfle T.
        • Fitzke E.
        • et al.
        Release factor eRF3 mediates premature translation termination on polylysine-stalled ribosomes in Saccharomyces cerevisiae.
        Mol. Cell. Biol. 2014; 34: 4062-4076
        • Agirrezabala X.
        • Samatova E.
        • Klimova M.
        • Zamora M.
        • Gil-Carton D.
        • Rodnina M.V.
        • et al.
        Ribosome rearrangements at the onset of translational bypassing.
        Sci. Adv. 2017; 3e1700147
        • Dinman J.D.
        Mechanisms and implications of programmed translational frameshifting.
        Wiley Interdiscip. Rev. RNA. 2012; 3: 661-673
        • Jones J.D.
        • Monroe J.
        • Koutmou K.S.
        A molecular-level perspective on the frequency, distribution, and consequences of messenger RNA modifications.
        Wiley Interdiscip. Rev. RNA. 2020; 11e1586
        • Ranjan N.
        • Rodnina M.V.
        tRNA wobble modifications and protein homeostasis.
        Translation. 2016; 4e1143076
        • Tsai T.-W.
        • Yang H.
        • Yin H.
        • Xu S.
        • Wang Y.
        High-efficiency “–1” and “–2” ribosomal frameshiftings revealed by force spectroscopy.
        ACS Chem. Biol. 2017; 12: 1629-1635
        • Yan S.
        • Wen J.-D.
        • Bustamante C.
        • Tinoco I.
        Ribosome excursions during mRNA translocation mediate broad branching of frameshift pathways.
        Cell. 2015; 160: 870-881
        • Monroe J.G.
        • Smith T.J.
        • Koutmou K.S.
        Investigating the consequences of mRNA modifications on protein synthesis using in vitro translation assays.
        Methods Enzymol. 2021; 658: 379-406
        • Kim H.-K.
        • Tinoco I.
        EF-G catalyzed translocation dynamics in the presence of ribosomal frameshifting stimulatory signals.
        Nucleic Acids Res. 2017; 45: 2865-2874
        • Klimova M.
        • Senyushkina T.
        • Samatova E.
        • Peng B.Z.
        • Pearson M.
        • Peske F.
        • et al.
        EF-G–induced ribosome sliding along the noncoding mRNA.
        Sci. Adv. 2019; 5eaaw9049
        • Chen J.
        • Petrov A.
        • Johansson M.
        • Tsai A.
        • O'Leary S.E.
        • Puglisi J.D.
        Dynamic pathways of -1 translational frameshifting.
        Nature. 2014; 512: 328-332
        • Lorenz C.
        • Lünse C.E.
        • Mörl M.
        tRNA modifications: impact on structure and thermal adaptation.
        Biomolecules. 2017; 7: 35
        • Agris P.F.
        • Narendran A.
        • Sarachan K.
        • Väre V.Y.P.
        • Eruysal E.
        The role of RNA modifications in translational fidelity.
        Enzymes. 2017; 41: 1-50
        • Bekaert M.
        • Rousset J.-P.
        An extended signal involved in eukaryotic -1 frameshifting operates through modification of the E site tRNA.
        Mol. Cell. 2005; 17: 61-68
        • Maehigashi T.
        • Dunkle J.A.
        • Miles S.J.
        • Dunham C.M.
        Structural insights into +1 frameshifting promoted by expanded or modification-deficient anticodon stem loops.
        Proc. Natl. Acad. Sci. U. S. A. 2014; 111: 12740-12745
        • Rezgui V.A.N.
        • Tyagi K.
        • Ranjan N.
        • Konevega A.L.
        • Mittelstaet J.
        • Rodnina M.V.
        • et al.
        tRNA tKUUU, tQUUG, and tEUUC wobble position modifications fine-tune protein translation by promoting ribosome A-site binding.
        Proc. Natl. Acad. Sci. U. S. A. 2013; 110: 12289-12294
        • Ranjan N.
        • Rodnina M.V.
        Thio-modification of tRNA at the wobble position as regulator of the kinetics of decoding and translocation on the ribosome.
        J. Am. Chem. Soc. 2017; 139: 5857-5864
        • Ahmed N.
        • Friedrich U.A.
        • Sormanni P.
        • Ciryam P.
        • Altman N.S.
        • Bukau B.
        • et al.
        Pairs of amino acids at the P- and A-sites of the ribosome predictably and causally modulate translation-elongation rates.
        J. Mol. Biol. 2020; 432: 166696
        • Goto Y.
        • Katoh T.
        • Suga H.
        Flexizymes for genetic code reprogramming.
        Nat. Protoc. 2011; 6: 779-790
        • Dale T.
        • Sanderson L.E.
        • Uhlenbeck O.C.
