CoChat
Sign Up
Cited by 463RNA and protein synthesis mechanismsRNA Research and SplicingGenomics and Chromatin Dynamics

The DOI matches its Crossref record; 71 of the 72 references it cites were found in a registry or an index.

Rate-Limiting Steps in Yeast Protein Translation

Open access
Shah, Premal, Ding, Yang, Niemczyk, Malwina, Kudla, Grzegorz, Plotkin, Joshua B.·2013·Cell
DOI 10.1016/j.cell.2013.05.049PMC PMC3694300
Open this paper in CoChat — free accountOpen PDF

Abstract

Deep sequencing now provides detailed snapshots of ribosome occupancy on mRNAs. We leverage these data to parameterize a computational model of translation, keeping track of every ribosome, tRNA, and mRNA molecule in a yeast cell. We determine the parameter regimes in which fast initiation or high codon bias in a transgene increases protein yield and infer the initiation rates of endogenous Saccharomyces cerevisiae genes, which vary by several orders of magnitude and correlate with 5' mRNA folding energies. Our model recapitulates the previously reported 5'-to-3' ramp of decreasing ribosome densities, although our analysis shows that this ramp is caused by rapid initiation of short genes rather than slow codons at the start of transcripts. We conclude that protein production in healthy yeast cells is typically limited by the availability of free ribosomes, whereas protein production under periods of stress can sometimes be rescued by reducing initiation or elongation rates.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1016/j.cell.2013.05.049

