160
Biochemical Society Transactions (2014) Volume 42, part 1
Controlling translation elongation efficiency: tRNA regulation of ribosome flux on the mRNA Barbara Gorgoni*1 , Elizabeth Marshall*†1 , Matthew R. McFarland*1 , M. Carmen Romano*† and Ian Stansfield*2 *Institute of Medical Sciences, University of Aberdeen, Aberdeen AB25 2ZD, U.K. †Institute of Complex Systems and Mathematical Biology, King’s College, University of Aberdeen, Aberdeen AB24 3UE, U.K.
Biochemical Society Transactions
www.biochemsoctrans.org
Abstract Gene expression can be regulated by a wide variety of mechanisms. One example concerns the growing body of evidence that the protein-production rate can be regulated at the level of translation elongation by controlling ribosome flux across the mRNA. Variations in the abundance of tRNA molecules cause different rates of translation of their counterpart codons. This, in turn, produces a variable landscape of translational rate across each and every mRNA, with the dynamic formation and deformation of ribosomal queues being regulated by both tRNA availability and the rates of translation initiation and termination. In the present article, a range of examples of tRNA control of gene expression are reviewed, and the use of mathematical modelling to develop a predictive understanding of the consequences of that regulation is discussed and explained. These findings encourage a view that predicting the protein-synthesis rate of each mRNA requires a holistic understanding of how each stage of translation, including elongation, contributes to the overall protein-production rate.
Introduction In all organisms, the dynamic and steady-state levels of cellular proteins can be controlled at multiple stages of the gene-expression process, through regulation of transcription, mRNA processing and turnover, mRNA translation, and protein turnover. Of these various stages, the translation of mRNA into protein by the ribosomal machinery is a factorylike assembly-line process that must be tightly regulated. Translation initiation, where ribosomes join the mRNA, must be orchestrated with the capacity of translation elongation to manufacture the polypeptide. For example, a very low rate of translation elongation, relative to the translation initiation rate, will cause ribosome queues to be established on the mRNA. Given sufficient time and an appropriate mRNA stability, this queuing will lead eventually to ribosome occupancy and occlusion of the 5 -end of that mRNA, causing a reduced translation efficiency. Both translation initiation and elongation are therefore subject to regulation; translation initiation can be regulated by a variety of mechanisms, using secondary structures in the mRNA 5 -UTR [1] or using phosphorylation of translation initiation factors to control ribosome joining [2]. The presence of regulatory upstream ORFs in the 5 -UTR can also control ribosomal access to the ORF [3]. Once ribosomes reach the ORF and begin elongation, their flux through the bulk of the ORF is regulated, in part, by the availability of charged tRNA species, which deliver amino acids to the Key words: ribosomal pausing, Saccharomyces cerevisiae, translational regulation, translation elongation, tRNA, tRNA modification. Abbreviations: Elp/elp, Elongator complex protein; mcm5 , methoxycarbonylmethyl group at C-5; TASEP, totally asymmetric simple exclusion process; s2 , thiol at C-2. 1 2
These authors contributed equally to this review. To whom correspondence should be addressed (email i.stansfi[email protected]).
C The
C 2014 Biochemical Society Authors Journal compilation
growing polypeptide chain. Particularly in microbial species, there can be considerable (>15-fold) variation in the abundance of different tRNA isotypes. Selection for translational efficiency has ensured that this tRNA variation is correlated with biased codon usage [4] of the cognate codon(s) in the transcriptome. This balances the codon demand/tRNA supply relationship (Figure 1). Despite such balancing, individual rare codons will nevertheless be translated slowly by correspondingly rare tRNAs; such translational pauses can potentially cause ribosome queues [5–7] (Figure 2). Although evidence exists that rare codons can cause translational pauses, predicting whether any given slow elongation step can inhibit protein synthesis by promoting queue formation back to the mRNA 5 -end is challenging, and depends upon the complex dynamics of queue formation and dissolution, combined with the initiation rate on that particular mRNA. The recent development of mathematical models of translation has, however, begun to make possible the simulation of ribosome flow along an mRNA through a field of codons at variable rates, allowing the global consequences of such regulation to be predicted [34,35,37,42,45]. In the present review, known examples of the tRNA-mediated control of elongation are considered alongside the use of mathematical models to predict the systems biological effects of this regulation.
Regulation of gene expression by tRNA One of the best established examples of tRNA regulation of gene expression is found in the Streptomyces genus. Mutation of the bldA gene in Streptomyces coelicolor produces cells that exhibit normal vegetative growth, but cannot form an aerial mycelium (giving these mutants their ‘bald’ phenotype) or Biochem. Soc. Trans. (2014) 42, 160–165; doi:10.1042/BST20130132
Translation UK 2013
Figure 1 tRNA abundance is matched to translational demand tRNA abundance, or ‘supply’, was estimated by expressing the tRNA gene copy number for a given isotype as a percentage of all S. cerevisiae tRNA gene copies. For each tRNA type, its corresponding cognate codon(s) were identified and the total number of those codons within the transcriptome enumerated and expressed as a percentage of all codons within the mRNA population using SAGE (serial analysis of gene expression) data [47] to gauge the mRNA copy numbers for each gene transcript type (I. Stansfield, unpublished work).
produce antibiotics and other secondary metabolites [8–10]. The bldA gene encodes the only copy of a UUA-cognate leucine tRNA (tRNALeu UUA) [9]. Of the 7800 genes in the S. coelicolor genome, only approximately 150 of them contain UUA codons, and many of them are involved in differentiation or antibiotic production, but do not appear to
have a role in normal vegetative growth [11,12]. This suggests that either the abundance or the translation efficiency of the UUA-decoding tRNA is compromised in the bldA mutant, leading to down-regulation of UUA-containing genes. In the wild-type cell, bldA tRNA is developmentally regulated, accumulating late in the batch growth cycle to coincide with the production of hyphae and antibiotics [13]. The failure to form a mycelium in bldA mutants is attributed to the presence of the cognate UUA codon in the adpA (also known as bldH) gene. This gene encodes a transcription factor that activates many of the genes responsible for differentiation and antibiotic production [14]. Consequently, mutations in the adpA gene also produce a ‘bald’ phenotype, with the failure to produce aerial hyphae and secondary metabolites. If the adpA UUA codon is mutated to another leucine codon, such as UUG, normal aerial hyphal growth is restored in bldA mutants, suggesting that its expression levels are regulated by the efficiency of UUA decoding [12]. Intriguingly, more recent experiments carried out in another Streptomycete (Streptomyces griseus) found that adpA is required for efficient bldA transcription [15]. The AdpA protein binds sites upstream of the bldA promoter and activates transcription of the tRNA. This sets up a positive-feedback loop between the two where low-level expression of tRNALeu UUA leads to AdpA protein production, in turn further activating bldA transcription, eventually producing sufficient levels of the protein to trigger morphological differentiation. Another example of gene regulation by a low-abundance tRNA is found in Saccharomyces cerevisiae, where mutations
Figure 2 Low-abundance tRNAs cause lower decoding rates The 30–40 genes encoding the range of tRNA isotypes are multi-copy in most organisms, but this copy number varies (ranging in yeast from one to 16 copies/genome), causing corresponding variations in tRNA abundance. Infrequently used ‘rare’ codons, represented by the light grey codon box, recognized by low-abundance tRNAs (light grey), will be decoded more slowly by the translation apparatus, potentially resulting in the formation of ribosome queues at such codon positions. Conversely, the abundant tRNAs (dark grey) will decode their cognate codon more quickly.
