The surprisingly complex tRNA subsystem: part 3—quality control mechanisms
Dozens of transfer RNAs (tRNAs) each requiring a precise structure and properties are indispensable to translate the genetic code. Cells possess sophisticated quality control mechanisms to identify and eliminate those which are improperly formed. Enzyme complexes can recognize very small errors in processed tRNAs, such as lack of an important methylation on a single nucleotide position (m1A58) in tRNAiMet. Controlled degradation of malformed tRNAs occurs through several mechanisms, helping to prevent errors in cellular processes and freeing up raw materials for reuse. Additional quality control measures include multiple proof-reading checks to ensure correct tRNA anticodon pairing to the correct mRNA codon and ensuring the correct amounts of tRNAs are made available at the right time and in the correct location.
Cells cannot survive nor reproduce without the tRNA subsystem, so this must have been in place in the very first one. The quality control processes, however, are protein-based and would not have been available before they were needed without a deliberate plan. Proteins are only produced by a functioning genetic code apparatus. A putative initial primitive genetic system which generated improperly formed tRNAs and lacked any kind of quality control processes could not have provided the precision needed to ensure survival generation after generation.
The tRNA subsystem is an indispensable component of a cell’s genetic apparatus, but tRNAs rely on biochemical processing by many protein-based molecular machines to work. On the other hand, the relevant proteins depend on the genetic system being already in place, possible only if the functional tRNAs already exist. The question a scientist faces is, what is supposed to have originated first?
In part 11 of this series we showed that tRNAs are not merely sections of RNA used as linker molecules to translate messenger RNA (mRNA) codons to produce proteins. The 5’ leader sequence must be removed from precursors using complex ribonucleases constructed by RNA and proteins having virtually no similarity across prokaryotes, mitochondria and eukaryotes. Removal of 3’ extensions from pre-tRNA involves both exo- and endonucleases, and introns must also be removed although the location of the splice junctions differ for eukaryotes and archaea.
Aminoacylation of tRNAs requires them to have the trinucleotide motif ‘CCA’ on the 3’ end, which is also necessary for processing in ribosomes.2 In some kinds of organism, the CCA is encoded on the tRNA genes and for others it is added by a complex enzyme.3 Both schemes require elaborate processing by additional protein-based molecular machines. The pattern of alternative sources of CCA on tRNAs is unexpected if all shared a common ancestor.
In part 1 we also pointed out that kinetic studies performed on E. coli revealed that the enzyme to repair damaged CCA sequences also has an innate ability to discriminate against damaged tRNAs by scrutinizing the integrity of the tRNA substrate. Moreover, the enzyme recognizes three kinds of tRNA flaws—a missing C74C75A76, C75A76, or A76—and reconstructs the CCA-3’ sequence as needed.4
Defective tRNAs are rapidly degraded by RNA surveillance mechanisms found in all cells, preventing them from entering the ribosome machinery.5 Structurally unstable tRNAs are also actively eliminated by first tagging them with a second CCA, leading to 3’-terminal CCACCA sequences, which serve as a specific degradation tag.6,7,8 Here in Part 3 we will examine quality control and related features in more detail.
In part 29 we drew attention to the fact that many chemical modifications of tRNA nucleotides are necessary in all organisms to produce the correct clover-leaf structure, ensure rigidity, prevent undesired interactions and permit correct processing at the ribosome. The modifications are dynamic, tissue specific, respond to environmental changes, and tRNA concentrations are used as a regulatory signal in several ways to regulate the amount of translation.
1. tRNA quality control mechanisms
tRNA degradation pathways to control expression levels
Cells possess multiple pathways to degrade inappropriately processed or misfolded tRNAs and, in addition, immature tRNAs help regulate translation by interacting with the protein synthesis machinery. They also serve as intracellular signalling molecules to communicate stress.10 But mature tRNAs are very stable with half-lives from ~9 h up to a few days.10 Mature mammalian tRNAs have half-lives around 100 h, so with cellular tRNA/ribosome ratios of about 10 and protein synthesis rate on the ribosome around 2–5 peptide bonds per second, charged tRNAs should be replenished every few seconds on average. But physiological needs vary greatly according to circumstances.11
Eraser enzymes12 can remove chemical modifications on tRNA nucleotides within a few minutes and together with multiple regulatory factors change tRNA lifetimes and processing under a wide range of physiological conditions.11 Other examples of regulatory dynamic fine-tuning are evident when lack of an essential amino acid and changes in glucose concentration lead to differential charging among tRNA isoacceptors.11,13
Surprisingly, two tRNA degradation pathways have been found, which we will summarize next.