Cites

71 of 71 references verified0 citation errors · 0 not checkable

Statistics

Figures

Graphical abstract
Graphical abstract CC BY · PubMed Central
Figure 1 A Computational Model of Protein Translation The model tracks the status of all ribosomes, tRNAs, and mRNAs in a cell in continuous time. At any time point, each tRNA and ribosome molecule is either diffusing freely in the cell or is bound to a specific mRNA molecule at a specific codon position. Translation initiation occurs when a free ribosome diffuses to an mRNA and subsequently, with an mRNA-dependent probability, scans to its start codon. The rate of elongation of each subsequent codon depends on the abundance of free cognate tRNAs and their diffusion to the bound ribosome. All rates are based on experimentally determined parameters, including the cell volume, numbers of mRNAs, total abundances of ribosomes and tRNAs, and their diffusion constants. A precise definition of the Markov state space, illustrative pseudocode, and the complete source code for simulation are provided in the Supplemental Information. See also Figure S1, Tables S1 and S2, and Data S1 and S2.
Figure 1 A Computational Model of Protein Translation The model tracks the status of all ribosomes, tRNAs, and mRNAs in a cell in continuous time. At any time point, each tRNA and ribosome molecule is either diffusing freely in the cell or is bound to a specific mRNA molecule at a specific codon position. Translation initiation occurs when a free ribosome diffuses to an mRNA and subsequently, with an mRNA-dependent probability, scans to its start codon. The rate of elongation of each subsequent codon depends on the abundance of free cognate tRNAs and their diffusion to the bound ribosome. All rates are based on experimentally determined parameters, including the cell volume, numbers of mRNAs, total abundances of ribosomes and tRNAs, and their diffusion constants. A precise definition of the Markov state space, illustrative pseudocode, and the complete source code for simulation are provided in the Supplemental Information. See also Figure S1, Tables S1 and S2, and Data S1 and S2.CC BY · PubMed Central
Figure 2 The Effects of Transgene Codon Bias on Protein Production (A and B) We simulated translation in a wild-type yeast cell with the addition of a transgene. Transgene mRNA levels were set at 1%, 10%, or 50% of all cellular mRNA. We measured the number of transproteins produced per transgene mRNA over 500 s in equilibrium (A). As (A) shows, increasing the codon bias of the transgene generally increases the efficiency of its translation. However, when the transgene is expressed a low levels (e.g., transgene mRNAs constituting 1% of transcriptome), then the gain in translation efficiency achieved by optimizing codon bias is moderate (∼50% gain, comparing CAI ≈ 1 to CAI ≈0, squares in [B]). By contrast, when the transgene mRNAs constitute a large fraction of the total transcriptome, then the gain in translation efficiency by optimizing codon bias is far greater (3.6-fold gain, triangles in [B]). See also Tables S3 and S4 and Figure S2.
Figure 2 The Effects of Transgene Codon Bias on Protein Production (A and B) We simulated translation in a wild-type yeast cell with the addition of a transgene. Transgene mRNA levels were set at 1%, 10%, or 50% of all cellular mRNA. We measured the number of transproteins produced per transgene mRNA over 500 s in equilibrium (A). As (A) shows, increasing the codon bias of the transgene generally increases the efficiency of its translation. However, when the transgene is expressed a low levels (e.g., transgene mRNAs constituting 1% of transcriptome), then the gain in translation efficiency achieved by optimizing codon bias is moderate (∼50% gain, comparing CAI ≈ 1 to CAI ≈0, squares in [B]). By contrast, when the transgene mRNAs constitute a large fraction of the total transcriptome, then the gain in translation efficiency by optimizing codon bias is far greater (3.6-fold gain, triangles in [B]). See also Tables S3 and S4 and Figure S2.CC BY · PubMed Central
Figure 3 The Effects of Transgene Codon Bias on the Pool of Free Ribosomes (A and B) As in Figure 2, we simulated translation of a transgene added to a wild-type yeast cell. Transgene mRNA levels were set at 1%, 10%, or 50% of all cellular mRNA. We measured the equilibrium fraction of ribosomes that are free (unbound). Increasing codon bias of the transgene reduces the number of ribosomes bound to its transcripts and thereby increases the pool of free ribosomes (A), especially when the transgene accounts for a large proportion of all cellular mRNA. For example, when transgene mRNAs comprise 50% of the total transcriptome, then optimizing codon bias of the transgene from CAI ≈ 0 to CAI ≈ 1 causes a 3.2-fold increase in the number of free ribosomes (triangles in [B]), which explains a large proportion of the corresponding gain in transprotein production. See also Table S4.