C The
C 2014 Biochemical Society Authors Journal compilation
161
162
Biochemical Society Transactions (2014) Volume 42, part 1
in the single-copy SUP70 gene [also known as CDC65 (celldivision cycle 65)], encoding a rare tRNAGln CUG , cause an unregulated pseudohyphal growth form, in contrast with the normal budding growth pattern of yeast [16]. The recessive sup70-65 mutations are located in the anticodon stem of the tRNA and cause base mispairing, which presumably leads to the observed instability and altered decoding properties of the mutant tRNA [16,17]. The use of reporter genes containing tandem arrays of the target CAG codon for the SUP70 tRNA showed clearly that the mutant tRNA was defective in its ability to decode CAG codons. This, in turn, suggests that one or more of the mRNAs in yeast whose protein products regulate pseudohyphal differentiation must be enriched in CAG codons, and be particularly sensitive to the abundance of the CAG-decoding SUP70 tRNA.
Regulation of gene expression by tRNA modification In addition to the regulation of gene expression by tRNA abundance, there is also evidence that tRNA modification status can specifically regulate subsets of yeast genes. tRNAs from all organisms are extensively modified and, interestingly, in most tRNAs containing a uridine in the wobble position (U34 ), this nucleoside is always modified (reviewed in [18]). In particular, the pyrimidine-rich tRNALys UUU , tRNAGlu UUC and tRNAGln UUG contain a double U34 modification: mcm5 (methoxycarbonylmethyl group at C-5) and s2 (thiol at C2). These tRNAs, when unmodified, present an unfavourable codon–anticodon pairing, and the modifications improve ribosome binding and the efficiency and accuracy of both cognate and wobble decoding [19]. The crucial cellular function of these tRNA modifications is highlighted by the fact that they are targeted by two different toxins, zymocin [20,21] and PaT (Pichia acaciae killer toxin) [22], which cleave the modified, but not the unmodified, nucleosides and cause cell death. Indeed, it was through a screen of yeast mutants resistant to zymocin that the genes responsible for the addition of the mcm5 group were identified [23]. These genes encode Elongator, a highly conserved multisubunit complex. Elongator is composed of six subunits, Elp1–Elp6 (Elongator complex protein 1–6) and associates with several other protein factors (reviewed in [24]). Importantly, deletion of each of the subunits in yeast (elp deletions) leads to similar phenotypes, indicating that they are all essential for Elongator function [24]. The Elongator complex was first identified in S. cerevisiae as a transcription elongation factor owing to its association with hyperphosphorylated RNA polymerase II and its histone acetyltransferase activity. It was subsequently implicated in a variety of cellular processes, including exocytosis, kinase signalling, transcriptional gene silencing and protein acetylation (for reviews, see [24,25]). This suggested that Elongator had complex multifunctional roles, in addition to its tRNA-modification function. However, a pivotal study in S. cerevisiae showed that several Elongator mutant phenotypes could be rescued by simply overexpressing two of the tRNAs that normally contain the U34 mcm5 and C The
C 2014 Biochemical Society Authors Journal compilation
s2 modifications [26]. Artificially increasing the levels of unmodified tRNALys UUU and tRNAGln UUG in elp-deleted yeast cells could suppress temperature-sensitivity and defects in chromatin remodelling, transcriptional activation and exocytosis [26]. Similarly, a more recent study demonstrated that elp-dependent defects in telomeric gene silencing and the DNA-damage response were also suppressed by overexpression of the same tRNAs [27]. These data strongly suggest that the Elongator’s main biological role is the regulation of translation, through modification of tRNAs. The pleiotropic phenotypes caused by its deletion are most probably due to reduced translation of several target mRNAs. This is corroborated by the observation that ncs2 (needs Cla4 to survive 2)-deletion mutants, which are missing the U34 s2 , but not the mcm5 , modification phenocopy the elp mutants and are rescued by overexpression of the same tRNAs [26]. Furthermore, these modifications appear to control the efficiency rather than the specificity of decoding, as it is the increased abundance of unmodified tRNAs that is able to restore the wild-type phenotype. A first intriguing insight into the mechanism by which tRNA modifications can regulate translation of specific mRNAs was provided by the work in fission yeast of Hermand and colleagues [28]. A screen for translation defects in an elp3-deletion mutant revealed that, although only a modest global effect on the proteome was observed, the translation of proteins belonging to specific functional groups was significantly affected, including proteins involved in cell division. Analysis of codon content of these groups and, in particular, of the cell-cycle kinase Cdr2, revealed an increased usage for the lysine AAA codon over the synonymous AAG codon, making these mRNAs more dependent on modified tRNALys UUU . For one functional group of mRNAs, this was also true for the glutamate GAA, compared with the GAG, codon. Cdcr2 expression was significantly reduced in the elp mutant background and its deletion mutant phenocopied the elp3 deletion. Importantly, replacing the lysine AAA codons with lysine AAG made this mRNA insensitive to the absence of a functional Elongator. These findings were extended to budding yeast and nematodes, where biased codon usage was observed in mRNAs related to telomeric gene silencing and α-tubulin acetylation respectively [29]. These data provide a strong rationale for the role of Elongator in translation regulation. Moreover, it suggests that specific genes and functional categories of genes may be enriched in codons targeted by tRNAs with Elp modifications; the coupling of tRNA modification with enhanced content of particular codons thus offers an evolutionarily conserved mechanism for the co-ordinated expression of specific group of genes.
Modelling ribosomal flux: theory and simulation Understanding how particular tRNAs, or tRNA modifications, can specifically regulate groups of genes can be approached using mathematical models of translation, where genome-wide simulation of translation is now possible.
Translation UK 2013
Some whole-cell-type models focus on the competition for translational resources imposed by the mRNA population, namely ribosomes and charged tRNAs, a strategy that allows global and mRNA-specific testing of the translational synthesis capacity of the cell [30]. Another type of model used with some success to simulate translation is the TASEP (totally asymmetric simple exclusion process). First used in 1968 to describe protein synthesis, it has since become a paradigmatic model in non-equilibrium statistical physics [31–33]. It has been applied to describe a broad range of different systems, such as transport by molecular motors [34,35] and traffic flow [33,36]. In its simplest form, an mRNA can be treated as a one-dimensional homogenous lattice of length N sites (representing codons), where singlesite particles take the role of ribosomes. The TASEP is a stochastic model, which mimics the inherent randomness of a tRNA diffusing through the cytoplasm to bind the mRNA via the ribosome. The particles enter on the left-hand side of the lattice with a given initiation rate α, elongate (hop) from site i to i + 1, if i + 1 is empty, with rate ki , and exit the lattice on the right-hand side with the termination rate β. The status of each site can be indicated with occupation numbers ni , where i = 1 . . . N. If site i is empty, ni = 0, and if it is occupied then ni = 1. The density of particles at this site can then be given as the time-averaged occupancy . Averaging this by the number of sites in the lattice gives the average density of the entire system . For a homogeneous lattice with ki = k for i = 1 . . . N, it can be shown that, depending upon the relative values of the initiation rate α, termination rate β and elongation rate k, there are three characteristic phases in which the lattice can exist, determined by the value of the density ρ and particle flow J: (i) low density, with α as the rate-limiting factor for J; (ii) high density, with β as the ratelimiting factor; and (iii) maximal current, with k restricting the particle current. Both the low- and high-density phases are suboptimal for particle flow (and therefore for protein production on mRNAs). In the maximal current phase, the density of particles is such that the particle flow reaches its maximum value; in this phase, the mRNA would synthesize proteins with the highest efficiency possible (Figure 3). In more recent years, the TASEP has been extended to model translation more realistically. For example, it has been generalized to particles with a realistic extended footprint of the ribosome [37]. Moreover, the mechanochemical cycle of the ribosome has been included in the TASEP approach, where it has been condensed into two main steps: capture of the correct tRNA and translocation to the next codon. This extended model then introduces twostate ribosomes (without and with tRNA), and it takes into consideration the important fact that when a ribosome is waiting for the next codon to become free, it can use the time to capture the tRNA (this fact was neglected in the simple TASEP approach). In addition, there is evidence that the elongation rate associated with each codon correlates with the concentration of tRNAs available to decode it, and the time taken for the correct tRNA to bind has been
Figure 3 The relative rates of initiation and termination define different ribosome density regimes on the mRNA Using the TASEP model of translating ribosomes, the initiation rate α and termination rate β can be controlled independently. Three different domains within the α–β phase plot are defined: low density, characterized by well-spaced ribosomes and low translational efficiency; high density, characterized by densely packed ribosomes and saturated protein production rate; and maximal current, where the optimum ribosomal loading produces the ideal spacing for efficient protein production at maximal rate.
