Ribosome profiling reveals ribosome stalling on tryptophan codons and ribosome queuing upon oxidative stress in fission yeast

Abstract

Translational control is essential in response to stress. We investigated the translational programmes launched by the fission yeast Schizosaccharomyces pombe upon five environmental stresses. We also explored the contribution of defence pathways to these programmes: The Integrated Stress Response (ISR), which regulates translation initiation, and the stress-response MAPK pathway. We performed ribosome profiling of cells subjected to each stress, in wild type cells and in cells with the defence pathways inactivated. The transcription factor Fil1, a functional homologue of the yeast Gcn4 and the mammalian Atf4 proteins, was translationally upregulated and required for the response to most stresses. Moreover, many mRNAs encoding proteins required for ribosome biogenesis were translationally downregulated. Thus, several stresses trigger a universal translational response, including reduced ribosome production and a Fil1-mediated transcriptional programme. Surprisingly, ribosomes stalled on tryptophan codons upon oxidative stress, likely due to a decrease in charged tRNA-Tryptophan. Stalling caused ribosome accumulation upstream of tryptophan codons (ribosome queuing/collisions), demonstrating that stalled ribosomes affect translation elongation by other ribosomes. Consistently, tryptophan codon stalling led to reduced translation elongation and contributed to the ISR-mediated inhibition of initiation. We show that different stresses elicit common and specific translational responses, revealing a novel role in Tryptophan-tRNA availability.

Figure 1.

General responses to stress. (A) Scatter plot comparing mRNA levels and translation efficiencies (log₂ ratios stress/control) upon cadmium treatment (15 min) in wild type, eIF2α-S52A and sty1Δ genetic backgrounds. CESR-induced genes are plotted in red. (B) Same dataset as in A, but CESR-repressed genes are highlighted in blue. (C) Western blots comparing eIF2α phosphorylation levels after the five treatments for the indicated times in wild type cells. In the first two lanes (left), eIF2α-S52A cells were used as negative control for the anti-eIF2α phosphorylation antibody. Tubulin was employed as a loading control. (D–F) Representative polysome profile traces before and after cadmium treatment (15 min) in wild type (D), eIF2α-S52A (E) and sty1Δ cells (F). (G) Quantification of polysomes to subpolysomes ratios after five stress treatments (15 min). The data are normalised to control unstressed cells. Unnormalized values are presented in Supplementary Figure S2B. The data are shown for two independent biological replicates of each experiment.

Figure 2.

Translational regulation upon stress exposure. (A) Scatter plot comparing mRNA levels and translation efficiencies (log₂ ratios stress/control) upon cadmium treatment (15 min) in wild type, eIF2α-S52A and sty1Δ genetic backgrounds. The fil1 gene is plotted in red and genes encoding ribosomal proteins in blue. Note that this is the same dataset shown in Figure 1A and B, but with different genes highlighted. (B) Boxplots comparing translation efficiencies of stressed and control cells (log₂ ratios stress/control) of genes encoding ribosomal proteins. Data are shown for wild type, eIF2α-S52A and sty1Δ cells. (C) Comparisons of translation efficiencies of stressed (15 min) and control cells (log₂ ratios stress/control). Only genes that showed significant TE upregulation in wild type cells in at least one stress are displayed. Data are presented for wild type, eIF2α-S52A and sty1Δ cells. (D) As in C, but TE changes have been normalised to those of wild type cells. (E) As in C, but only genes that showed significant translational downregulation are displayed. Dots corresponding to fil1 gene are shown in red. (F) As in D, but data are displayed for significantly downregulated genes.

Figure 3.

Fil1 is the major translational responder to stress. (A) Comparison of translation efficiency of the fil1 gene between stressed and control cells (log₂ ratios stress/control). The dotted lines indicate 1.5-fold changes. Data are presented for wild type, eIF2α-S52A and sty1Δ cells. (B) As in C, but for fil1 mRNA changes. The dotted lines indicate 2-fold changes. (C) Western blots to measure Fil1-TAP protein levels after cadmium treatment for the indicated times. Data are presented for wild type, eIF2α-S52A and sty1Δ cells. Tubulin was used as a loading control. (D) As in C, but after H₂O₂ treatment.

Figure 4.

