High-Resolution Ribosome Profiling Defines Discrete Ribosome Elongation States and Translational Regulation during Cellular Stress

Abstract

Ribosomes undergo substantial conformational changes during translation elongation to accommodate incoming aminoacyl-tRNAs and translocate along the mRNA template. We used multiple elongation inhibitors and chemical probing to define ribosome conformational states corresponding to differently sized ribosome-protected mRNA fragments (RPFs) generated by ribosome profiling. We show, using various genetic and environmental perturbations, that short 20-22 or classical 27-29 nucleotide RPFs correspond to ribosomes with open or occupied ribosomal A sites, respectively. These distinct states of translation elongation are readily discerned by ribosome profiling in all eukaryotes we tested, including fungi, worms, and mammals. This high-resolution ribosome profiling approach reveals mechanisms of translation-elongation arrest during distinct stress conditions. Hyperosmotic stress inhibits translocation through Rck2-dependent eEF2 phosphorylation, whereas oxidative stress traps ribosomes in a pre-translocation state, independent of Rck2-driven eEF2 phosphorylation. These results provide insights and approaches for defining the molecular events that impact translation elongation throughout biology.

Keywords: eEF2 phosphorylation; ribosome functional states; ribosome profiling; translation elongation.

Figure 1.. Assigning ribosome functional states to distinct footprint sizes (21 and 28 nt RPFs) from ribosome profiling samples.

(A) Schematic representation of the eukaryotic elongation cycle. PreAcc: pre-accommodation; PrePT: Pre-peptide bond formation; PreTrans: Pre-translocation. (B and C) Scatter plots showing mutation rates of 25S rRNA (a function of DMS reactivity) comparing CHX-pretreated (B) or ANS-pretreated (C) relative to mock pretreatment. Nucleotides protected by CHX or ANS are color-coded and labeled. (D) Fold change in mutation rates of nucleotides A1755 and A1756 for CHX-pretreatment with in vivo DMS modification (CHX in vivo), ANS-pretreatment with in vivo DMS modification (ANS in vivo), ANS-pretreatment with DMS modification in lysate (ANS lysate), and TIG with DMS modification in lysate (TIG lysate) (n=2, ±SD). (E) Average ribosome occupancies aligned at stop codons using 21 nt RPFs (top), 28 nt RPFs (middle) and both RPFs (bottom) for WT and eRF1-depleted (eRF1d) cells. (F) Scatter plot comparing 64 codon-specific ribosome occupancies (pause scores) for 21 nt RPFs in WT and eRF1d cells.

Figure 2.. 21 nt RPFs correlate with the tRNA abundance metrics.

(A) Spearman rank correlations of A-site codon-specific occupancies with the inverse of tRNA adaptation index (tAI) from ribosome profiling libraries prepared with CHX (our samples or (Matsuo et al., 2017)) or no elongation inhibitor in the lysates (n≥2, ±SD). (Lareau et al., 2014; Tunney et al., 2017) (B) Length distributions of ribosome footprints comparing libraries prepared with CHX only (grey), CHX/TIG (purple), and CHX/ANS (green) in the lysates. (C) Spearman rank correlations of A-site codon-specific occupancies with 1/tAI from samples described in (B) (n≥2, ±SD). p-values for 21 nt RPFs from Student’s t-test after Fisher’s z-transformation is indicated by asterisks. ****, p< 0.0001. (D and E) Scatter plots of codon-specific ribosome occupancies for 21 nt RPFs comparing gamma-toxin treated to untreated cells from libraries prepared with CHX only (D) or CHX/TIG (E) in the lysates. GAA codons are colored in blue and labeled. (F and G) Scatter plots of codon-specific ribosome occupancies for 21 nt RPFs comparing 3-AT treated relative to untreated cells from libraries prepared with CHX only (F) or CHX/TIG (G) in the lysates. Both histidine codons are colored in green and labeled.

Figure 3.. Conserved ribosome functional states in metazoan cells visualized by distinct footprint sizes.

(A) Size distribution of ribosome footprints in C. elegans embryos. (B) Size distributions of ribosome footprints in HeLa and MDA-MB-231 cells. (C) Size distributions of ribosome footprints mapped to glutamine codons in untreated and starved cells. (D) Scatter plot of codon-specific ribosome occupancies for 21 nt RPFs comparing starved relative to untreated cells.

Figure 4.. Ribosome profiling reveals translation elongation regulation under hyperosmotic stress.

(A) Size distributions of ribosome footprints for WT cells under hyperosmotic (blue), oxidative (orange) and stationary phase (green) stress conditions relative to untreated (black). (B) Similar to (A), size distributions of ribosome footprints for rck2Δ cells under hyperosmotic (blue) or oxidative (orange) stress conditions compared to WT unstressed cells (black). (C) Scatter plot of codon-specific occupancies for 28 nt RPFs comparing oxidative stress relative to unstressed condition in WT cells. r, Pearson correlation coefficient. (D) Similar to (C), comparing rck2Δ cells under oxidative stress relative to WT unstressed cells. (E) Immunoblot of a Phos-tag gel for eEF2 phosphorylation under unstressed, hyperosmotic or oxidative stress conditions in WT and rck2Δ cells with or without λ phosphatase treatment. (F) Scatter plot of codon-specific ribosome occupancies for 21 nt RPFs comparing WT cells under oxidative stress relative to unstressed condition. All proline codons are colored in red and labeled. (G) Northern blots for Pro-tRNA^AGG, Glu-tRNA^UUC and Thr-tRNA^AGU in unstressed, hyperosmotic and oxidative (10 and 45 min) stress conditions.