shown to be the rate-limiting step in elongation rather than the translocation step [38,39]. It has been shown that the dynamics of the ribosomes in this biologically realistic regime change considerably with respect to the TASEP, and the twostate ribosome dynamics leads to a substantially enlarged maximal current phase in the α–β parameter space [40]. Hence this model predicts that the optimal region for protein synthesis is accessible at lower values of the initiation rate. The configuration of slow codons is an important factor in protein production. Grouping the slow codons into clusters (a ‘bottleneck’) has a profound effect on particle flow, with closely spaced slow codons reducing protein synthesis by a factor of ∼2–4 [41]. The positioning of this bottleneck determines whether an mRNA undergoes an abrupt increase in particle density by transitioning from low to high density (a Type I mRNA), or no phase transition, where J depends smoothly on the initiation rate α (Type II) [42]. A Type I mRNA has slow codons either individually or clustered somewhere remote from the 5 -end of the mRNA (far enough to allow the initiation-rate-dependent formation of a ribosome queue), leading to an abrupt change in density as ribosomal ‘traffic jams’ form. In contrast, a Type II mRNA either has no slow codons or a ‘ramp’ at the beginning of the ORF, a feature conserved across multiple species, potentially to reduce the queuing of ribosomes at some point downstream [43,44]. mRNAs can thus be classified according to the type of transition their sequence undergoes as initiation increases [42]. Genome-wide application of such a classification has also shown a strong correlation between an mRNA’s transition type and biological function; for example, those synthesising regulatory proteins are overrepresented in C The
C 2014 Biochemical Society Authors Journal compilation
163
164
Biochemical Society Transactions (2014) Volume 42, part 1
the ‘abrupt’ Type I class [45]. Furthermore, simulations of translation across the entire transcriptome of S. cerevisiae using the two-state ribosome model [40] have now been used for the first time to identify the range of in vivo translation initiation rates for all yeast mRNAs, showing a distribution of rates with the majority clustered around one ribosome joining a given mRNA type per 5–10 s [45]. Interestingly, these calculations, based on experimental data and translation simulation, were well matched with later estimates obtained using a whole-cell model of translation [46]. Models of the types described above are now being subjected to experimental validation. For example, a two-state TASEP simulation of translation on a real mRNA containing tandem arrays of CAG codons was able to successfully, and quantitatively, predict the translation-inhibiting effects of a yeast sup70-65 mutation in tRNAGln CUG caused by ribosomal queuing back to the mRNA 5 -end [17]. Mathematical modelling of mRNA translation is thus a rapidly developing field with the potential in the very near future to provide an in silico test bed for the translational regulation of gene expression in a range of species. The use of such simulations will provide access to quantitative predictions that would otherwise be experimentally challenging to obtain.
Future prospects The development of mathematical models of translation provides revealing new insights into how this cellular production line has evolved to operate smoothly without the inefficiency of ribosomal queuing. Moreover, the expectation is that such approaches, in future, will encourage a more integrated view of translation without artificial modularization into the initiation, elongation and termination stages. Progress in this area will also undoubtedly bring important insights into the rational design of gene sequences for biotechnological protein expression processes, side-stepping the slightly heuristic approaches currently used to optimize gene expression.
Funding Work in the laboratories of I.S. and M.C.R. has been supported by the Biotechnology and Biological Sciences Research Council [grant numbers BB/G010722/1 (to I.S. and M.C.R.), BB/I020926/1 (to I.S.) and BB/F00513X/1 (to M.C.R.)].
References 1 Hentze, M.W. and Kuhn, L.C. (1996) Molecular control of vertebrate iron metabolism: mRNA-based regulatory circuits operated by iron, nitric oxide, and oxidative stress. Proc. Natl. Acad. Sci. U.S.A. 93, 8175–8182 2 Proud, C.G. (2006) Regulation of protein synthesis by insulin. Biochem. Soc. Trans. 34, 213–216 3 Hinnebusch, A.G. (2005) Translational regulation of GCN4 and the general amino acid control of yeast. Annu. Rev. Microbiol. 59, 407–450 4 Sharp, P.M. and Li, W.H. (1987) The codon adaptation index: a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res. 15, 1281–1295 C The
C 2014 Biochemical Society Authors Journal compilation
5 Belcourt, M.F. and Farabaugh, P.J. (1990) Ribosomal frameshifting in the yeast retrotransposon ty: tRNAs induce slippage on a 7 nucleotide minimal site. Cell 62, 339–352 6 Ramachandiran, V., Kramer, G., Horowitz, P.M. and Hardesty, B. (2002) Single synonymous codon substitution eliminates pausing during chloramphenicol acetyl transferase synthesis on Escherichia coli ribosomes in vitro. FEBS Lett. 512, 209–212 7 Buchan, J.R. and Stansfield, I. (2007) Halting a cellular production line: responses to ribosomal pausing during translation. Biol. Cell 99, 475–487 8 Merrick, M.J. (1976) A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 96, 299–315 9 Lawlor, E.J., Baylis, H.A. and Chater, K.F. (1987) Pleiotropic morphological and antibiotic deficiencies result from mutations in a gene encoding a tRNA-like product in Streptomyces coelicolor A3(2). Genes Dev. 1, 1305–1310 10 Chater, K.F. and Chandra, G. (2008) The use of the rare UUA codon to define “expression space” for genes involved in secondary metabolism, development and environmental adaptation in Streptomyces. J. Microbiol. 46, 1–11 11 Bentley, S.D., Chater, K.F., Cerdeno-Tarraga, A., Challis, G. L., Thomson, N.R., James, K.D., Harris, D.E., Quail, M.A., Kieser, H., Harper, D. et al. (2002) Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417, 141–147 12 Takano, E., Tao, M., Long, F., Bibb, M.J., Wang, L., Li, W., Buttner, M.J., Bibb, M.J., Deng, Z.X. and Chater, K.F. (2003) A rare leucine codon in adpA is implicated in the morphological defect of bldA mutants of Streptomyces coelicolor. Mol. Microbiol. 50, 475–486 13 Leskiw, B.K., Mah, R., Lawlor, E.J. and Chater, K.F. (1993) Accumulation of bldA-specified tRNA is temporally regulated in Streptomyces coelicolor A3(2). J. Bacteriol. 175, 1995–2005 14 Ohnishi, Y., Yamazaki, H., Kato, J., Tomono, A. and Horinouchi, S. (2005) AdpA, a central transcriptional regulator in the A-factor regulatory cascade that leads to morphological development and secondary metabolism in Streptomyces griseus. Biosci. Biotechnol. Biochem. 