3’-5’ exonucleolytic degradation by the nuclear exosome
tRNAiMet , used to identify where mRNA translation is to begin, is unstable if it lacks the m1A58 methylation. This modification is critically important in bacteria, archaea, and eukaryotes.14 Unmodified tRNAiMet is enzymatically polyadenylated at the 3’ end and then degraded by a multi-protein complex called an exosome. It is interesting that poly(A) tails on mRNA generally specify stability for mRNA transcripts,15 whereas in E. coli tRNA degradation proceeds by poly(A) addition.10,16 Turnover is not a trivial matter, requiring Mtr4, an RNA-dependent helicase, and other proteins comprising the TRAMP complex17 including the two poly(A) polymerases, Trf4 and Trf5, and either Air1 or Air2, RNA binding proteins.10
This 3’ to 5’ turnover machinery likely serves as a quality control pathway that monitors both appropriate tRNA nuclear modification as well as 3’ end maturation. Recent genome-wide studies indicate that as much as 50% of pre-tRNAs may be rapidly degraded by the exosome.10
5’ to 3’ exonucleolyic degradation by the RTD (Rapid Turnover) pathway
The RTD pathway occurs in the nucleus or the cytoplasm, and acts upon tRNAs missing particular modifications on tRNAs. It appears that nuclear enzyme Rat1 and cytoplasmic Xrn1 individually contribute to tRNA turnover.10
Although the mature tRNAs normally have half-lives in the order of hours to days, the defective ones are degraded on the minute to hour timescale, similar to mRNA half-lives.10
Further screening by elongation factor Tu
tRNAs charged with activated amino acids (aa-tRNAs) are screened for correctness by the ribosome protein elongation factor Tu, providing an additional checkpoint for their use for translation purposes. EF-Tu can distinguish both precursor noncognate tRNAs and other incorrectly charged tRNA species.18 Elongation factors, which deliver aa-tRNAs to a ribosome, are known in bacteria as EF-Tu (Elongation Factor Thermo unstable), and eEF1A in eukaryotes. These are not simple biochemicals. In E. coli, EF-Tu is comprised of three functional protein parts known as domain I (amino acids 1–200), domain II (amino acids 209–299) and domain III (amino acids 301–393)19 and in humans the canonical isoform P68104-1 has a length of 462 residues.20
Incorrect expression levels cause disease
Generating too few of some tRNAs would hinder producing the necessary proteins but generating too many imposes a deleterious cost to the organism. Incorrect expression of tRNA has been shown to lead to severe consequences for cells, and to cause human diseases such as cancer, and especially mitochondrial tRNA mutations are responsible for a wide range of disorders.11,21-23
2. Quality control of codon-anticodon selection
The thermodynamics of mRNA–tRNA base pairing are not sufficient to explain the high fidelity and efficiency of aminoacyl-tRNA (aa-tRNA) selection by the ribosome.24 During initial association of ternary complex (aa-tRNA*EFTu* GTP) with the ribosome, codon-anticodon interactions are formed between aa-tRNA and mRNA. Based on differences in hydrogen-bond energies, aa-tRNA interaction with the correct codon is only slightly more stable than a near-cognate one. Such small differences in stability can explain a preference for correct vs incorrect tRNAs of approximately 100:1,24 but the reliability of initial codon-anticodon recognition varies considerably. For example, anomalously high initial misreading in vitro of near-cognate codon by tRNAHis and tRNAGlu has been reported.25 This makes the need for subsequent proof-reading steps important, and indeed in bacteria, the average accuracy of translation is roughly 3,000:1.24 The translation design strategy must optimize a tradeoff between speed vs accuracy of decoding.26
The principle of proofreading was formulated by Hopfield27 and Ninio28 whereby the free energy difference, ΔG0, between enzyme-bound noncognate and cognate substrate can be used in both an initial selection (I) and subsequent proofreading selection (F) to boost the total accuracy (A), A=I × F.29 Laboratory distillation columns are based on a similar logic.30 A general way to implement this strategy in cells is to use the energy provided from hydrolysis of GTP or ATP coupled to discarding of incorrect substrates. In the case of ribosomes, the existence of proofreading was originally verified by biochemical data that implicated at least two distinct steps during tRNA selection (that is, initial selection and proofreading).24
Multistep proofreading is a good design strategy to deal with the obligatory tradeoff between efficiency and accuracy in substrate selection by enzymes occurring when a single transition state reaction is at play. When the literature refers to the proofreading step during codon translation, the aatRNA accommodation step is usually meant, although this can consist of more than one transition step, each contributing an enrichment factor to the desired outcome.24,26,29 Several inter- and intramolecular interactions on aa-tRNA can impact the degree of proofreading, and these appear to be sensitive to the tRNA species.24
Each aa-tRNA is delivered to the ribosome attached to elongation factor Tu (EF-Tu) and GTP. At this point the aatRNA occupies the ‘A site’ on the 30S ribosomal subunit while still interacting with EF-Tu. The free energy differences due to base pair mismatches between the mRNA codon and the tRNA anticodon are too small to provide the observed high accuracy of tRNA selection, and the initial codon-anticodon selection is effective by a factor of about 100:1 vs similar anticodons. At least one follow-up proofreading step is necessary (figure 1).