        The affinity of elongation factor Tu for an aminoacyl-tRNA is modulated by the esterified amino acid.
        Biochemistry. 2004; 43: 6159-6166
        • Yikilmaz E.
        • Chapman S.J.
        • Schrader J.M.
        • Uhlenbeck O.C.
        The interface between Escherichia coli elongation factor Tu and aminoacyl-tRNA.
        Biochemistry. 2014; 53: 5710-5720
        • Tesina P.
        • Lessen L.N.
        • Buschauer R.
        • Cheng J.
        • Wu C.C.-C.
        • Berninghausen O.
        • et al.
        Molecular mechanism of translational stalling by inhibitory codon combinations and poly(A) tracts.
        EMBO J. 2020; 39e103365
        • Chandrasekaran V.
        • Juszkiewicz S.
        • Choi J.
        • Puglisi J.D.
        • Brown A.
        • Shao S.
        • et al.
        Mechanism of ribosome stalling during translation of a poly(A) tail.
        Nat. Struct. Mol. Biol. 2019; 26: 1132-1140
        • Liu B.
        • Shi H.
        • Rangadurai A.
        • Nussbaumer F.
        • Chu C.-C.
        • Erharter K.A.
        • et al.
        A quantitative model predicts how m6A reshapes the kinetic landscape of nucleic acid hybridization and conformational transitions.
        Nat. Commun. 2021; 12: 5201
        • Shi H.
        • Liu B.
        • Nussbaumer F.
        • Rangadurai A.
        • Kreutz C.
        • Al-Hashimi H.M.
        NMR chemical exchange measurements reveal that N6-methyladenosine slows RNA annealing.
        J. Am. Chem. Soc. 2019; 141: 19988-19993
        • Roost C.
        • Lynch S.R.
        • Batista P.J.
        • Qu K.
        • Chang H.Y.
        • Kool E.T.
        Structure and thermodynamics of N6-methyladenosine in RNA: a spring-loaded base modification.
        J. Am. Chem. Soc. 2015; 137: 2107-2115
        • Linder B.
        • Grozhik A.V.
        • Olarerin-George A.O.
        • Meydan C.
        • Mason C.E.
        • Jaffrey S.R.
        Single-nucleotide resolution mapping of m6A and m6Am throughout the transcriptome.
        Nat. Methods. 2015; 12: 767-772
        • Zhang Z.
        • Chen L.-Q.
        • Zhao Y.-L.
        • Yang C.-G.
        • Roundtree I.A.
        • Zhang Z.
        • et al.
        Single-base mapping of m6A by an antibody-independent method.
        Sci. Adv. 2019; 5eaax0250
        • Leininger S.E.
        • Rodriguez J.
        • Vu Q.V.
        • Jiang Y.
        • Li M.S.
        • Deutsch C.
        • et al.
        Ribosome elongation kinetics of consecutively charged residues are coupled to electrostatic force.
        Biochemistry. 2021; 60: 3223-3235
        • Dimitrova L.N.
        • Kuroha K.
        • Tatematsu T.
        • Inada T.
        Nascent peptide-dependent translation arrest leads to Not4p-mediated protein degradation by the proteasome.
        J. Biol. Chem. 2009; 284: 10343-10352
        • Licznar P.
        • Mejlhede N.
        • Prère M.-F.
        • Wills N.
        • Gesteland R.F.
        • Atkins J.F.
        • et al.
        Programmed translational –1 frameshifting on hexanucleotide motifs and the wobble properties of tRNAs.
        EMBO J. 2003; 22: 4770-4778
        • Johansson M.J.O.
        • Esberg A.
        • Huang B.
        • Björk G.R.
        • Byström A.S.
        Eukaryotic wobble uridine modifications promote a functionally redundant decoding system.
        Mol. Cell. Biol. 2008; 28: 3301-3312
        • Rozov A.
        • Demeshkina N.
        • Khusainov I.
        • Westhof E.
        • Yusupov M.
        • Yusupova G.
        Novel base-pairing interactions at the tRNA wobble position crucial for accurate reading of the genetic code.
        Nat. Commun. 2016; 7: 10457
        • Karlsborn T.
        • Tükenmez H.
        • Chen C.
        • Byström A.S.
        Familial dysautonomia (FD) patients have reduced levels of the modified wobble nucleoside mcm5s2U in tRNA.
        Biochem. Biophys. Res. Commun. 2014; 454: 441-445
        • Björk G.R.
        • Huang B.
        • Persson O.P.
        • Byström A.S.
        A conserved modified wobble nucleoside (mcm5s2U) in lysyl-tRNA is required for viability in yeast.
        RNA. 2007; 13: 1245-1255
        • Suzuki T.
        The expanding world of tRNA modifications and their disease relevance.
        Nat. Rev. Mol. Cell Biol. 2021; 22: 375-392
        • Torres A.G.