69, 431–439 15 Higo, A., Horinouchi, S. and Ohnishi, Y. (2011) Strict regulation of morphological differentiation and secondary metabolism by a positive feedback loop between two global regulators AdpA and BldA in Streptomyces griseus. Mol. Microbiol. 81, 1607–1622 16 Murray, L.E., Rowley, N., Dawes, I.W., Johnston, G.C. and Singer, R.A. (1998) A yeast glutamine tRNA signals nitrogen status for regulation of dimorphic growth and sporulation. Proc. Natl. Acad. Sci. U.S.A. 95, 8619–8624 17 Kemp, A.J., Betney, R., Ciandrini, L., Schwenger, A.C.M., Romano, M.C. and Stansfield, I. (2013) A yeast tRNA mutant that causes pseudohyphal growth exhibits reduced rates of CAG codon translation. Mol. Microbiol. 87, 284–300 18 Agris, P.F., Malkiewicz, A., Kraszewski, A., Everett, K., Nawrot, B., Sochacka, E., Jankowska, J. and Guenther, R. (1995) Site-selected introduction of modified purine and pyrimidine ribonucleosides into RNA by automated phosphoramidite chemistry. Biochimie 77, 125–134 19 Vendeix, F.A., Murphy, IV, F.V., Cantara, W.A., Leszczynska, G., Gustilo, E.M., Sproat, B., Malkiewicz, A. and Agris, P.F. (2012) Human tRNALys3 UUU is pre-structured by natural modifications for cognate and wobble codon binding through keto–enol tautomerism. J. Mol. Biol. 416, 467–485 20 Lu, J., Huang, B., Esberg, A., Johansson, M.J. and Bystrom, A.S. (2005) The Kluyveromyces lactis γ -toxin targets tRNA anticodons. RNA 11, 1648–1654 21 Jablonowski, D., Zink, S., Mehlgarten, C., Daum, G. and Schaffrath, R. (2006) tRNAGlu wobble uridine methylation by Trm9 identifies elongator’s key role for zymocin-induced cell death in yeast. Mol. Microbiol. 59, 677–688 22 Klassen, R., Paluszynski, J.P., Wemhoff, S., Pfeiffer, A., Fricke, J. and Meinhardt, F. (2008) The primary target of the killer toxin from Pichia acaciae is tRNAGln . Mol. Microbiol. 69, 681–697 23 Huang, B., Johansson, M. J. and Bystrom, A.S. (2005) An early step in wobble uridine tRNA modification requires the elongator complex. RNA 11, 424–436 24 Glatt, S. and Muller, ¨ C.W. (2013) Structural insights into elongator function. Curr. Opin. Struct. Biol. 23, 235–242 25 Svejstrup, J. Q. (2007) Elongator complex: how many roles does it play? Curr. Opin. Cell Biol. 19, 331–336 26 Esberg, A., Huang, B., Johansson, M.J.O. and Bystrom, ¨ A.S. (2006) Elevated levels of two tRNA species bypass the requirement for Elongator complex in transcription and exocytosis. Mol. Cell 24, 139–148
Translation UK 2013
27 Chen, C., Huang, B., Eliasson, M., Ryden, P. and Bystrom, A.S. (2011) Elongator complex influences telomeric gene silencing and DNA damage response by its role in wobble uridine tRNA modification. PLoS Genet. 7, e1002258 28 Bauer, F., Matsuyama, A., Candiracci, J., Dieu, M., Scheliga, J., Wolf, D.A., Yoshida, M. and Hermand, D. (2012) Translational control of cell division by Elongator. Cell Rep. 1, 424–433 29 Bauer, F. and Hermand, D. (2012) A coordinated codondependent regulation of translation by Elongator. Cell Cycle 11, 4524–4529 30 Chu, D. and von der Haar, T. (2012) The architecture of eukaryotic translation. Nucleic Acids Res. 40, 10098–10106 31 MacDonald, C.T., Gibbs, J.H. and Pipkin, A.C. (1968) Kinetics of biopolymerization on nucleic acid templates. Biopolymers 6, 1–5 32 Derrida, B., Domany, E. and Mukamel, D. (1992) An exact solution of a one-dimensional asymmetric exclusion model with open boundaries. J. Stat. Phys. 69, 667–687 33 Proeme, A., Blythe, R.A. and Evans, M.R. (2011) Dynamical transition in the open-boundary totally asymmetric exclusion process. J. Phys. A 44, 035003 34 Chowdhury, D., Schadschneider, A. and Nishinari, K. (2005) Physics of transport and traffic phenomena in biology: from molecular motors and cells to organisms. Phys. Life Rev. 2, 318–352 35 Neri, I., Kern, N. and Parmeggiani, A. (2013) Modeling cytoskeletal traffic: an interplay between passive diffusion and active transport. Phys. Rev. Lett. 110, 098102 36 Popkov, V., Santen, L., Schadschneider, A. and Schutz, G.M. (2001) Empirical evidence for a boundary-induced nonequilibrium phase transition. J. Phys. A 34, L45–L52 37 Shaw, L.B., Zia, R.K. and Lee, K.H. (2003) Totally asymmetric exclusion process with extended objects: a model for protein synthesis. Phys. Rev. E: Stat. Nonlin. Soft Matter Phys. 68, 021910
38 Varenne, S., Buc, J., Lloubes, R. and Lazdunski, C. (1984) Translation is a non-uniform process: effect of tRNA availability on the rate of elongation of nascent polypeptide chains. J. Mol. Biol. 180, 549–576 39 Liljenstrom, H. and von Heijne, G. (1987) Translation rate modification by preferential codon usage: intragenic position effects. J. Theor. Biol. 124, 43–55 40 Ciandrini, Stansfield, I. and Romano. (2010) Role of the particle’s stepping cycle in an asymmetric exclusion process: a model of mRNA translation. Phys. Rev. E: Stat. Nonlin. Soft Matter Phys. 81, 051904 41 Chou, T. and Lakatos, G. (2004) Clustered bottlenecks in mRNA translation and protein synthesis. Phys. Rev. Lett. 93, 198101 42 Romano, M. C., Thiel, M., Stansfield, I. and Grebogi, C. (2009) Queueing phase transition: theory of translation. Phys. Rev. Lett. 102, 198104 43 Tuller, T., Carmi, A., Vestsigian, K., Navon, S., Dorfan, Y., Zaborske, J., Pan, T., Dahan, O., Furman, I. and Pilpel, Y. (2010) An evolutionarily conserved mechanism for controlling the efficiency of protein translation. Cell 141, 344–354 44 Pechmann, S. and Frydman, J. (2013) Evolutionary conservation of codon optimality reveals hidden signatures of cotranslational folding. Nat. Struct. Mol. Biol. 20, 237–243 45 Ciandrini, L., Stansfield, I. and Romano, M. C. (2013) Ribosome traffic on mRNAs maps to gene ontology: genome-wide quantification of translation initiation rates and polysome size regulation. PLoS Comput. Biol. 9, e1002866 46 Shah, P., Ding, Y., Niemczyk, M., Kudla, G. and Plotkin, J.B. (2013) Rate-limiting steps in yeast protein translation. Cell 153, 1589–1601 47 Velculescu, V.E., Zhang, L., Zhou, W., Vogelstein, J., Basrai, M.A., Bassett, Jr, D.E., Hieter, P., Vogelstein, B. and Kinzler, K.W. (1997) Characterization of the yeast transcriptome. Cell 88, 243–251
Received 2 July 2013 doi:10.1042/BST20130132
C The
C 2014 Biochemical Society Authors Journal compilation
165
Copyright of Biochemical Society Transactions is the property of Portland Press Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Biochemical Society Transactions (2014) Volume 42, part 1
Controlling translation elongation efficiency: tRNA regulation of ribosome flux on the mRNA Barbara Gorgoni*1 , Elizabeth Marshall*†1 , Matthew R. McFarland*1 , M. Carmen Romano*† and Ian Stansfield*2 *Institute of Medical Sciences, University of Aberdeen, Aberdeen AB25 2ZD, U.K. †Institute of Complex Systems and Mathematical Biology, King’s College, University of Aberdeen, Aberdeen AB24 3UE, U.K.
