Absolute Quantification of Protein and mRNA Abundances Demonstrate Variability in Gene-Specific Translation Efficiency in Yeast

The quote only partly states this claim.

mRNA abundance was the dominant factor controlling protein synthesis, explaining 61% of the variance in final protein abundance.

“Non-linear regression analysis detected that mRNA abundance and translation elongation were the dominant factors controlling protein synthesis, explaining 61% and 15% of its variance.”

• Also stated in: § Translation Efficiency and Its Regulation · p. 7 · plain

  “We found that 61% of the final protein abundances can be predicted based on their corresponding mRNA abundance”

The quote only partly states this claim.

Across all conditions, the overall correlation between mRNA and protein abundances was low, with a Pearson correlation of about 0.46 (R^2 = 0.45).

“The overall correlation between mRNA and protein abundances across all conditions was low (0.46)”

• Also stated in: § Quantitative Transcriptome and Proteome under Carbon-Limited Conditions · p. 5 · plain

  “a Pearson correlation of R 2 = 0.51 was detected for the reference condition and R 2 = 0.45 when leveraging on all the studied conditions”

May combine two claims.

The authors measured absolute levels for 5,354 transcripts and 2,198 proteins in S. cerevisiae across ten environmental conditions, plus turnover for 1,384 proteins under a reference condition.

“We quantified absolute abundances of 5,354 mRNAs and 2,198 proteins in Saccharomyces cerevisiae under ten environmental conditions and protein turnover for 1,384 proteins under a reference condition.”

May combine two claims.

Of those pairs, 164 changed at both levels, accounting for 81% of differentially expressed proteins and 15% of differentially expressed transcripts.

“Amongthese pairs, 164 showed differential expression at both the transcript and protein level, which is 81% of all differentially expressed proteins and 15% of the differentially expressed transcripts.”

May combine two claims.

Combining transcript and protein measurements produced 1,864 matched pairs; 1,205 transcripts and 202 proteins were differentially expressed.

“Our integrated transcript-protein data resulted in 1,864 comparable transcript-protein pairs where differential expression wasobservedin1,205transcripts and202proteins.”

May combine two claims.

Linear correlation predicted only 46% of protein levels correctly after accounting for noise; roughly 10% of detected proteins deviated from prediction by over 10-fold.

“However, only 46% of protein levels were predicted correctly based on linear correlation while experimental noise was taken into account for the calculations (STAR Methods), where approximately 10% of the detected proteins showed more than a 10-fold difference from the predicted value.”

May combine two claims.

In the reference condition, the median protein level was 491 molecules per pg dry weight, or 6,384 molecules per cell.

“Under the reference condition, we found for each gene a median of 0.7 mRNA molecules and 491 protein molecules per picogram of DW corresponding to 9 and 6,384 molecules per cell under the reference condition, respectively (Figure 1B).”

May combine two claims.

In the reference condition, the median transcript level was 0.7 molecules per pg dry weight, or 9 molecules per cell.

“Under the reference condition, we found for each gene a median of 0.7 mRNA molecules and 491 protein molecules per picogram of DW corresponding to 9 and 6,384 molecules per cell under the reference condition, respectively (Figure 1B).”

Probably not a substantial claim: may not pass S3 (load-bearing).

The estimated confidence intervals deviated from the median by 36% on average.

“On average, we observed a 36% deviation from the median on each estimated confidence interval (Figure S3 and Table S5).”

May combine two claims.

They estimated one cell to weigh 13 ± 1 pg, which corresponds to a volume of 35 fL.

“We then estimated a single cell weight equal to 13 ± 1 pg, similar to values reported in the literature and corresponding to a cell volume of 35 fL (Bryan et al., 2010; Klis et al., 2014; McMurrough and Rose, 1967).”

• cites Bryan et al., 2010 sourceNot checked yet

  Not checked in the free run.

• cites Klis et al., 2014 sourceNot checked yet

  Not checked in the free run.

• cites McMurrough and Rose, 1967Not checked yet

  Not checked in the free run.

May combine two claims.

From these data, protein turnover was estimated for 1,384 proteins and translation efficiency was calculated for 1,117 proteins.

“Using these data, we estimated the protein turnover rate (for 1,384 proteins) and calculated the translation efficiency for 1,117 proteins.”