        • Batlle E.
        • Ribas de Pouplana L.
        Role of tRNA modifications in human diseases.
        Trends Mol. Med. 2014; 20: 306-314
        • Yasukawa T.
        • Suzuki T.
        • Ishii N.
        • Ohta S.
        • Watanabe K.
        Wobble modification defect in tRNA disturbs codon–anticodon interaction in a mitochondrial disease.
        EMBO J. 2001; 20: 4794-4802
        • Hoernes T.P.
        • Heimdörfer D.
        • Köstner D.
        • Faserl K.
        • Nußbaumer F.
        • Plangger R.
        • et al.
        Eukaryotic translation elongation is modulated by single natural nucleotide derivatives in the coding sequences of mRNAs.
        Genes (Basel). 2019; 10: 84
        • Hoernes T.P.
        • Faserl K.
        • Juen M.A.
        • Kremser J.
        • Gasser C.
        • Fuchs E.
        • et al.
        Translation of non-standard codon nucleotides reveals minimal requirements for codon-anticodon interactions.
        Nat. Commun. 2018; 9: 4865
        • You C.
        • Dai X.
        • Wang Y.
        Position-dependent effects of regioisomeric methylated adenine and guanine ribonucleosides on translation.
        Nucleic Acids Res. 2017; 45: 9059-9067
        • Choi J.
        • Ieong K.-W.
        • Demirci H.
        • Chen J.
        • Petrov A.
        • Prabhakar A.
        • et al.
        N 6 -methyladenosine in mRNA disrupts tRNA selection and translation-elongation dynamics.
        Nat. Struct. Mol. Biol. 2016; 23: 110-115
        • Ieong K.-W.
        • Indrisiunaite G.
        • Prabhakar A.
        • Puglisi J.D.
        • Ehrenberg M.
        N 6-Methyladenosines in mRNAs reduce the accuracy of codon reading by transfer RNAs and peptide release factors.
        Nucleic Acids Res. 2021; 49: 2684-2699
        • Kladwang W.
        • Topkar V.V.
        • Liu B.
        • Rangan R.
        • Hodges T.L.
        • Keane S.C.
        • et al.
        Anomalous reverse transcription through chemical modifications in polyadenosine stretches.
        Biochemistry. 2020; 59: 2154-2170
        • Caliskan N.
        • Katunin V.I.
        • Belardinelli R.
        • Peske F.
        • Rodnina M.V.
        Programmed –1 frameshifting by kinetic partitioning during impeded translocation.
        Cell. 2014; 157: 1619-1631
        • Eyler D.E.
        • Franco M.K.
        • Batool Z.
        • Wu M.Z.
        • Dubuke M.L.
        • Dobosz-Bartoszek M.
        • et al.
        Pseudouridinylation of mRNA coding sequences alters translation.
        Proc. Natl. Acad. Sci. U. S. A. 2019; 116: 23068-23074
        • Liao P.-Y.
        • Gupta P.
        • Petrov A.N.
        • Dinman J.D.
        • Lee K.H.
        A new kinetic model reveals the synergistic effect of E-, P- and A-sites on +1 ribosomal frameshifting.
        Nucleic Acids Res. 2008; 36: 2619-2629
        • Sitron C.S.
        • Brandman O.
        Detection and degradation of stalled nascent chains via ribosome-associated quality control.
        Annu. Rev. Biochem. 2020; 89: 417-442
        • Meydan S.
        • Guydosh N.R.
        Disome and trisome profiling reveal genome-wide targets of ribosome quality control.
        Mol. Cell. 2020; 79: 588-602.e6
        • Juszkiewicz S.
        • Hegde R.S.
        Initiation of quality control during poly(A) translation requires site-specific ribosome ubiquitination.
        Mol. Cell. 2017; 65: 743-750.e4
        • Guydosh N.R.
        • Green R.
        Translation of poly(A) tails leads to precise mRNA cleavage.
        RNA. 2017; 23: 749-761
        • Youngman E.M.
        • Brunelle J.L.
        • Kochaniak A.B.
        • Green R.
        The active site of the ribosome is composed of two layers of conserved nucleotides with distinct roles in peptide bond formation and peptide release.
        Cell. 2004; 117: 589-599
        • Zaher H.S.
        • Green R.
        Quality control by the ribosome following peptide bond formation.
        Nature. 2009; 457: 161-166
        • Murakami H.
        • Ohta A.
        • Ashigai H.
        • Suga H.
        A highly flexible tRNA acylation method for non-natural polypeptide synthesis.
        Nat. Methods. 2006; 3: 357-359
        • Walker S.E.
        • Fredrick K.
        Preparation and evaluation of acylated tRNAs.
        Methods. 2008; 44: 81-86