Figure 3 The Effects of Transgene Codon Bias on the Pool of Free Ribosomes (A and B) As in Figure 2, we simulated translation of a transgene added to a wild-type yeast cell. Transgene mRNA levels were set at 1%, 10%, or 50% of all cellular mRNA. We measured the equilibrium fraction of ribosomes that are free (unbound). Increasing codon bias of the transgene reduces the number of ribosomes bound to its transcripts and thereby increases the pool of free ribosomes (A), especially when the transgene accounts for a large proportion of all cellular mRNA. For example, when transgene mRNAs comprise 50% of the total transcriptome, then optimizing codon bias of the transgene from CAI ≈ 0 to CAI ≈ 1 causes a 3.2-fold increase in the number of free ribosomes (triangles in [B]), which explains a large proportion of the corresponding gain in transprotein production. See also Table S4.CC BY · PubMed Central
Figure 4 The Effects of Initiation Probabilities on Protein Production and Pool of Free Ribosomes (A and B) As in Figure 2, we simulated translation of a transgene added to a wild-type yeast cell. Transgene mRNA levels were set at 25% of all cellular mRNA. We measured the number of transgene proteins produced per transgene mRNA (A), as well as the equilibrium fraction of ribosomes that are free (B); both quantities are expressed relative to the case of transgene with CAI ≈ 0. The dashed vertical line denotes the average initiation probability of endogenous yeast genes. Increasing codon bias of a transgene significantly increases the rate of protein production only when the transgene’s initiation probability exceeds the average initiation probability of endogenous genes. See also Table S4 and Figure S2.
Figure 4 The Effects of Initiation Probabilities on Protein Production and Pool of Free Ribosomes (A and B) As in Figure 2, we simulated translation of a transgene added to a wild-type yeast cell. Transgene mRNA levels were set at 25% of all cellular mRNA. We measured the number of transgene proteins produced per transgene mRNA (A), as well as the equilibrium fraction of ribosomes that are free (B); both quantities are expressed relative to the case of transgene with CAI ≈ 0. The dashed vertical line denotes the average initiation probability of endogenous yeast genes. Increasing codon bias of a transgene significantly increases the rate of protein production only when the transgene’s initiation probability exceeds the average initiation probability of endogenous genes. See also Table S4 and Figure S2.CC BY · PubMed Central
Figure 5 Fast Initiation of Short Genes Causes a 5′ Ribosomal Ramp (A) Yeast genes with weak 5′ mRNA structure initiate quickly. We inferred the initiation probabilities of 3,795 endogenous yeast genes from ribosomal profiling data (Ingolia et al., 2009). A gene’s initiation probability correlates strongly with the estimated energy of its 5′ mRNA structure. The gray bars indicate 1 SD of folding energies of binned genes. (B) Initiation probabilities of yeast genes also correlate with ORF lengths, suggesting that short genes have experienced selection for faster initiation. (C) Simulations of translation in a wild-type yeast cell recapitulate the “ramp” of 5′ ribosomes observed in empirical ribosomal profiling data (Ingolia et al., 2009). The figure shows the density of ribosomes bound to mRNAs as a function of codon position, averaged across the simulated transcriptome (black). The ramp is preserved in simulations that permute the codons within each gene (blue), but the ramp is disrupted when permuting the initiation probabilities among genes (red). Thus, we infer that the ramp of 5′ ribosomes is caused primarily by the trend toward faster initiation in short genes, rather than by the ordering of codons within each gene. See also Figures S3, S4, and S5.
Figure 5 Fast Initiation of Short Genes Causes a 5′ Ribosomal Ramp (A) Yeast genes with weak 5′ mRNA structure initiate quickly. We inferred the initiation probabilities of 3,795 endogenous yeast genes from ribosomal profiling data (Ingolia et al., 2009). A gene’s initiation probability correlates strongly with the estimated energy of its 5′ mRNA structure. The gray bars indicate 1 SD of folding energies of binned genes. (B) Initiation probabilities of yeast genes also correlate with ORF lengths, suggesting that short genes have experienced selection for faster initiation. (C) Simulations of translation in a wild-type yeast cell recapitulate the “ramp” of 5′ ribosomes observed in empirical ribosomal profiling data (Ingolia et al., 2009). The figure shows the density of ribosomes bound to mRNAs as a function of codon position, averaged across the simulated transcriptome (black). The ramp is preserved in simulations that permute the codons within each gene (blue), but the ramp is disrupted when permuting the initiation probabilities among genes (red). Thus, we infer that the ramp of 5′ ribosomes is caused primarily by the trend toward faster initiation in short genes, rather than by the ordering of codons within each gene. See also Figures S3, S4, and S5.CC BY · PubMed Central