The quote only partly states this claim.

Proteins with high translation efficiency had elevated adenine content, particularly at positions -1 and -3.

“we discovered that highly translated proteins showed: (1) high adenine content, especially at positions 1 and 3;”

May combine two claims.

Those genes were strongly enriched for oxidative phosphorylation (p<e-14) and aromatic amino acid biosynthesis (p<e-5).

“Enrichmentanalysisofthosegenesshowedstrong enrichment toward oxidative phosphorylation genes (p &lt; e-14) and aromatic amino acid biosynthetic processes (p &lt; e-5).”

• AssumptionThe authors state this as an assumption: The flux simulation assumed that yeast metabolism maximizes ATP consumption, i.e., that maximum ATP yield is the objective function.

May combine two claims.

The sole exception was growth at 38°C, where 334 proteins, about 17%, differed significantly, with equal numbers increasing and decreasing relative to the reference.

“The only exception was high-temperature stress (38 C), where 334 proteins (17%) showed significant differences, equally split in increased versus decreased translation efficiency compared with reference (Figure S5).”

May combine two claims.

Elongation effects were driven mostly by codon and tRNA adaptation indices, with TT and GT sequence frequency also contributing.

“Translation elongation was contributing mainly through codon and tRNA adaptation indexes, but also due to the high frequency of TT and GT sequences (for more information see Table S9).”

May combine two claims.

Changes in protein levels correlated well with transcript changes, but translation efficiency varied widely among individual proteins.

“We found that although changes in protein levels are well correlated with the changes in the transcripts, the translation efficiency of proteins is highly variable between individual proteins.”

May combine two claims.

Although the medians differed by 33%, the distributions of individual protein turnover rates overlapped almost completely (Figure S6).

“Despite a 33% difference in the median value, histograms illustrating the distribution of individual protein turnover rates showed an almost complete overlap (Figure S6).”

May combine two claims.

This yeast dataset is larger, quantifying constant protein-mRNA ratios for more than 1,800 proteins across ten environmental conditions.

“Our data, however, represent a larger global study for yeast, whereby constant protein-mRNA ratios were quantified for more than 1,800 proteins across the ten environmental conditions studied.”

The quote only partly states this claim.

Protein turnover rates for 1,384 proteins in the reference condition followed a log-normal distribution with a median of 0.043 hr−1.

“The rates have a log-normal distribution with a medianvalueof 0.043 hr 1 .”

May combine two claims.

Translation elongation features accounted for 15% of the control, and translation initiation features for only 4%.

“In addition, 15% of the control accounted for the features related to translation elongation and only 4% for features related to translation initiation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The MARS algorithm selected 27 features that contributed more than 0.1% to the final protein abundance.

“Twenty-seven features were selected by the algorithm to contribute to the final protein abundance (impacting the final amount more than 0.1%).”

• AssumptionRests on an assumption: The linear relationship between QuantiGene absolute mRNA measurements and RNA-seq FPKM values, established for 18 mRNAs, applies to all mRNAs quantified.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Metabolic fluxes largely shape cellular phenotypes, and how cells regulate metabolism is a major biological question.

“Cellular phenotypes are determined largely by metabolic fluxes, andunderstanding how the cell is controlling metabolism is a central question in biology (Nielsen, 2003, 2011).”

• cites Nielsen, 2003, 2011Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Classified as probably not a substantial claim.

Protein quantification for 2,198 proteins used an internal standard of biomass labeled with heavy lysine and Orbitrap mass spectrometry.

“Heavy ( 15 N, 13 C) lysinelabeled biomass as an internal standard and Orbitrap mass spectrometry (MS) were used for quantification of 2,198 proteins (Table S3).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Prior work has linked energy limitation and proteome allocation to control of key cellular processes, including overflow metabolism, in bacteria and yeast.

“Recently, energetic limitations and proteome allocation has been determined as one of the main contributors controlling important cellular functions, such as overflow metabolism, in bacteria and yeast (Hui et al., 2015; Nilsson and Nielsen, 2016; Peebo et al., 2015; Schmidt et al., 2016).”

• cites Hui et al., 2015 sourceNot checked yet

  Not checked in the free run.

• cites Nilsson and Nielsen, 2016 sourceNot checked yet

  Not checked in the free run.