The codon–anticodon interaction with mRNA triggers GTPase activation and the necessary energy to trigger conformation change through hydrolysis. GTP hydrolysis is followed by release of a phosphate group (Pi) and a conformational change (~100 Å) which precedes peptide bond formation. Importantly, the kinetic energy of transition state of the cognate codon-anticodon interaction is lower than for the erroneous interaction, leading to a proofreading contribution factor of typically between 15 and 60.24
The details of how codon recognition in the 30S subunit leads to GTP hydrolysis are still being worked out and several ribosome components are involved.24,31 Rearrangements occurring during aa-tRNA accommodation include: (1) aatRNA moves from the initial conformation A/T to a so-called ‘Elbow-Accommodated’ (EA) conformation, (2) the aa-tRNA arm accommodates into the A-site, and (3) the 3’-CCA end of the tRNA enters the peptidyl transferase centre (PTC).24
A single kinetic proofreading step is not enough to account for the fidelity of translation,26 and both structural and kinetic data combined with energy landscape calculations indicate that the translation process involves multiple EF-Tu independent metastable states.24,26 The rates of the various processes have been measured by pre-steady-state kinetics and single-molecule fluorescence resonance energy transfer (FRET) experiments.26
3. Cellular location of tRNA processing
Eukaryotes typically have much longer lifetimes and lower population sizes than prokaryotes or archaea. A robust design therefore requires more complex features to proactively enhance survivability.
tRNAs move back and forth between the nucleus and the cytoplasm in yeast, protozoa, and vertebrate cells with the aid of importin and exportin protein complexes (figure 2).32 The tRNA biogenesis steps occur at numerous distinct subcellular locations, as discussed next.
Primary pre-tRNAs form in the nucleolus
In eukaryotes, tRNA transcription occurs in the nucleolus to generate primary pre-tRNAs from which different enzymes then remove the 5’ leader and 3’ tail. Some modifications occur (such as pseudouridylation) and the CCA trinucleotide gets added, generating pre-tRNAs.
Pre-tRNAs are exported to the cytoplasm
Large biomolecules such as carbohydrates, lipids, RNAs, and proteins can only cross the nuclear envelope with the help of special nuclear pore complexes, which consist of dozens of distinct proteins.33 The pre-tRNAs mentioned above are exported to the cytoplasm34 in a step called tRNA primary nuclear export, through nucleopore channels in an energy-dependent mechanism.10 This exportin serves as a quality control check to enhance the delivery of correctly structured and processed pre-tRNAs to the cytoplasm.34 In addition, the level of tRNA available is regulated by tRNA nuclear export.10 This process has to be properly guided, since biomolecules don’t simply diffuse into and out of the nucleus. In the model yeast S. pombe the Xpot protein attaches to the chemically modified TψC (figure 3) and D loops of tRNA and also interacts with both the 5’ and 3’ tRNA termini.