Figure 6 Codon Usage and Ribosomal Pausing and Stalling (A and B) The influence of codon usage and amino acid sequence on ribosomal pausing and stalling for a simulated transgene with either (A) low or (B) high codon adaptation, expressed at 50% transcript abundance. Gray bars indicate the probability of finding a ribosome bound at a given codon position x, and black bars indicate the probability of finding a ribosome stalled at position x-10 (i.e., a ribosome whose further elongation is obstructed by another ribosome). High codon adaptation reduces both ribosome density and ribosome interference. The probability of a ribosome stalling at a position correlates strongly with the probability of a ribosome pausing 10 positions ahead (R = 0.947 for high CAI and R = 0.644 for low CAI). For a transgene with high CAI, the probability of finding a ribosome bound at a given codon position is strongly anticorrelated with the abundance of iso-accepting tRNAs for that codon (R = 0.786), but not for a transgene with low CAI (R = −0.042).
Figure 6 Codon Usage and Ribosomal Pausing and Stalling (A and B) The influence of codon usage and amino acid sequence on ribosomal pausing and stalling for a simulated transgene with either (A) low or (B) high codon adaptation, expressed at 50% transcript abundance. Gray bars indicate the probability of finding a ribosome bound at a given codon position x, and black bars indicate the probability of finding a ribosome stalled at position x-10 (i.e., a ribosome whose further elongation is obstructed by another ribosome). High codon adaptation reduces both ribosome density and ribosome interference. The probability of a ribosome stalling at a position correlates strongly with the probability of a ribosome pausing 10 positions ahead (R = 0.947 for high CAI and R = 0.644 for low CAI). For a transgene with high CAI, the probability of finding a ribosome bound at a given codon position is strongly anticorrelated with the abundance of iso-accepting tRNAs for that codon (R = 0.786), but not for a transgene with low CAI (R = −0.042).CC BY · PubMed Central
Figure 7 The Effects of Amino Acid Starvation and Cellular Stress Response on Protein Production (A) For each amino acid, we reduced the relative abundance of its (charged) cognate tRNAs by either 2-, 5-, or 10-fold and then simulated translation in the cell. Although a stronger stress always leads to lower protein production, the strength of this effect varies widely among amino acids. (B) During starvation, cells often respond by repressing ribosome production, which we modeled by reducing the total number of ribosomes in the cell and resimulating translation. Reducing ribosomes under normal or mild stress conditions always reduces protein production. However, during severe stress, reducing ribosomes by a moderate amount can partly rescue protein production, as in the case of serine shown in (B). See also Figures S6 and S7 and Table S5.
Figure 7 The Effects of Amino Acid Starvation and Cellular Stress Response on Protein Production (A) For each amino acid, we reduced the relative abundance of its (charged) cognate tRNAs by either 2-, 5-, or 10-fold and then simulated translation in the cell. Although a stronger stress always leads to lower protein production, the strength of this effect varies widely among amino acids. (B) During starvation, cells often respond by repressing ribosome production, which we modeled by reducing the total number of ribosomes in the cell and resimulating translation. Reducing ribosomes under normal or mild stress conditions always reduces protein production. However, during severe stress, reducing ribosomes by a moderate amount can partly rescue protein production, as in the case of serine shown in (B). See also Figures S6 and S7 and Table S5.CC BY · PubMed Central
Figure S1 Validation of Analytical Approximations of Simulations Model, Related to Figure 1 (A) We have derived analytic approximations for the steady-state density of ribosomes on each mRNA molecule. These equations provide a good approximation to the true behavior of our stochastic simulation. (B) We can invert these equations to infer the initiation probability of each gene, based on its observed equilibrium ribosome density. Panel B shows a validation experiment in which we applied this inference method to data that had been simulated with known initiation probabilities, confirming that we can accurately infer these parameters. (C) Correlation between gene length and initiation probabilities does not bias our inference of initiation probabilities. We validated that we can reliably infer initiation probabilities from observed ribosome densities, even when gene length and initiation probabilities are positively correlated. To do so we inverted the rank order of initiation probabilities assigned to yeast genes, so that the genes with longer ORFs were now assigned greater initiation probabilities. (D) When we simulated a cell with this new assignment of initiation probabilities we were still able to reliably infer the simulated initiation probabilities based on the simulated ribosome density for each gene. These results confirm that the negative correlation between gene length and inferred initiation probability observed in the real data (main text Figure 5B) is not an artifact of our inference procedure.