• cites Peebo et al., 2015 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

Because RNA sequencing has high coverage and accuracy, ribosome profiling is regarded as the preferred approach for measuring translation efficiency.

“Due to the high coverage and accuracy in RNA sequencing, ribosome profiling is considered the best method for translation efficiency estimations.”

May combine two claims.

For mRNA, residual SD averaged 5.2%, and bootstrap analysis gave a mean abundance error of 1.16-fold (95% CI).

“For the absolute mRNA levels, the average SD of the residuals was calculated as 5.2%, and a bootstrapping method estimated an average abundance error of 1.16-fold with a 95% confidence interval.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Flux balance analysis allows metabolic fluxes to be compared with transcriptome and proteome measurements and yields genome-scale information on flux control.

“This approach, called flux balance analysis (Orth et al., 2010), enables comparison of metabolic fluxes with transcriptome and proteomedataandprovidesinformationaboutthe control of fluxes at the genome scale (Bordel et al., 2010).”

• cites Orth et al., 2010 sourceNot checked yet

  Not checked in the free run.

• cites Bordel et al., 2010 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Turnover is a substantial component of cellular upkeep and influences energy metabolism.

“From these analyses, however, it is clear that protein turnover constitutes a significant part of cellular maintenance and affects cellular energy metabolism (Hong et al., 2012).”

• cites Hong et al., 2012 sourceNot checked yet

  Not checked in the free run.

Classified as probably not a substantial claim.

Relative changes between conditions were called significant using p values corrected for false discovery rate.

“The significance of relative changes between different studied conditions was estimated using the false discovery rate (FDR) corrected p values (STAR Methods).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Median turnover rates reported earlier for yeast span roughly 0.078 to 0.98 per hour.

“Previously reported protein turnover measurements have yielded varying results, with reported median protein turnover rates in the range of 0.078-0.98 hr 1 (Belle et al., 2006; Helbig et al., 2011; Hong et al., 2012; Khmelinskii et al., 2012; Pratt et al., 2002).”

• cites Belle et al., 2006 sourceNot checked yet

  Not checked in the free run.

• cites Helbig et al., 2011 sourceNot checked yet

  Not checked in the free run.

• cites Hong et al., 2012 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

For proteins, residual SD averaged 87.4%, and bootstrap analysis gave a mean abundance error of 2.0-fold (95% CI).

“For the proteome data, the average SD of the residuals was 87.4%, and bootstrapping based on average abundance error with a 95% confidence interval was 2.0-fold.”

The quote only partly states this claim.

The mRNA-protein correlations for those 202 proteins held when the confidence threshold was varied.

“These correlations were robust to changes in the selected confidence level.”

Classified as probably not a substantial claim.

The authors reasoned that variation in turnover and translation efficiency should account for post-transcriptional differences.

“As differences in protein turnover and translation efficiency are expected to explain the observed differences in post-transcriptional control, we quantified individual protein turnover under the reference condition”

Probably not a substantial claim: may not pass S3 (load-bearing).

For each turnover fit, a 95% confidence interval was computed for every parameter estimate to gauge accuracy.

“The accuracy of each fit was determined by calculating the 95% confidence interval for each estimate of the parameters.”

May combine two claims.

Each peptide bond formed costs slightly over four ATP, while its degradation is estimated to consume one ATP.

“Synthesisofeachpeptidebondrequires a little more than four ATP molecules (Stouthamer, 1973; Stephanopoulos et al., 1998), and degradation of a peptide bond has been estimated to utilize one ATP (Benaroudj et al., 2003).”

• cites Stouthamer, 1973 sourceNot checked yet

  Not checked in the free run.

• cites Stephanopoulos et al., 1998Not checked yet

  Not checked in the free run.

• cites Benaroudj et al., 2003 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Thus, protein turnover consumes roughly 21% of the energy used for biomass synthesis, making it a major energy consumer in yeast.

“Therefore, 21% of the energy generated for biomass synthesis is used for protein turnover, which makes protein turnover a major client of energy metabolism in yeast.”

May combine two claims.

Using the protein-to-mRNA ratio as a proxy for translation efficiency ignores protein degradation, yet in the reference condition the resulting values differed by no more than about 30% on average.