Many nucleosides are modified in the nucleoplasm and at the inner nuclear membrane, as discussed in Part 2. In yeast, tRNAs containing an intron are spliced at the surface of the mitochondria by the SEN complex and the two halves are ligated by another enzyme complex, Rlg1/Trl1.32 In vertebrates, however, intron splicing occurs in the nucleoplasm. tRNA traffic in yeast is also known to be coordinated with the formation of P-bodies in the cytoplasm.10
tRNA retrograde import back into the nucleus
Spliced eukaryote tRNAs undergo a second trafficking step called retrograde nuclear import back into the nucleus, where spliced tRNAs no longer containing introns can be further modified and then be aminoacylated (‘charged’) in the nucleus.10
Charged tRNAs re-exported to the cytoplasm
Finally, charged tRNAs are then re-exported from the nucleus to the cytoplasm to be utilized in translation. A second importin-β family member, Msn5, is now involved. Msn5 likely exports tRNAs that have no introns to the cytoplasm and appears to specifically export those tRNAs that have been spliced in the cytoplasm and imported from the cytoplasm to the nucleus back to the cytoplasm via the tRNA retrograde re-export process.10 Just how the appropriately processed and/or aminoacylated tRNAs are recognized by the tRNA re-export step is still poorly understood.10
Quality control purpose of tRNA retrograde process
Although all tRNAs have a similar tertiary structure, each species has a different RNA sequence and set of biochemical modifications added in the nucleus. There is a limited number of exporters (Los1, Msn5, and at least one unknown exporter) to transfer tRNAs from the nucleus to the cytoplasm.35 Therefore, a given exporter must recognize multiple different tRNA sequences. This can result in errors whereby immature and/ or hypomodified tRNAs are mistakenly exported to the cytoplasm.10
Thus, the tRNA retrograde pathway seems to serve to remove these aberrant tRNAs from the cytoplasm, returning them to the nucleus for repair and/or turnover by the nuclear RTD or TRAMP pathways, providing another mechanism for tRNA quality control.10
tRNA quality control by the TRAMP complex. This tRNA turnover pathway ensures that correctly crafted tRNAs are exported to the cytoplasm. An example flaw would lack m1A58 transformation in tRNAMeti leading to an altered interaction between the tRNA D and T loops. The pathway (which includes the complex of proteins comprised of Trf4, Air1/Air2, and Mtr4) recognizes unmodified tRNAMeti and activates nuclear exosome to degrade it.34
tRNA quality control using a rapid tRNA decay pathway. This occurs in both the nucleus and the cytoplasm. Unmodified tRNAs in yeast are degraded via a second pathway—the rapid tRNA decay (RTD) pathway. For example, tRNAVal(AAC) lacking the m7G modification (figure 3, catalyzed by Trm8) and m5C (figure 3, catalyzed by Trm4) has a half-life typical of only minutes, instead of hours or days typical of fully modified tRNA.34
4. Charged tRNAs participate in autoregulation
A recent RNA immunoprecipitation study revealed for every tested aaRS (aminoacyl-tRNA synthetase) a far better association with the mRNA coding for it than to other mRNAs.36 How might this function in this case? Detailed work using the HisRS protein showed that the binding location was very similar to the tRNAHis anticodon loop. The authors increased the levels of uncharged tRNAHis and this led to increased HisRS translation. Apparently HisRS binding to the mRNA was hindered, and translation could proceed.
Other RNA-binding proteins (RBPs) are known to be autoregulated and some ribosomal proteins are also known to bind their own mRNA.36 mRNA binding leads to translation inhibition, although the mechanisms are not understood yet.36
Discussion and conclusions
tRNAs must satisfy a daunting number of constraints in order to translate all coding codons specified by the genetic code. They all must possess a very narrowly defined folded shape to allow proper recognition by the ribosome components. But simultaneously, in order to obtain sufficient diversity to permit unambiguous recognition by the synthetases to charge the correct activated amino acid to the correct tRNA CCA sequence, an unusually complex series of chemical modifications are needed (discussed in part 2), requiring dozens of specialized enzymes for all organisms.
All forms of life on Earth require a functioning genetic code, and evolution could not prepare one or a few tRNAs first, then countless generations later add another tRNA gene, and then another, since during the evolving process most of the codons could not be translated. There is no simple starting point.
In the section above we drew attention to how cells perform many complex quality control processes, such as:
- Multiple proof-reading steps to ensure correct codon-anticodon pairing
- Two tRNA degradation pathways to ensure that unsuitable tRNAs are eliminated
- Elongation factors can identify incorrectly charged tRNA species
- Degradation of tRNAiMet which lack the m1A58 chemical transformation, indispensable to ensure a stable tRNA structure.
However, the many proteins needed for each of these processes could not have been produced before tRNAs existed, so naturalists must assume none of the quality control mechanisms were available until long after the genetic code was fully functional. Recall from part 1 that the genetic system consists of several mutually dependent subsystems, and without a functional genetic code cells cannot survive. Therefore, a putative genetic system with no quality control anywhere would mean uncontrollable errors generated during each of the processes involved, such as during DNA to mRNA transcription, DNA to DNA replication, charging of amino acids by synthetases, translation of codons, folding of proteins, etc.