Figure S1 Validation of Analytical Approximations of Simulations Model, Related to Figure 1 (A) We have derived analytic approximations for the steady-state density of ribosomes on each mRNA molecule. These equations provide a good approximation to the true behavior of our stochastic simulation. (B) We can invert these equations to infer the initiation probability of each gene, based on its observed equilibrium ribosome density. Panel B shows a validation experiment in which we applied this inference method to data that had been simulated with known initiation probabilities, confirming that we can accurately infer these parameters. (C) Correlation between gene length and initiation probabilities does not bias our inference of initiation probabilities. We validated that we can reliably infer initiation probabilities from observed ribosome densities, even when gene length and initiation probabilities are positively correlated. To do so we inverted the rank order of initiation probabilities assigned to yeast genes, so that the genes with longer ORFs were now assigned greater initiation probabilities. (D) When we simulated a cell with this new assignment of initiation probabilities we were still able to reliably infer the simulated initiation probabilities based on the simulated ribosome density for each gene. These results confirm that the negative correlation between gene length and inferred initiation probability observed in the real data (main text Figure 5B) is not an artifact of our inference procedure.CC BY · PubMed Central
Figure S2 Experimental Validation of Roles of Initiation and Elongation Rates on Protein Production in Yeast, Related to Figures 2 and 4 (A and B) The relationship between protein production, codon bias, and 5′ mRNA folding for 141 synthetic GFP genes expressed in Saccharomyces cerevisiae. We constructed 141 GFP genes, all encoding the same protein sequence but varying randomly at synonymous sites, as previously described (Kudla et al., 2009). BY4741 yeast was transformed with a Gal-induced 2-micron plasmid, then grown for 24h in 2% Raf +2% Gal. Fluorescence was then measures by FACS. Measurements typically represent three biological replicates, on the same day. Panel (A) shows the relationship between fluorescence and each GFP gene’s yeast Codon Adaptation Index; Panel (B) shows the relationship between fluorescence and each GFP gene’s predicted 5′ mRNA folding. 5′ mRNA folding was estimated using mfold, as previously described (Kudla et al., 2009). There is no significant correlation between expression level and codon bias, suggesting that elongation rates do not modulate protein production in these constructs. By contrast, there is a significant correlation between expression level and mRNA folding near the translation initiation site, indicating that initiation rates strongly influence protein production in these constructs.
Figure S2 Experimental Validation of Roles of Initiation and Elongation Rates on Protein Production in Yeast, Related to Figures 2 and 4 (A and B) The relationship between protein production, codon bias, and 5′ mRNA folding for 141 synthetic GFP genes expressed in Saccharomyces cerevisiae. We constructed 141 GFP genes, all encoding the same protein sequence but varying randomly at synonymous sites, as previously described (Kudla et al., 2009). BY4741 yeast was transformed with a Gal-induced 2-micron plasmid, then grown for 24h in 2% Raf +2% Gal. Fluorescence was then measures by FACS. Measurements typically represent three biological replicates, on the same day. Panel (A) shows the relationship between fluorescence and each GFP gene’s yeast Codon Adaptation Index; Panel (B) shows the relationship between fluorescence and each GFP gene’s predicted 5′ mRNA folding. 5′ mRNA folding was estimated using mfold, as previously described (Kudla et al., 2009). There is no significant correlation between expression level and codon bias, suggesting that elongation rates do not modulate protein production in these constructs. By contrast, there is a significant correlation between expression level and mRNA folding near the translation initiation site, indicating that initiation rates strongly influence protein production in these constructs.CC BY · PubMed Central
Figure S3 Evidence that the Ramp of 5′ Ribosomes in Profiling Data Is Not Caused by Codon Ordering, Related to Figure 5 (A) We removed all positional information in the profiling data, and used instead only the average ribosome density along each transcript, as reported by Ingolia et al. (2009). When we assume that ribosomes are uniformly distributed along each message, then the figure shows the resulting density of ribosomes as a function of codon position, averaged across the expression-weighted transcriptome. Similar results were obtained using the unweighted transcriptome (data not shown). In both cases we observe a ramp of higher 5′ ribosome density. These results demonstrate that the ramp can be attributed primarily to a greater overall density of ribosomes on shorter genes, even if ribosomes are distributed uniformly along each message. (B) Evidence that individual genes in yeast do not exhibit a trend toward a 5′ ramp of ribosomes. We analyzed the raw profiling data of Ingolia et al. (2009) and computed, for each yeast gene, the regression between codon position and average ribosome density at that position. A gene with a ramp of elevated 5′ ribosome densities would exhibit a negative slope. Instead, we found that just as many genes exhibit a positive slope as exhibit a negative slope. The mean slope among genes is not significantly different from zero (p = 0.58). Thus, there is no systematic trend toward elevated 5′ ribosome densities on individual genes, in the data of Ingolia et al. (2009).
