mTORC1 cooperates with tRNA wobble modification to sustain the protein synthesis machinery

Abstract

Synthesizing the cellular proteome is a demanding process that is regulated by numerous signaling pathways and RNA modifications. How precisely these mechanisms control the protein synthesis machinery to generate specific proteome subsets remains unclear. Here, through genome-wide CRISPR screens we identify genes that enable mammalian cells to adapt to inactivation of the kinase mechanistic target of rapamycin complex 1 (mTORC1), the central driver of protein synthesis. When mTORC1 is inactive, enzymes that modify tRNAs at wobble uridines (U₃₄-enzymes), Elongator and Ctu1/2, become critically essential for cell growth in vitro and in tumors. By integrating quantitative nascent proteomics, steady-state proteomics and ribosome profiling, we demonstrate that the loss of U₃₄-enzymes particularly impairs the synthesis of ribosomal proteins. However, when mTORC1 is active, this biosynthetic defect only mildly affects steady-state protein abundance. By contrast, simultaneous suppression of mTORC1 and U₃₄-enzymes depletes cells of ribosomal proteins, globally inhibiting translation. Thus, mTORC1 cooperates with tRNA U₃₄-enzymes to sustain the protein synthesis machinery and support the high translational requirements of cell growth.

Fig. 1. A CRISPR screen identifies tRNA wobble enzymes as essential for cell proliferation during mTOR inhibition.

a Population doublings of EPP2 iCas9 cells during the CRISPR screen in the presence of torin 1 [300 nM] or dimethyl sulfoxide (DMSO). b Gene-level enrichment or depletion of sgRNAs in the CRISPR screen upon torin 1 treatment. Dashed lines demarcate genes that are significantly depleted or enriched (log₂ fold change (log₂FC) > |2| , adj. p < 0.01). Selected hits are highlighted. c Schematic representation of tRNA wobble uridine (U₃₄) modification by the Elongator complex (Elp1-Elp6) and Cytosolic Thiouridylase (Ctu1-Ctu2). Enzyme subunits that scored as selectively essential during torin 1 treatment in the screen are highlighted. d, e Fold change (FC) in cell number of (d) Ctu1 iKO and (e) Elp3 iKO EPP2 cells after 3 days in culture ± torin 1 [25 nM] or rapamycin [12.5 nM]. f, g Fold change in cell number of Ctu1 KO f) EPP2 cells and (g) KRPC cells with ectopic expression of Ctu1 cDNA after 3 days ± rapamycin [12.5 nM]. h Fold change in cell number of Elp3 iKO EPP2 cells with ectopic expression of Elp3 cDNA after 3 days ± rapamycin [12.5 nM]. d–h Data are represented as replicate mean ± SD (d, e) n = 4, f–h n = 3 independent experiments in 3 technical replicates); p values were calculated by two-tailed unpaired t-test with Welch correction. Source data are provided as a Source Data file.

Fig. 2. Loss of tRNA wobble modification sensitizes cancer cells to mTORC1 pathway inhibition in vitro and in tumors.

a–c Fold change in cell number of single cell-derived clonal Ctu1 KO or control EPP2 cells expressing sgRNAs against (a) Raptor, (b) Rictor ± rapamycin [12.5 nM] and (c) Rheb after 3 days in culture. d Fold change in cell number of Ctu1 iKO EPP2 cells after 3 days in culture ± GDC-0941 [1 µM] or MK-2206 [2 µM]. a–d Data are represented as mean ± SD (n = 3 technical replicates). e Weight of orthotopic pancreatic tumors from Ctu1 iKO or control EPP2 cells in C57BL/6 J Rag2^−/− mice after 9 days treatment with rapamycin [5 mg/kg/day] or vehicle. Centre line represents median, upper and lower bounds of the box 75th and 25th percentiles, whiskers min to max (control ± rapamycin n = 6, Ctu1 iKO ± rapamycin n = 7); p values were calculated by unpaired two-tailed t-test with Welch correction. Source data are provided as a Source Data file.