Consider the impact of multiple sources of errors. Part 11 and part 29 of this series showed that without many chemical transformations by protein-based enzymes most tRNAs are unsuitable. Chemical transformations are particularly important to ensure that Start and Stop codons are not misinterpreted by coding triplets having only one nucleotide difference.37 Now in part 3 we see that the flawed ones would not be recognized and degraded in an evolutionary scenario. Consider the outcome. tRNAiMet is used to identify where mRNA translation is to begin. If not positioned just right the 3-nucleotide reading frame to translate codon after codon won’t align correctly. (Of course, in a primitive ribosome it is not clear why each tRNA would only find its cognate codon right next to the just translated codon instead of anywhere on the mRNA, so utter translational chaos would be expected). Now we learn that an enzymatic transformation on tRNAiMet is necessary but that a quality control process to identify the malfunctional ones would not have existed initially.
Finally, without the necessary protein-based molecular machines, the correct amount of the right kinds of tRNAs under the correct circumstances cannot be regulated. The result would be a genetic system hopelessly lacking key components, with high error rates for each key process, over and under generating the proportion of biomolecules needed.
We conclude in part 4 of this series38 that evolution cannot create a minimally functional genetic code based only on our analysis of the tRNA subsystem, and also that evolving various additional complex tRNA-related features lies beyond what theoretical evolutionary processes could achieve.
References and notes
- Truman, R., The Surprisingly Complex tRNA Subsystem: part 1—generation and chemical modifications, J. Creation 34(3):80–86, 2020. Return to text.
- Korostelev, A., Trakhanov, S., Laurberg, M., and Noller, H.F., Crystal structure of a 70S ribosome-tRNA complex reveals functional interactions and rearrangements, Cell 126:1065–1077, 2006. Return to text.
- Hou, Y.M., CCA addition to tRNA: implications for tRNA quality control, IUBMB Life 62(4):251–260, 2010. Return to text.
- Tomita, K. and Yamashita, S., Molecular mechanisms of template-independent RNS polymerization by tRNA nucleotidyltransferases, Frontiers in Genetics 5(36)1:12, 2014. Return to text.
- Hou, Y.M., CCA addition to tRNA: implications for tRNA quality control, IUBMB Life 62(4):251–260, 2010. Return to text.
- Betat, H. and Mörl, M., The CCA-adding enzyme: a central scrutinizer in tRNA quality control, Bioessays, 37(9):975–982, 2015. Return to text.
- Wilusz, J.E., Whipple, J.M., Phizicky, E.M., and Sharp, P.A., tRNAs marked with CCACCA are targeted for degradation, Science 334(6057):817–821, 2011. Return to text.
- Wellner, K., Betat, H., and Mörl, M., A tRNA’s fate is decided at its 3’ end: collaborative actions of CCA-adding enzyme and RNases involved in tRNA processing and degradation, Biochimica et Biophysica Acta (BBA)—Gene Regulatory Mechanisms 1861(4):433–441, 2018. Return to text.
- Truman, R., The surprisingly complex tRNA subsystem: part 2—biochemical modifications, J. Creation 34(3):87–94, 2020. Return to text.
- Hopper, A.K., Transfer RNA post-transcriptional processing, turnover, and subcellular dynamics in the Yeast Saccharomyces cerevisiae, Genetics 194:43–67, May 2013. Return to text.
- Pan, T., Modifications and functional genomics of human transfer RNA, Cell Res 28:395–404, 2018. Return to text.
- Huber, S.M., Leonardi, A., Dedon, P.C., and Begley, T.J., The versatile roles of the tRNA epitranscriptome during cellular responses to toxic exposures and environmental stress, Toxics 7(1):17, 2019. Return to text.
- Isoacceptors are tRNAs having different anticodons but are charged with the same amino acid. (Anticodons are the three tRNA nucleotides which bind to an mRNA codon). Isoacceptors are used to translate synonymous codons which represent the same amino acid in the genetic code. Return to text.
- Zhang, C. and Jia, G., Reversible RNA modification N1-methyladenosine (m1A) in mRNA and tRNA, Genomics Proteomics Bioinformatics 16:155–161, 2018. Return to text.
- Slobodin, B. et al., Transcription dynamics regulate poly(a) tails and expression of the RNA degradation machinery to balance mRNA levels, Molecular Cell 78(3):434–444.e5, 2020. Return to text.