Figure S3 Evidence that the Ramp of 5′ Ribosomes in Profiling Data Is Not Caused by Codon Ordering, Related to Figure 5 (A) We removed all positional information in the profiling data, and used instead only the average ribosome density along each transcript, as reported by Ingolia et al. (2009). When we assume that ribosomes are uniformly distributed along each message, then the figure shows the resulting density of ribosomes as a function of codon position, averaged across the expression-weighted transcriptome. Similar results were obtained using the unweighted transcriptome (data not shown). In both cases we observe a ramp of higher 5′ ribosome density. These results demonstrate that the ramp can be attributed primarily to a greater overall density of ribosomes on shorter genes, even if ribosomes are distributed uniformly along each message. (B) Evidence that individual genes in yeast do not exhibit a trend toward a 5′ ramp of ribosomes. We analyzed the raw profiling data of Ingolia et al. (2009) and computed, for each yeast gene, the regression between codon position and average ribosome density at that position. A gene with a ramp of elevated 5′ ribosome densities would exhibit a negative slope. Instead, we found that just as many genes exhibit a positive slope as exhibit a negative slope. The mean slope among genes is not significantly different from zero (p = 0.58). Thus, there is no systematic trend toward elevated 5′ ribosome densities on individual genes, in the data of Ingolia et al. (2009).CC BY · PubMed Central
Figure S4 Evidence that Genes Binned by Lengths Do Not Exhibit a Trend toward a 5′ Ramp of Ribosomes, Related to Figure 5 We analyzed the raw profiling data of Ingolia et al. (2009), and computed position-specific ribosome density for groups of genes binned by their ORF lengths. This figures is analogous to Figure S11 of Ingolia et al. (2009), but with more stringent bins on length: each panel contains 5% of genes in the Ingolia data set, sorted according to their lengths. We find no consistent 5′-to-3′ ramp (some panels show 3′-to 5′-ramps) in the data of Ingolia et al. (2009).
Figure S4 Evidence that Genes Binned by Lengths Do Not Exhibit a Trend toward a 5′ Ramp of Ribosomes, Related to Figure 5 We analyzed the raw profiling data of Ingolia et al. (2009), and computed position-specific ribosome density for groups of genes binned by their ORF lengths. This figures is analogous to Figure S11 of Ingolia et al. (2009), but with more stringent bins on length: each panel contains 5% of genes in the Ingolia data set, sorted according to their lengths. We find no consistent 5′-to-3′ ramp (some panels show 3′-to 5′-ramps) in the data of Ingolia et al. (2009).CC BY · PubMed Central
Figure S5 Comparison between Initiation-Limited and Elongation-Limited Regimes of Protein Translation, Related to Figure 5 We simulated our model of translation in a wild-type yeast cell with either the standard number of ribosomes and tRNAs, as used in all other simulations in our study, or alternatively with 10-fold as many tRNAs and an elevated number ribosomes. (A) We ran multiple simulations that varied the scaling coefficient applied to the number of ribosomes, from 1-fold to 35-fold more than their empirically measured abundance. The number of ribosomes (or, more precisely, the number of free ribosomes) determines the initiation rate of any mRNA whose 5′ end is not currently obstructed by a bound ribosome; and so this quantity corresponds to the initiation rate λ in the ribosome-flow model (Reuveni et al., 2011). Black lines show the total rate of protein production, as a function of the fold-increase in ribosome abundance. As the number of ribosomes is increased, our simulation model behaves like the ribosome-flow model, in which free ribosomes are always abundant and each mRNA is initiated at a fixed rate independent of how many ribosomes are being used to translate other mRNAs; in this regime protein production is elongation-limited. However, the number of ribosomes required to achieve this regime is unrealistic: many more ribosomes are bound to mRNAs (red curve) in this regime than the total number of ribosomes in a real yeast cell. (B) Simulations under elongation-limited regime of ribosome flow models do produce a “ramp” of decreasing ribosome densities with codon position. However, this decrease persists regardless of codon ordering within genes and to a lesser degree with randomized gene-specific initiation probabilities. This indicates that the ramp produced under the assumptions of ribosome-flow model is an artifact of the underlying elongation-limiting assumption.