“this approximation neglects protein turnover; however, differences in translation efficiency values were on average within 30% in the reference condition; Figures S4B and S4C; Table S4”

Probably not a substantial claim: may not pass S3 (load-bearing).

A possible reason for transcriptional control of oxidative phosphorylation is that it coordinates flux through all steps of the pathway, preventing reactive oxygen species formation and cytochrome c leakage that would trigger apoptosis.

“A possible explanation for transcriptional control of oxidative phosphorylation may be that it is essential to coordinate flux control through all the steps of oxidative phosphorylation in order to avoid the formation of reactive oxidative species and cytochrome c leakage, resulting in apoptosis.”

• AssumptionThe authors state this as an assumption: The flux simulation assumed that yeast metabolism maximizes ATP consumption, i.e., that maximum ATP yield is the objective function.

Classified as probably not a substantial claim.

Earlier modeling with flux balance analysis put net cellular ATP production at about 59 mmol per gDW per hour.

“Previously, using genome-scale models and flux balance analysis it had been estimated that the net production of ATP in the cell forms approximately 59 mmol ATP gDW 1 h 1 (Forster et al., 2003).”

• cites Forster et al., 2003 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Another study reported constant tissue-independent protein-mRNA ratios for 55 proteins measured from the same sample (Edfors et al., 2016).

“Recently, another study reported constant tissue-independent protein-mRNA ratios for 55 proteins, for the first time measured from the same sample (Edfors et al., 2016).”

• cites Edfors et al., 2016 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Under carbon limitation, whole-proteome turnover measured with uniformly labeled glucose varied by amino acid and had a median of 0.078 hr−1 (Hong et al., 2012).

“Under carbon limitation, protein turnover has been measured for the whole proteome using uniformly labeled glucose, whereby measurements showed different values for different amino acids with a median of 0.078 hr 1 (Hong et al., 2012).”

• cites Hong et al., 2012 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Protein turnover was measured only in the reference condition, so further work is needed to determine its growth-rate and environmental dependence.

“In the current study, protein turnover was measured only in the reference condition, and further analysis is required to determine the growth rate and dependence on environmental conditions of protein turnover.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Ribosome profiling has reported similar variation (Ingolia et al., 2009).

“Similar variation has been reported based on ribosome profiling (Ingolia et al., 2009).”

• cites Ingolia et al., 2009 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

In prior chemostat work under nitrogen limitation at a dilution rate of 0.1 hr−1, turnover was estimated for 641 proteins with a median of 0.064 hr−1 (Helbig et al., 2011).

“Previously, in chemostats at a dilution rate of 0.1 hr 1 and under nitrogen limitation conditions, protein turnover was estimated for 641 proteins at a median value of 0.064 hr 1 (Helbig et al., 2011).”

• cites Helbig et al., 2011 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Prior measurements of absolute protein levels used tandem affinity purification or GFP tags (Chong et al., 2015; Ghaemmaghami et al., 2003) and selected reaction monitoring proteomics (Picotti et al., 2009).

“Absolute protein levels have been measured before using tandem affinity purification- or GFP-tagged proteins (Chong et al., 2015; Ghaemmaghamiet al., 2003) and by a targeted proteomics approach based on selected reaction monitoring (Picotti et al., 2009).”

• cites Chong et al., 2015; Ghaemmaghamiet al., 2003Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Picotti et al., 2009 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Human tissue data from combined databases also showed relatively constant protein-mRNA ratios for individual proteins (Wilhelm et al., 2014).

“Similar results have been detected for human tissue, whereby data collected from databases were combined and relatively constant protein-mRNA ratios were found for individual proteins (Wilhelm et al., 2014).”

• cites Wilhelm et al., 2014 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

These results suggest that enzyme properties account for most metabolic control, except perhaps for mitochondrial proteins, and that global studies of enzyme activities are worth pursuing.

“These results suggest that most of the control in metabolism is attributable to enzyme properties (perhaps with the exception of mitochondrial proteins, as noted above) and opens a possibility to concentrate on global studies of the behavior of enzyme activities.”

• AssumptionThe authors state this as an assumption: The flux simulation assumed that yeast metabolism maximizes ATP consumption, i.e., that maximum ATP yield is the objective function.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Imbalanced flux in this pathway would cause reactive oxygen species accumulation and cytochrome c leakage to the cytoplasm, inducing apoptosis.