Role of Fil1 in the transcriptional responses to stress. (A–C) Boxplots comparing mRNA levels of stressed and control cells (log₂ ratios stress/control) of Fil1 targets. Data are shown for wild type and fil1Δ cells at the indicated times and stresses. (D) Venn diagram showing the overlap between genes expressed at lower levels in fil1Δ mutant relative to wild type cells after heat shock (15 min), and Fil1 targets in unstressed cells. The P value of the observed overlap is shown. (E) As in D, but genes expressed at lower levels in fil1Δ mutant relative to wild type after H₂O₂ treatment (60 min) were compared to Fil1 targets in unstressed cells. (F) As in A to C, but after cadmium, heat shock and H₂O₂ treatments, in wild type and eIF2α-S52A strains. (G) As in F, but after H₂O₂ treatment and with an additional time point. (H) As in F, but in the wild type and sty1Δ strains.

Figure 5.

Levels of charged tRNA-Trp are affected by oxidative stress. (A) Scatter plots showing log₂ relative codon occupancies before and after H₂O₂ treatment for 15 min in wild type cells. The dot corresponding to UGG codon encoding tryptophan is shown in black. (B) Metagene depicting average read density of RPFs around tryptophan codons (UGG) or one of the histidine codons (CAC). The cartoon shows the interpretation of the results of the experiment, with ribosomes queuing upstream of the tryptophan codon-stalled ribosome. (C) Changes in intracellular tryptophan levels in response to H₂O₂ exposure. Tryptophan levels were measured at the indicated times after H₂O₂ addition to the culture medium. Each dot corresponds to an independent biological replicate (n = 4), and the horizontal lines indicate the means. No adjustment for multiple testing was performed. (D) Representative northern blot for the determination of tRNA-Trp charging levels before and after H₂O₂ exposure. The top blot was hybridised with a probe against tRNA-Trp, and the bottom one with a probe against the U5 snRNA. In the upper blot, the top band corresponds to charged tRNA, and the bottom to the uncharged form. tRNA-Trp samples were either deacylated to remove the linked amino acid from charged tRNAs (sample D) or oxidised to remove the unprotected 3′ nucleotides from uncharged tRNAs by beta-elimination (sample B–E) (see Materials and Methods for details). U5 snRNA was used as a loading control. (E) Quantification of tRNA-Trp charging ratios from experiment D. Ratios between charged and uncharged tRNA were calculated. Each dot corresponds to an independent biological replicate (n = 5), and the horizontal line indicates the mean. (F) As in D, but using a probe against tRNA-His (top panel) or U5 snRNA (bottom). (G) Quantification of tRNA-His charging from the experiment shown in F (n = 2 independent replicates). (H) Northern blot as in D, to explore the effects of supplementing the culture medium with tryptophan. Cells were grown in the presence of tryptophan for 0, 15 or 120 min, and H₂O₂ was added at the indicated times (0, 15 min) before the end of the incubation with tryptophan. Control deacylated RNA (sample D) is used to identify the location of uncharged tRNA. (I) Quantification of tRNA-Trp charging levels from experiment H, right panels (n = 2 independent replicates). (J) Scatter plots showing log₂ relative codon occupancies before and after H₂O₂ and tryptophan treatment for 15 min in wild type cells. The dot corresponding to the UGG codon, encoding tryptophan, is shown in black. (K) Metagene depicting average read density of RPFs around tryptophan codons (UGG) or one of the histidine codons (CAC) in different conditions.

Figure 6.

Oxidative stress affects the translation efficiency of tryptophan codon-enriched genes, and decreased tRNA-Trp charging may affect eIF2α phosphorylation. (A) Western blots to investigate the effect of tryptophan on eIF2α phosphorylation levels after H₂O₂ treatment. Cells were treated with H₂O₂ for 15 min and supplemental tryptophan was added as indicated. Tubulin was used as a loading control. (B) Quantification of eIF2α phosphorylation normalized to total eIF2α from the experiment shown in A (n = 4 independent replicates). (C) Boxplots showing changes in translation efficiency upon oxidative stress (log₂ TE ratios stress/control, 15 min treatment) according to tryptophan codon content. Genes were binned into 11 categories based on the fraction of tryptophan codons in their coding sequences (the first group contains 269 genes without tryptophan, codons and the other 10 groups have 234 genes each). The horizontal red dashed line indicates the median of the second group. (D) As above, but displaying coding sequence lengths.