Figure S5 Comparison between Initiation-Limited and Elongation-Limited Regimes of Protein Translation, Related to Figure 5 We simulated our model of translation in a wild-type yeast cell with either the standard number of ribosomes and tRNAs, as used in all other simulations in our study, or alternatively with 10-fold as many tRNAs and an elevated number ribosomes. (A) We ran multiple simulations that varied the scaling coefficient applied to the number of ribosomes, from 1-fold to 35-fold more than their empirically measured abundance. The number of ribosomes (or, more precisely, the number of free ribosomes) determines the initiation rate of any mRNA whose 5′ end is not currently obstructed by a bound ribosome; and so this quantity corresponds to the initiation rate λ in the ribosome-flow model (Reuveni et al., 2011). Black lines show the total rate of protein production, as a function of the fold-increase in ribosome abundance. As the number of ribosomes is increased, our simulation model behaves like the ribosome-flow model, in which free ribosomes are always abundant and each mRNA is initiated at a fixed rate independent of how many ribosomes are being used to translate other mRNAs; in this regime protein production is elongation-limited. However, the number of ribosomes required to achieve this regime is unrealistic: many more ribosomes are bound to mRNAs (red curve) in this regime than the total number of ribosomes in a real yeast cell. (B) Simulations under elongation-limited regime of ribosome flow models do produce a “ramp” of decreasing ribosome densities with codon position. However, this decrease persists regardless of codon ordering within genes and to a lesser degree with randomized gene-specific initiation probabilities. This indicates that the ramp produced under the assumptions of ribosome-flow model is an artifact of the underlying elongation-limiting assumption.CC BY · PubMed Central
Figure S6 Effects of Amino Acid Starvation on Protein Production, Number of Free Ribosomes, Overall Elongation Rates, and Ribosomal Stalling, Related to Figure 7 All quantities are shown relative to a wild-type cell without any amino acid starvation. We simulated starvation of a particular amino acid by reducing the abundance of all its (charged) cognate tRNAs by either two-, five-, or ten-fold. The rate of protein production decreases under stress, and it can decrease extensively when starvation is severe. Starvation of different amino acids has radically different effects on protein production. (A–D) Decreased protein synthesis upon starvation is caused primarily by a corresponding decrease in the pool of free ribosomes (comparing panel A to panel B). Starving a cell of amino acids reduces the number of free ribosomes, and this in turn leads to fewer free tRNAs. The combined effects of these two processes lead to an overall decrease in (C) total elongation rates across all mRNAs. Starving a cell of amino acids leads to higher densities of ribosomes bound to mRNAs (D). As a result, greater proportions of bound ribosomes are stalled - that is, they are obstructed from further elongation due to a neighboring ribosome bound 10 codons ahead on the same mRNA.
Figure S6 Effects of Amino Acid Starvation on Protein Production, Number of Free Ribosomes, Overall Elongation Rates, and Ribosomal Stalling, Related to Figure 7 All quantities are shown relative to a wild-type cell without any amino acid starvation. We simulated starvation of a particular amino acid by reducing the abundance of all its (charged) cognate tRNAs by either two-, five-, or ten-fold. The rate of protein production decreases under stress, and it can decrease extensively when starvation is severe. Starvation of different amino acids has radically different effects on protein production. (A–D) Decreased protein synthesis upon starvation is caused primarily by a corresponding decrease in the pool of free ribosomes (comparing panel A to panel B). Starving a cell of amino acids reduces the number of free ribosomes, and this in turn leads to fewer free tRNAs. The combined effects of these two processes lead to an overall decrease in (C) total elongation rates across all mRNAs. Starving a cell of amino acids leads to higher densities of ribosomes bound to mRNAs (D). As a result, greater proportions of bound ribosomes are stalled - that is, they are obstructed from further elongation due to a neighboring ribosome bound 10 codons ahead on the same mRNA.CC BY · PubMed Central
Figure S7 The Effects of Repressing Ribosome Synthesis on Protein Production during Amino Acid Starvation, Related to Figure 7 During starvation, cells usually respond by repressing ribosome production. Reducing ribosomes under normal or mild-stress conditions always reduces protein production. However, for many amino acids (e.g., A, E, L, P, Q, S, V), reducing ribosomes by small amounts partly rescues protein production under amino-acid starvation. The y axis shows the total rate of protein production relative to a wild-type cell without any amino-acid starvation.
Figure S7 The Effects of Repressing Ribosome Synthesis on Protein Production during Amino Acid Starvation, Related to Figure 7 During starvation, cells usually respond by repressing ribosome production. Reducing ribosomes under normal or mild-stress conditions always reduces protein production. However, for many amino acids (e.g., A, E, L, P, Q, S, V), reducing ribosomes by small amounts partly rescues protein production under amino-acid starvation. The y axis shows the total rate of protein production relative to a wild-type cell without any amino-acid starvation.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 72 references