“Lack of flux balance in this pathway will lead to accumulation of reactive oxygen species and possible leakage of cytochrome c to the cytoplasm, resulting in induction of apoptosis.”

• AssumptionThe authors state this as an assumption: The flux simulation assumed that yeast metabolism maximizes ATP consumption, i.e., that maximum ATP yield is the objective function.

Probably not a substantial claim: may not pass S3 (load-bearing).

Transcriptional control of mitochondrial biogenesis and function has also been observed in higher eukaryotes (Hock and Kralli, 2009).

“Transcriptional control of mitochondrial biogenesis and function has also been observed for higher eukaryotes (Hock and Kralli, 2009)”

• cites Hock and Kralli, 2009Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Earlier work assumed translation initiation was the key step in E. coli translation efficiency, while folding energy and codon bias were claimed to dominate translation control in S. cerevisiae (Tuller et al., 2010).

“Previous studies analyzing the translational control for Escherichia coli have assumed that translation initiation was the most critical step in regulating translation efficiency, whereas folding energy and codon bias had earlier been claimed to account for the majority of translation control in S. cerevisiae (Tuller et al., 2010).”

• cites Tuller et al., 2010 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

A similar approach previously explained 67% of the variability in a human cell line (Vogel et al., 2010).

“A similar approach has been used previously, whereby 67% of the variability was described for a human cell line (Vogel et al., 2010).”

• cites Vogel et al., 2010 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

The study showed that mRNA abundance and translational elongation were the two main determinants of total protein synthesis and thus of translation efficiency.

“We showed that mRNA abundances and elongation of translation were the two major factors for determining the amount of total protein synthesis and, hence, describing the translation efficiency.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The net ATP production required for growth and non-growth processes under reference conditions was 59 mmol-ATP gDW−1 h−1.

“ATP requirement (growth and non-growth related ATP net production) under the reference environmental conditions was 59 mmol-ATP gDW 1 h 1 (Forster et al., 2003).”

• cites Forster et al., 2003 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

P-values in differential expression analysis were adjusted for multiple testing using the Benjamini-Hochberg procedure.

“P-values were adjusted for multiple testing using the Benjamini-Hochberg procedure (Benjamini and Hochberg, 1995) as implemented in DESeq.”

• cites Benjamini and Hochberg, 1995Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

RNA integrity was verified with a 2100 Bioanalyzer instrument.

“Integrity of the product was verified using 2100 Bioanalyzer instrument according to its user's manual (Agilent Technologies, Santa Clara, California, United States).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Protein turnover data were missing for the perturbed environmental conditions, preventing calculation of total synthesized protein for those conditions.

“Although we were missing protein turnover under those conditions to calculate total amount of synthesized proteins”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

RNA sequencing samples had 8.3-16.2 million mappable reads, with an average mapping rate of 88%.

“Each sample had between 8.3 to 16.2 million mappable reads, with an average map rate of 88%.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The fully labelled biomass standard had 98% of its proteogenic lysine labelled.

“fully labelled biomass was produced (in which 98% of proteogenic lysine was labelled, data not shown)”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

For each fitted protein turnover curve, 95% confidence intervals were calculated and lower and upper bounds were determined.

“For every fitted line 95% confidence intervals were calculated (2.5%-97.5% confidence interval) and lower bounds (LB) and upper bounds (UB) were determined with a median value of 36% (Table S5 and Figure S3).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Normalized heavy-to-light ratios, obtained by shifting the median peptide log ratio to zero, were used in all downstream quantitative analyses to correct for mixing errors.

“To account for any mixing errors normalized H/L ratios (by shifting median peptide log H/L ratio to zero) were used for all down-stream quantitative analyses (Cox and Mann, 2008).”

• cites Cox and Mann, 2008 sourceNot checked yet

  Not checked in the free run.

Classified as probably not a substantial claim. May combine two claims.

The false discovery rate for peptide-spectrum matches and proteins was maintained under 1% by a target-decoy method.

“Peptide-spectrum match and protein false discovery rate (FDR) was kept below 1% using a target-decoy approach.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The combined contribution of features is smaller than the sum of their individual contributions because the contributions overlap.

“It is important to note that combined contribution is smaller than the sum of individual contributions as there will be some overlap in the contributions.”