Biochemical Society Transactions
www.biochemsoctrans.org
Abstract Gene expression can be regulated by a wide variety of mechanisms. One example concerns the growing body of evidence that the protein-production rate can be regulated at the level of translation elongation by controlling ribosome flux across the mRNA. Variations in the abundance of tRNA molecules cause different rates of translation of their counterpart codons. This, in turn, produces a variable landscape of translational rate across each and every mRNA, with the dynamic formation and deformation of ribosomal queues being regulated by both tRNA availability and the rates of translation initiation and termination. In the present article, a range of examples of tRNA control of gene expression are reviewed, and the use of mathematical modelling to develop a predictive understanding of the consequences of that regulation is discussed and explained. These findings encourage a view that predicting the protein-synthesis rate of each mRNA requires a holistic understanding of how each stage of translation, including elongation, contributes to the overall protein-production rate.
Introduction In all organisms, the dynamic and steady-state levels of cellular proteins can be controlled at multiple stages of the gene-expression process, through regulation of transcription, mRNA processing and turnover, mRNA translation, and protein turnover. Of these various stages, the translation of mRNA into protein by the ribosomal machinery is a factorylike assembly-line process that must be tightly regulated. Translation initiation, where ribosomes join the mRNA, must be orchestrated with the capacity of translation elongation to manufacture the polypeptide. For example, a very low rate of translation elongation, relative to the translation initiation rate, will cause ribosome queues to be established on the mRNA. Given sufficient time and an appropriate mRNA stability, this queuing will lead eventually to ribosome occupancy and occlusion of the 5 -end of that mRNA, causing a reduced translation efficiency. Both translation initiation and elongation are therefore subject to regulation; translation initiation can be regulated by a variety of mechanisms, using secondary structures in the mRNA 5 -UTR [1] or using phosphorylation of translation initiation factors to control ribosome joining [2]. The presence of regulatory upstream ORFs in the 5 -UTR can also control ribosomal access to the ORF [3]. Once ribosomes reach the ORF and begin elongation, their flux through the bulk of the ORF is regulated, in part, by the availability of charged tRNA species, which deliver amino acids to the Key words: ribosomal pausing, Saccharomyces cerevisiae, translational regulation, translation elongation, tRNA, tRNA modification. Abbreviations: Elp/elp, Elongator complex protein; mcm5 , methoxycarbonylmethyl group at C-5; TASEP, totally asymmetric simple exclusion process; s2 , thiol at C-2. 1 2
These authors contributed equally to this review. To whom correspondence should be addressed (email i.stansfi[email protected]).
C The
C 2014 Biochemical Society Authors Journal compilation
growing polypeptide chain. Particularly in microbial species, there can be considerable (>15-fold) variation in the abundance of different tRNA isotypes. Selection for translational efficiency has ensured that this tRNA variation is correlated with biased codon usage [4] of the cognate codon(s) in the transcriptome. This balances the codon demand/tRNA supply relationship (Figure 1). Despite such balancing, individual rare codons will nevertheless be translated slowly by correspondingly rare tRNAs; such translational pauses can potentially cause ribosome queues [5–7] (Figure 2). Although evidence exists that rare codons can cause translational pauses, predicting whether any given slow elongation step can inhibit protein synthesis by promoting queue formation back to the mRNA 5 -end is challenging, and depends upon the complex dynamics of queue formation and dissolution, combined with the initiation rate on that particular mRNA. The recent development of mathematical models of translation has, however, begun to make possible the simulation of ribosome flow along an mRNA through a field of codons at variable rates, allowing the global consequences of such regulation to be predicted [34,35,37,42,45]. In the present review, known examples of the tRNA-mediated control of elongation are considered alongside the use of mathematical models to predict the systems biological effects of this regulation.
Regulation of gene expression by tRNA One of the best established examples of tRNA regulation of gene expression is found in the Streptomyces genus. Mutation of the bldA gene in Streptomyces coelicolor produces cells that exhibit normal vegetative growth, but cannot form an aerial mycelium (giving these mutants their ‘bald’ phenotype) or Biochem. Soc. Trans. (2014) 42, 160–165; doi:10.1042/BST20130132
Translation UK 2013
Figure 1 tRNA abundance is matched to translational demand tRNA abundance, or ‘supply’, was estimated by expressing the tRNA gene copy number for a given isotype as a percentage of all S. cerevisiae tRNA gene copies. For each tRNA type, its corresponding cognate codon(s) were identified and the total number of those codons within the transcriptome enumerated and expressed as a percentage of all codons within the mRNA population using SAGE (serial analysis of gene expression) data [47] to gauge the mRNA copy numbers for each gene transcript type (I. Stansfield, unpublished work).
produce antibiotics and other secondary metabolites [8–10]. The bldA gene encodes the only copy of a UUA-cognate leucine tRNA (tRNALeu UUA) [9]. Of the 7800 genes in the S. coelicolor genome, only approximately 150 of them contain UUA codons, and many of them are involved in differentiation or antibiotic production, but do not appear to
have a role in normal vegetative growth [11,12]. This suggests that either the abundance or the translation efficiency of the UUA-decoding tRNA is compromised in the bldA mutant, leading to down-regulation of UUA-containing genes. In the wild-type cell, bldA tRNA is developmentally regulated, accumulating late in the batch growth cycle to coincide with the production of hyphae and antibiotics [13]. The failure to form a mycelium in bldA mutants is attributed to the presence of the cognate UUA codon in the adpA (also known as bldH) gene. This gene encodes a transcription factor that activates many of the genes responsible for differentiation and antibiotic production [14]. Consequently, mutations in the adpA gene also produce a ‘bald’ phenotype, with the failure to produce aerial hyphae and secondary metabolites. If the adpA UUA codon is mutated to another leucine codon, such as UUG, normal aerial hyphal growth is restored in bldA mutants, suggesting that its expression levels are regulated by the efficiency of UUA decoding [12]. Intriguingly, more recent experiments carried out in another Streptomycete (Streptomyces griseus) found that adpA is required for efficient bldA transcription [15]. The AdpA protein binds sites upstream of the bldA promoter and activates transcription of the tRNA. This sets up a positive-feedback loop between the two where low-level expression of tRNALeu UUA leads to AdpA protein production, in turn further activating bldA transcription, eventually producing sufficient levels of the protein to trigger morphological differentiation. Another example of gene regulation by a low-abundance tRNA is found in Saccharomyces cerevisiae, where mutations
Figure 2 Low-abundance tRNAs cause lower decoding rates The 30–40 genes encoding the range of tRNA isotypes are multi-copy in most organisms, but this copy number varies (ranging in yeast from one to 16 copies/genome), causing corresponding variations in tRNA abundance. Infrequently used ‘rare’ codons, represented by the light grey codon box, recognized by low-abundance tRNAs (light grey), will be decoded more slowly by the translation apparatus, potentially resulting in the formation of ribosome queues at such codon positions. Conversely, the abundant tRNAs (dark grey) will decode their cognate codon more quickly.