Research works printed but not linked: 1

  1. Tuller, 2010, An evolutionarily conserved mechanism for controlling the efficiency of protein translation, Cell, 141, 344, 10.1016/j.cell.2010.03.031 · DOI · Paper page

Cited by

Sorted by citations

See all 463

Related

Metadata

Authors
Shah, P., Ding, Y., Niemczyk, M., Kudla, G., & Plotkin, J. B.
Affiliations
  • University of Pennsylvania (US)
  • Wellcome Centre for Cell Biology (GB)
Corresponding author
Plotkin, Joshua B.
ORCID
Venue
Cell
Publisher
Elsevier BV
Published
2013-06
Volume
153
Issue
7
Pages
1589-1601
ISSN
0092-8674 (ISSN-L)
DOI
10.1016/j.cell.2013.05.049
PMC
PMC3694300
OpenAlex
W2011503994
Cited by
463
References
72
Citations by year
2026: 15 · 2025: 24 · 2024: 31 · 2023: 33 · 2022: 37 · 2021: 27 · 2020: 40 · 2019: 37 · 2018: 42 · 2017: 50 · 2016: 48 · 2015: 26 · 2014: 44 · 2013: 9
FWCI
18.98
Type
journal article
Language
en
Open access
Licence
Locations
  • Cell (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by)
Funders
  • Burroughs Wellcome Fund
  • Army Research Office (D12AP00025, W911NF-12-1-0552)
  • James S. McDonnell Foundation
  • David and Lucile Packard Foundation
  • U.S. Department of the Interior
  • Wellcome Trust (077248)
  • Alfred P. Sloan Foundation (RFP-12-16)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › RNA and protein synthesis mechanisms
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › RNA Research and Splicing
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
Keywords
mRNA translation, ribosome occupancy, translation initiation rate, codon usage bias, free ribosomes, protein production, Saccharomyces cerevisiae, tRNA availability, stress response
MeSH terms
Codon, Markov Chains, Models, Genetic, Ribosomes, RNA, Messenger, RNA, Transfer, Saccharomyces cerevisiae, Protein Biosynthesis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review