Skip to main content
. 2022 May 27;11:e73466. doi: 10.7554/eLife.73466

Figure 1. Identification of translational regulators using polysomal proteomics.

(A) Growth curves and doubling times for unstressed, 0.45 mM hydrogen peroxide (H2O2)-treated and 10 mM 3-amino-1,2,4-triazole (3-AT)-treated cultures (n = 3). The time of H2O2 or 3-AT addition is indicated. (B) Quantification of polysome-to-monosome (P/M) ratios under the three conditions (n = 4–19). Error bars show standard deviation (SD). The t-test was used to compare the conditions: ns – not significant (p > 0.05), **p < 0.01, ***p < 0.001. (C) Overview of polysomal proteomics. Monosomal (F1) and polysomal (F2–F5) fractions were isolated from unstressed, H2O2- and 3-AT-treated extracts and analysed using label-free mass spectrometry (MS). (D) The number of proteins identified reproducibly (≥2 replicates) in each fraction. Venn-style diagrams of overlaps between conditions for proteins found across (E) all five fractions or (F) the first four fractions.

Figure 1.

Figure 1—figure supplement 1. Formaldehyde crosslinking prevents ribosome run-off and retains RNA-binding proteins (RBPs) in polysome fractions similar to cycloheximide treatment.

Figure 1—figure supplement 1.

(A) Left: sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) analysis of fractions from unstressed and hydrogen peroxide (H2O2)-treated extracts run on 15–50% sucrose gradients. Right: entire lanes were quantified (LI-COR Image Studio) and the percentage of the total signal that each fraction accounts for was calculated. (B) Polysome profiles and western blots comparing cycloheximide treatment (Cyc) and formaldehyde crosslinking (XL) of cultures for preparing extracts for polysome profiling. (C) Bands from (B) were quantified, as for panel A.
Figure 1—figure supplement 2. Polysomal proteomics data are reproducible and can separate the different fractions.

Figure 1—figure supplement 2.

(A) r2 values for pairwise correlations between replicates (R1–R4) for each sample. (B) Principal component analysis (PCA) of the different samples under each condition. The percentage of the variation explained by each of the first two principal components is indicated. The axes are not equivalent for the different conditions. Note that only protein groups present in all samples for a given condition could be included in the comparison. (C) Pairwise comparisons between conditions for the totals. r2 values from linear regressions are indicated. Protein groups are coloured by their functional category: RP – ribosomal protein, TF – translation factor, RBP – RNA-binding protein, non-RBP – other protein.
Figure 1—figure supplement 3. Nascent peptides are not a major contributor to protein signal.

Figure 1—figure supplement 3.

We hypothesised that proteins represented in polysome fractions by nascent chains and/or co-translational complex formation would show greater intensity from N-terminal peptides in F1–F5 than in totals, as they are synthesised in the N-to-C direction. (A) Peptide midpoint calculation and example cumulative peptide intensity distribution. Steps in the cumulative intensity profiles indicate peptide midpoints. For a protein detected in polysome fractions (Fraction) through only nascent chains, a shift in signal accumulation compared with a cytoplasmic extract (Total) would be expected, such that a much greater proportion of the total peptide intensity would be expected from the N-terminal end of the protein. (B) Median number of peptides and median sequence coverage (percentage of the amino acid residues that were detected by mass spectrometry (MS) in at least one peptide) for individual proteins in unstressed samples (371 uniquely detected proteins). Proteins in the shaded area were used for the analysis (142 proteins). Cumulative peptide intensity distributions for fractions and totals of named proteins. Differences between fractions and totals were tested using the Kolmogorov–Smirnov test. (C) The proteins with significantly different fraction distributions compared with totals (p < 0.05). The fractions that are significantly different are indicated by number. Only 6 of 142 showed some evidence of a bias towards the N-terminus in one or more fractions (Asc1, Cdc60, Gpm1, Gpp1, Map1, and eIF4A/Tif2), while 5 others (Fba1, Rpl16a, Rps7a, Scp160, and eEF1A/Tef2) showed different patterns. (D) Representative examples of proteins without significant differences between fractions and totals. Dashed lines are y = x.