C The
C 2014 Biochemical Society Authors Journal compilation
161
162
Biochemical Society Transactions (2014) Volume 42, part 1
in the single-copy SUP70 gene [also known as CDC65 (celldivision cycle 65)], encoding a rare tRNAGln CUG , cause an unregulated pseudohyphal growth form, in contrast with the normal budding growth pattern of yeast [16]. The recessive sup70-65 mutations are located in the anticodon stem of the tRNA and cause base mispairing, which presumably leads to the observed instability and altered decoding properties of the mutant tRNA [16,17]. The use of reporter genes containing tandem arrays of the target CAG codon for the SUP70 tRNA showed clearly that the mutant tRNA was defective in its ability to decode CAG codons. This, in turn, suggests that one or more of the mRNAs in yeast whose protein products regulate pseudohyphal differentiation must be enriched in CAG codons, and be particularly sensitive to the abundance of the CAG-decoding SUP70 tRNA.
Regulation of gene expression by tRNA modification In addition to the regulation of gene expression by tRNA abundance, there is also evidence that tRNA modification status can specifically regulate subsets of yeast genes. tRNAs from all organisms are extensively modified and, interestingly, in most tRNAs containing a uridine in the wobble position (U34 ), this nucleoside is always modified (reviewed in [18]). In particular, the pyrimidine-rich tRNALys UUU , tRNAGlu UUC and tRNAGln UUG contain a double U34 modification: mcm5 (methoxycarbonylmethyl group at C-5) and s2 (thiol at C2). These tRNAs, when unmodified, present an unfavourable codon–anticodon pairing, and the modifications improve ribosome binding and the efficiency and accuracy of both cognate and wobble decoding [19]. The crucial cellular function of these tRNA modifications is highlighted by the fact that they are targeted by two different toxins, zymocin [20,21] and PaT (Pichia acaciae killer toxin) [22], which cleave the modified, but not the unmodified, nucleosides and cause cell death. Indeed, it was through a screen of yeast mutants resistant to zymocin that the genes responsible for the addition of the mcm5 group were identified [23]. These genes encode Elongator, a highly conserved multisubunit complex. Elongator is composed of six subunits, Elp1–Elp6 (Elongator complex protein 1–6) and associates with several other protein factors (reviewed in [24]). Importantly, deletion of each of the subunits in yeast (elp deletions) leads to similar phenotypes, indicating that they are all essential for Elongator function [24]. The Elongator complex was first identified in S. cerevisiae as a transcription elongation factor owing to its association with hyperphosphorylated RNA polymerase II and its histone acetyltransferase activity. It was subsequently implicated in a variety of cellular processes, including exocytosis, kinase signalling, transcriptional gene silencing and protein acetylation (for reviews, see [24,25]). This suggested that Elongator had complex multifunctional roles, in addition to its tRNA-modification function. However, a pivotal study in S. cerevisiae showed that several Elongator mutant phenotypes could be rescued by simply overexpressing two of the tRNAs that normally contain the U34 mcm5 and C The
C 2014 Biochemical Society Authors Journal compilation
s2 modifications [26]. Artificially increasing the levels of unmodified tRNALys UUU and tRNAGln UUG in elp-deleted yeast cells could suppress temperature-sensitivity and defects in chromatin remodelling, transcriptional activation and exocytosis [26]. Similarly, a more recent study demonstrated that elp-dependent defects in telomeric gene silencing and the DNA-damage response were also suppressed by overexpression of the same tRNAs [27]. These data strongly suggest that the Elongator’s main biological role is the regulation of translation, through modification of tRNAs. The pleiotropic phenotypes caused by its deletion are most probably due to reduced translation of several target mRNAs. This is corroborated by the observation that ncs2 (needs Cla4 to survive 2)-deletion mutants, which are missing the U34 s2 , but not the mcm5 , modification phenocopy the elp mutants and are rescued by overexpression of the same tRNAs [26]. Furthermore, these modifications appear to control the efficiency rather than the specificity of decoding, as it is the increased abundance of unmodified tRNAs that is able to restore the wild-type phenotype. A first intriguing insight into the mechanism by which tRNA modifications can regulate translation of specific mRNAs was provided by the work in fission yeast of Hermand and colleagues [28]. A screen for translation defects in an elp3-deletion mutant revealed that, although only a modest global effect on the proteome was observed, the translation of proteins belonging to specific functional groups was significantly affected, including proteins involved in cell division. Analysis of codon content of these groups and, in particular, of the cell-cycle kinase Cdr2, revealed an increased usage for the lysine AAA codon over the synonymous AAG codon, making these mRNAs more dependent on modified tRNALys UUU . For one functional group of mRNAs, this was also true for the glutamate GAA, compared with the GAG, codon. Cdcr2 expression was significantly reduced in the elp mutant background and its deletion mutant phenocopied the elp3 deletion. Importantly, replacing the lysine AAA codons with lysine AAG made this mRNA insensitive to the absence of a functional Elongator. These findings were extended to budding yeast and nematodes, where biased codon usage was observed in mRNAs related to telomeric gene silencing and α-tubulin acetylation respectively [29]. These data provide a strong rationale for the role of Elongator in translation regulation. Moreover, it suggests that specific genes and functional categories of genes may be enriched in codons targeted by tRNAs with Elp modifications; the coupling of tRNA modification with enhanced content of particular codons thus offers an evolutionarily conserved mechanism for the co-ordinated expression of specific group of genes.
Modelling ribosomal flux: theory and simulation Understanding how particular tRNAs, or tRNA modifications, can specifically regulate groups of genes can be approached using mathematical models of translation, where genome-wide simulation of translation is now possible.
Translation UK 2013
Some whole-cell-type models focus on the competition for translational resources imposed by the mRNA population, namely ribosomes and charged tRNAs, a strategy that allows global and mRNA-specific testing of the translational synthesis capacity of the cell [30]. Another type of model used with some success to simulate translation is the TASEP (totally asymmetric simple exclusion process). First used in 1968 to describe protein synthesis, it has since become a paradigmatic model in non-equilibrium statistical physics [31–33]. It has been applied to describe a broad range of different systems, such as transport by molecular motors [34,35] and traffic flow [33,36]. In its simplest form, an mRNA can be treated as a one-dimensional homogenous lattice of length N sites (representing codons), where singlesite particles take the role of ribosomes. The TASEP is a stochastic model, which mimics the inherent randomness of a tRNA diffusing through the cytoplasm to bind the mRNA via the ribosome. The particles enter on the left-hand side of the lattice with a given initiation rate α, elongate (hop) from site i to i + 1, if i + 1 is empty, with rate ki , and exit the lattice on the right-hand side with the termination rate β. The status of each site can be indicated with occupation numbers ni , where i = 1 . . . N. If site i is empty, ni = 0, and if it is occupied then ni = 1. The density of particles at this site can then be given as the time-averaged occupancy . Averaging this by the number of sites in the lattice gives the average density of the entire system . For a homogeneous lattice with ki = k for i = 1 . . . N, it can be shown that, depending upon the relative values of the initiation rate α, termination rate β and elongation rate k, there are three characteristic phases in which the lattice can exist, determined by the value of the density ρ and particle flow J: (i) low density, with α as the rate-limiting factor for J; (ii) high density, with β as the ratelimiting factor; and (iii) maximal current, with k restricting the particle current. Both the low- and high-density phases are suboptimal for particle flow (and therefore for protein production on mRNAs). In the maximal current phase, the density of particles is such that the particle flow reaches its maximum value; in this phase, the mRNA would synthesize proteins with the highest efficiency possible (Figure 3). In more recent years, the TASEP has been extended to model translation more realistically. For example, it has been generalized to particles with a realistic extended footprint of the ribosome [37]. Moreover, the mechanochemical cycle of the ribosome has been included in the TASEP approach, where it has been condensed into two main steps: capture of the correct tRNA and translocation to the next codon. This extended model then introduces twostate ribosomes (without and with tRNA), and it takes into consideration the important fact that when a ribosome is waiting for the next codon to become free, it can use the time to capture the tRNA (this fact was neglected in the simple TASEP approach). In addition, there is evidence that the elongation rate associated with each codon correlates with the concentration of tRNAs available to decode it, and the time taken for the correct tRNA to bind has been
Figure 3 The relative rates of initiation and termination define different ribosome density regimes on the mRNA Using the TASEP model of translating ribosomes, the initiation rate α and termination rate β can be controlled independently. Three different domains within the α–β phase plot are defined: low density, characterized by well-spaced ribosomes and low translational efficiency; high density, characterized by densely packed ribosomes and saturated protein production rate; and maximal current, where the optimum ribosomal loading produces the ideal spacing for efficient protein production at maximal rate.