- This illustrates the principle that biochemical features often need to be understood as informative signals for logic processing and not mere physical interactions. Return to text.
- Pan, K., Huang, Z., Lee, J.T., and Wong, C., Current perspectives on the role of TRAMP in nuclear RNA surveillance and quality control, Research and Reports in Biochemistry 5:111–117, 2015. Return to text.
- Ibba, M. and Söll, D., Aminoacyl-tRNAs: setting the limits of the genetic code, Genes & Dev. 18:731–738, 2004. Return to text.
- Kjeldgaard, M. and Nyborg, J., Refined structure of elongation factor EF-Tu from Escherichia coli, J. Molecular Biology 223:721–742, 1992. Return to text.
- From UniProtKB Entry P68104: uniprot.org/uniprot/P68104. Return to text.
- Torres, A.G., Reina, O., Attolini, C.S-O., and Ribas de Pouplana, L., Differential expression of human tRNA genes drives the abundance of tRNA-derived fragments, PNAS 116(17):8451–8456, 23 April 2019. Return to text.
- Kwon, N.H., Lee, M.R., Kong, J. et al., Transfer-RNA-mediated enhancement of ribosomal proteins S6 kinases signaling for cell proliferation, RNA Biol. 15(4–5):635–648, 2017. Return to text.
- Goodarzi, H. et al., Modulated expression of specific tRNAs drives gene expression and cancer progression, Cell 165:1416–1427, 2016. Return to text.
- Noel, J.K. and Whitford, P.C., How EF-Tu can contribute to efficient proofreading of aa-tRNA by the ribosome, Nature Communications 7:13314, 2016. Return to text.
- Zhang, J., Ieong, K.-W., Mellenius, H., and Ehrenberg, M., Proofreading neutralizes potential error hotspots in genetic code translation by transfer RNAs, RNA 22:896–904, 2016. Return to text.
- Savir, Y. and Tlusty, T., The ribosome as an optimal decoder: a lesson in molecular recognition, Cell 153:471–479, 2013. Return to text.
- Hopfield, J.J., Kinetic proofreading: A new mechanism for reducing errors in biosynthetic processes requiring high specificity, PNAS 71(10):4135–4139, 1974. Return to text.
- Ninio, J., Kinetic amplification of enzyme discrimination, Biochimie 57(5):587–595, 1975. Return to text.
- Leong, K.-W., Uzuna, Ü., Selmera, M., and Ehrenberg, M., Two proofreading steps amplify the accuracy of genetic code translation, PNAS 113(48):13744–13749, 2016. Return to text.
- A conceptual illustration is the design of distillation columns for use in laboratories or chemical manufacturing. Surfaces are provided in the column between which the liquid being distilled can equilibrate. The more ‘theoretical plates’ the better the discrimination between substances having similar boiling points, but of course the more time and energy are then required. Return to text.
- Villa, E., et al., Ribosome-induced changes in elongation factor Tu conformation control GTP hydrolysis, PNAS 106(4):1063–1068, 2009. Return to text.
- Hopper, A.K. and Nostramo, R.T., tRNA processing and subcellular trafficking proteins multitask in pathways for other RNAs, Frontier in Genetics 10(96):1–14, 2019. Return to text.
- Lin, D.H., et al., Architecture of the symmetric core of the nuclear pore, Science 352(6283):aaf1015, 2016. Return to text.
- Hopper, A.K., Pai, D.A., and Engelke, D.R., Cellular dynamics of tRNAs and their genes, FEBS Letters 584:310–317, 2010. Return to text.
- Murthi, A., Shaheen, H.H., Huang, H-Y, Preston, M.A., Lai, T-P, Phizicky, E.M., and Hopper, A.K., Regulation of tRNA bidirectional nuclear-cytoplasmic trafficking in Saccharomyces cerevisiae, Mol. Biol. of the Cell. 21:639–649, 2010. Return to text.
- Levi, O. and AravaI, Y., mRNA association by aminoacyl tRNA synthetase occurs at a putative anticodon mimic and autoregulates translation in response to tRNA levels, PLoS Biol 17(5):e3000274, 2019. Return to text.
- Shepherd, J. and Ibba, M., Bacterial transfer RNAs, FEMS Microbiology Reviews 39(3):280–300, 2015. Return to text.
- Truman, R., The surprisingly complex tRNA subsystem: part 4—tRNA fragments regulate processes, J. Creation 35(1):70–76, 2020. Return to text.