shown to be the rate-limiting step in elongation rather than the translocation step [38,39]. It has been shown that the dynamics of the ribosomes in this biologically realistic regime change considerably with respect to the TASEP, and the twostate ribosome dynamics leads to a substantially enlarged maximal current phase in the α–β parameter space [40]. Hence this model predicts that the optimal region for protein synthesis is accessible at lower values of the initiation rate. The configuration of slow codons is an important factor in protein production. Grouping the slow codons into clusters (a ‘bottleneck’) has a profound effect on particle flow, with closely spaced slow codons reducing protein synthesis by a factor of ∼2–4 [41]. The positioning of this bottleneck determines whether an mRNA undergoes an abrupt increase in particle density by transitioning from low to high density (a Type I mRNA), or no phase transition, where J depends smoothly on the initiation rate α (Type II) [42]. A Type I mRNA has slow codons either individually or clustered somewhere remote from the 5 -end of the mRNA (far enough to allow the initiation-rate-dependent formation of a ribosome queue), leading to an abrupt change in density as ribosomal ‘traffic jams’ form. In contrast, a Type II mRNA either has no slow codons or a ‘ramp’ at the beginning of the ORF, a feature conserved across multiple species, potentially to reduce the queuing of ribosomes at some point downstream [43,44]. mRNAs can thus be classified according to the type of transition their sequence undergoes as initiation increases [42]. Genome-wide application of such a classification has also shown a strong correlation between an mRNA’s transition type and biological function; for example, those synthesising regulatory proteins are overrepresented in C The
C 2014 Biochemical Society Authors Journal compilation
163
164
Biochemical Society Transactions (2014) Volume 42, part 1
the ‘abrupt’ Type I class [45]. Furthermore, simulations of translation across the entire transcriptome of S. cerevisiae using the two-state ribosome model [40] have now been used for the first time to identify the range of in vivo translation initiation rates for all yeast mRNAs, showing a distribution of rates with the majority clustered around one ribosome joining a given mRNA type per 5–10 s [45]. Interestingly, these calculations, based on experimental data and translation simulation, were well matched with later estimates obtained using a whole-cell model of translation [46]. Models of the types described above are now being subjected to experimental validation. For example, a two-state TASEP simulation of translation on a real mRNA containing tandem arrays of CAG codons was able to successfully, and quantitatively, predict the translation-inhibiting effects of a yeast sup70-65 mutation in tRNAGln CUG caused by ribosomal queuing back to the mRNA 5 -end [17]. Mathematical modelling of mRNA translation is thus a rapidly developing field with the potential in the very near future to provide an in silico test bed for the translational regulation of gene expression in a range of species. The use of such simulations will provide access to quantitative predictions that would otherwise be experimentally challenging to obtain.
Future prospects The development of mathematical models of translation provides revealing new insights into how this cellular production line has evolved to operate smoothly without the inefficiency of ribosomal queuing. Moreover, the expectation is that such approaches, in future, will encourage a more integrated view of translation without artificial modularization into the initiation, elongation and termination stages. Progress in this area will also undoubtedly bring important insights into the rational design of gene sequences for biotechnological protein expression processes, side-stepping the slightly heuristic approaches currently used to optimize gene expression.
Funding Work in the laboratories of I.S. and M.C.R. has been supported by the Biotechnology and Biological Sciences Research Council [grant numbers BB/G010722/1 (to I.S. and M.C.R.), BB/I020926/1 (to I.S.) and BB/F00513X/1 (to M.C.R.)].
References 1 Hentze, M.W. and Kuhn, L.C. (1996) Molecular control of vertebrate iron metabolism: mRNA-based regulatory circuits operated by iron, nitric oxide, and oxidative stress. Proc. Natl. Acad. Sci. U.S.A. 93, 8175–8182 2 Proud, C.G. (2006) Regulation of protein synthesis by insulin. Biochem. Soc. Trans. 34, 213–216 3 Hinnebusch, A.G. (2005) Translational regulation of GCN4 and the general amino acid control of yeast. Annu. Rev. Microbiol. 59, 407–450 4 Sharp, P.M. and Li, W.H. (1987) The codon adaptation index: a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res. 15, 1281–1295 C The
C 2014 Biochemical Society Authors Journal compilation
5 Belcourt, M.F. and Farabaugh, P.J. (1990) Ribosomal frameshifting in the yeast retrotransposon ty: tRNAs induce slippage on a 7 nucleotide minimal site. Cell 62, 339–352 6 Ramachandiran, V., Kramer, G., Horowitz, P.M. and Hardesty, B. (2002) Single synonymous codon substitution eliminates pausing during chloramphenicol acetyl transferase synthesis on Escherichia coli ribosomes in vitro. FEBS Lett. 512, 209–212 7 Buchan, J.R. and Stansfield, I. (2007) Halting a cellular production line: responses to ribosomal pausing during translation. Biol. Cell 99, 475–487 8 Merrick, M.J. (1976) A morphological and genetic mapping study of bald colony mutants of Streptomyces coelicolor. J. Gen. Microbiol. 96, 299–315 9 Lawlor, E.J., Baylis, H.A. and Chater, K.F. (1987) Pleiotropic morphological and antibiotic deficiencies result from mutations in a gene encoding a tRNA-like product in Streptomyces coelicolor A3(2). Genes Dev. 1, 1305–1310 10 Chater, K.F. and Chandra, G. (2008) The use of the rare UUA codon to define “expression space” for genes involved in secondary metabolism, development and environmental adaptation in Streptomyces. J. Microbiol. 46, 1–11 11 Bentley, S.D., Chater, K.F., Cerdeno-Tarraga, A., Challis, G. L., Thomson, N.R., James, K.D., Harris, D.E., Quail, M.A., Kieser, H., Harper, D. et al. (2002) Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417, 141–147 12 Takano, E., Tao, M., Long, F., Bibb, M.J., Wang, L., Li, W., Buttner, M.J., Bibb, M.J., Deng, Z.X. and Chater, K.F. (2003) A rare leucine codon in adpA is implicated in the morphological defect of bldA mutants of Streptomyces coelicolor. Mol. Microbiol. 50, 475–486 13 Leskiw, B.K., Mah, R., Lawlor, E.J. and Chater, K.F. (1993) Accumulation of bldA-specified tRNA is temporally regulated in Streptomyces coelicolor A3(2). J. Bacteriol. 175, 1995–2005 14 Ohnishi, Y., Yamazaki, H., Kato, J., Tomono, A. and Horinouchi, S. (2005) AdpA, a central transcriptional regulator in the A-factor regulatory cascade that leads to morphological development and secondary metabolism in Streptomyces griseus. Biosci. Biotechnol. Biochem. 69, 431–439 15 Higo, A., Horinouchi, S. and Ohnishi, Y. (2011) Strict regulation of morphological differentiation and secondary metabolism by a positive feedback loop between two global regulators AdpA and BldA in Streptomyces griseus. Mol. Microbiol. 81, 1607–1622 16 Murray, L.E., Rowley, N., Dawes, I.W., Johnston, G.C. and Singer, R.A. (1998) A yeast glutamine tRNA signals nitrogen status for regulation of dimorphic growth and sporulation. Proc. Natl. Acad. Sci. U.S.A. 95, 8619–8624 17 Kemp, A.J., Betney, R., Ciandrini, L., Schwenger, A.C.M., Romano, M.C. and Stansfield, I. (2013) A yeast tRNA mutant that causes pseudohyphal growth exhibits reduced rates of CAG codon translation. Mol. Microbiol. 87, 284–300 18 Agris, P.F., Malkiewicz, A., Kraszewski, A., Everett, K., Nawrot, B., Sochacka, E., Jankowska, J. and Guenther, R. (1995) Site-selected introduction of modified purine and pyrimidine ribonucleosides into RNA by automated phosphoramidite chemistry. Biochimie 77, 125–134 19 Vendeix, F.A., Murphy, IV, F.V., Cantara, W.A., Leszczynska, G., Gustilo, E.M., Sproat, B., Malkiewicz, A. and Agris, P.F. (2012) Human tRNALys3 UUU is pre-structured by natural modifications for cognate and wobble codon binding through keto–enol tautomerism. J. Mol. Biol. 416, 467–485 20 Lu, J., Huang, B., Esberg, A., Johansson, M.J. and Bystrom, A.S. (2005) The Kluyveromyces lactis γ -toxin targets tRNA anticodons. RNA 11, 1648–1654 21 Jablonowski, D., Zink, S., Mehlgarten, C., Daum, G. and Schaffrath, R. (2006) tRNAGlu wobble uridine methylation by Trm9 identifies elongator’s key role for zymocin-induced cell death in yeast. Mol. Microbiol. 59, 677–688 22 Klassen, R., Paluszynski, J.P., Wemhoff, S., Pfeiffer, A., Fricke, J. and Meinhardt, F. (2008) The primary target of the killer toxin from Pichia acaciae is tRNAGln . Mol. Microbiol. 69, 681–697 23 Huang, B., Johansson, M. J. and Bystrom, A.S. (2005) An early step in wobble uridine tRNA modification requires the elongator complex. RNA 11, 424–436 24 Glatt, S. and Muller, ¨ C.W. (2013) Structural insights into elongator function. Curr. Opin. Struct. Biol. 23, 235–242 25 Svejstrup, J. Q. (2007) Elongator complex: how many roles does it play? Curr. Opin. Cell Biol. 19, 331–336 26 Esberg, A., Huang, B., Johansson, M.J.O. and Bystrom, ¨ A.S. (2006) Elevated levels of two tRNA species bypass the requirement for Elongator complex in transcription and exocytosis. Mol. Cell 24, 139–148
Translation UK 2013
27 Chen, C., Huang, B., Eliasson, M., Ryden, P. and Bystrom, A.S. (2011) Elongator complex influences telomeric gene silencing and DNA damage response by its role in wobble uridine tRNA modification. PLoS Genet. 7, e1002258 28 Bauer, F., Matsuyama, A., Candiracci, J., Dieu, M., Scheliga, J., Wolf, D.A., Yoshida, M. and Hermand, D. (2012) Translational control of cell division by Elongator. Cell Rep. 1, 424–433 29 Bauer, F. and Hermand, D. (2012) A coordinated codondependent regulation of translation by Elongator. Cell Cycle 11, 4524–4529 30 Chu, D. and von der Haar, T. (2012) The architecture of eukaryotic translation. Nucleic Acids Res. 40, 10098–10106 31 MacDonald, C.T., Gibbs, J.H. and Pipkin, A.C. (1968) Kinetics of biopolymerization on nucleic acid templates. Biopolymers 6, 1–5 32 Derrida, B., Domany, E. and Mukamel, D. (1992) An exact solution of a one-dimensional asymmetric exclusion model with open boundaries. J. Stat. Phys. 69, 667–687 33 Proeme, A., Blythe, R.A. and Evans, M.R. (2011) Dynamical transition in the open-boundary totally asymmetric exclusion process. J. Phys. A 44, 035003 34 Chowdhury, D., Schadschneider, A. and Nishinari, K. (2005) Physics of transport and traffic phenomena in biology: from molecular motors and cells to organisms. Phys. Life Rev. 2, 318–352 35 Neri, I., Kern, N. and Parmeggiani, A. (2013) Modeling cytoskeletal traffic: an interplay between passive diffusion and active transport. Phys. Rev. Lett. 110, 098102 36 Popkov, V., Santen, L., Schadschneider, A. and Schutz, G.M. (2001) Empirical evidence for a boundary-induced nonequilibrium phase transition. J. Phys. A 34, L45–L52 37 Shaw, L.B., Zia, R.K. and Lee, K.H. (2003) Totally asymmetric exclusion process with extended objects: a model for protein synthesis. Phys. Rev. E: Stat. Nonlin. Soft Matter Phys. 68, 021910
38 Varenne, S., Buc, J., Lloubes, R. and Lazdunski, C. (1984) Translation is a non-uniform process: effect of tRNA availability on the rate of elongation of nascent polypeptide chains. J. Mol. Biol. 180, 549–576 39 Liljenstrom, H. and von Heijne, G. (1987) Translation rate modification by preferential codon usage: intragenic position effects. J. Theor. Biol. 124, 43–55 40 Ciandrini, Stansfield, I. and Romano. (2010) Role of the particle’s stepping cycle in an asymmetric exclusion process: a model of mRNA translation. Phys. Rev. E: Stat. Nonlin. Soft Matter Phys. 81, 051904 41 Chou, T. and Lakatos, G. (2004) Clustered bottlenecks in mRNA translation and protein synthesis. Phys. Rev. Lett. 93, 198101 42 Romano, M. C., Thiel, M., Stansfield, I. and Grebogi, C. (2009) Queueing phase transition: theory of translation. Phys. Rev. Lett. 102, 198104 43 Tuller, T., Carmi, A., Vestsigian, K., Navon, S., Dorfan, Y., Zaborske, J., Pan, T., Dahan, O., Furman, I. and Pilpel, Y. (2010) An evolutionarily conserved mechanism for controlling the efficiency of protein translation. Cell 141, 344–354 44 Pechmann, S. and Frydman, J. (2013) Evolutionary conservation of codon optimality reveals hidden signatures of cotranslational folding. Nat. Struct. Mol. Biol. 20, 237–243 45 Ciandrini, L., Stansfield, I. and Romano, M. C. (2013) Ribosome traffic on mRNAs maps to gene ontology: genome-wide quantification of translation initiation rates and polysome size regulation. PLoS Comput. Biol. 9, e1002866 46 Shah, P., Ding, Y., Niemczyk, M., Kudla, G. and Plotkin, J.B. (2013) Rate-limiting steps in yeast protein translation. Cell 153, 1589–1601 47 Velculescu, V.E., Zhang, L., Zhou, W., Vogelstein, J., Basrai, M.A., Bassett, Jr, D.E., Hieter, P., Vogelstein, B. and Kinzler, K.W. (1997) Characterization of the yeast transcriptome. Cell 88, 243–251
Received 2 July 2013 doi:10.1042/BST20130132
C The
C 2014 Biochemical Society Authors Journal compilation
165
Copyright of Biochemical Society Transactions is the property of Portland Press Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