Fig. 3. Loss of tRNA wobble enzymes leads to a global decrease in protein synthesis.

a, b Changes in protein synthesis of (a) Ctu1 iKO and (b) Elp3 iKO EPP2 cells, measured using the mass spectrometry (MS) based quantitative analysis of the newly synthesized proteome (QuaNPA) workflow (n = 3 experimental replicates). c Codon-specific changes in ribosome A site occupancy in Ctu1 iKO EPP2 cells, analyzed by diricore. VAA codons (i.e., AAA/CAA/GAA), which are recognized by U₃₄ wobble tRNAs, are highlighted. Data are represented as mean (n = 5 independent experiments); p values were calculated by out-of-frame analysis and are shown for significant increases. d, e Correlation of mRNA codon usage deviation and change in the newly synthesized proteome (NSP) of (d) Ctu1 iKO and (e) Elp3 iKO EPP2 cells (n = 3 experimental replicates). Newly synthesized proteome data are represented as decile ranks of log₂FC (i.e., 1 = 10 % most decreased proteins in iKO/Ctrl). In box plots, centre line represents median, upper and lower bounds of the box 75th and 25th percentiles, whiskers 10th to 90th percentiles; p values were calculated by two-tailed Wilcoxon rank test. Grey line represents median codon usage of all quantified proteins. Source data are provided as a Source Data file.

Fig. 4. tRNA wobble modification promotes ribosomal protein synthesis.

a GSEA showing highly depleted or enriched protein groups in the newly synthesize proteome of Ctu1 iKO EPP2 cells. NES, normalized enrichment score. b, c Scatter plot comparing changes in the newly synthesized proteome versus transcriptome of (b) Ctu1 iKO and (c) Elp3 iKO EPP2 cells. Transcript levels were quantified by RNA sequencing (n = 2 independent experiments). d Changes in ribosome A site occupancy of U₃₄-wobble enzyme-dependent/independent VAA/VAG codons on ribosomal protein mRNAs in Ctu1 iKO EPP2 cells, analyzed by diricore. Data are represented as mean (n = 5 independent experiments); p values were calculated by out-of-frame analysis and are shown for significant increases. e, f Correlation of mRNA codon usage deviation for (e) AAA and (f) VAA with changes in the newly synthesized proteome of Ctu1 iKO and Elp3 iKO EPP2 cells. Data are represented as median and interquartile range. g, h Correlation of mRNA codon usage deviation and change in the steady-state proteome of (g) Ctu1 iKO and (h) Elp3 iKO EPP2 cells. Steady-state proteomes were quantified by label-free mass spectrometry (n = 5 independent experiments); proteome data are represented as decile ranks of log₂FC (i.e., 1 = 10% most decreased proteins in iKO / Ctrl). In box plots, centre line represents median, upper and lower bounds of the box 75th and 25th percentiles, whiskers 10th to 90th percentiles; p values were calculated by two-tailed Wilcoxon rank test. Grey line represents median codon usage of all quantified proteins. i GSEA showing the top depleted protein groups in the steady-state proteome of Ctu1 iKO EPP2 cells. j, k Scatter plot comparing changes in the newly synthesized proteome versus steady-state proteome of ( j) Ctu1 iKO and (k) Elp3 iKO EPP2 cells. Source data are provided as a Source Data file.

Fig. 5. mTORC1 and tRNA wobble enzymes converge on sustaining ribosomal protein synthesis.

a, b Codon-specific changes in ribosome A site occupancy in (a) control and (b) Ctu1 iKO EPP2 cells after 2 h ± torin 1 [300 nM], analyzed by diricore. VAA codons cognate for U₃₄ wobble tRNAs are highlighted. Data are represented as mean (n = 6 independent experiments); p values were calculated by out-of-frame analysis and are shown for significant increases and for all U₃₄-enzyme-dependent codons. c, d Changes in global and ribosomal protein abundance in the newly synthesized proteome of EPP2 cells after 5 h ± (c) torin 1 [50 nM] and (d) rapamycin [50 nM], quantified by the QuaNPA workflow (n = 3 experimental replicates). In box plots, centre line represents median, upper and lower bounds of the box 75th and 25th percentiles, whiskers 10th to 90th percentiles. e, f Correlation of mRNA AAA codon usage deviation with changes in the newly synthesized proteome of EPP2 cells after 5 h ± (e) torin 1 [50 nM] and (f) rapamycin [50 nM]. Shown are protein groups whose synthesis is particularly dependent on Ctu1. Data are represented as median and interquartile range. Source data are provided as a Source Data file.

Fig. 6. Ctu1 promotes efficient synthesis of ribosomal proteins in a codon-dependent manner.

a, b Scatter plot comparing the newly synthesized proteome of EPP2 cells after 5 h mTORC1 inhibitor ± Ctu1 iKO (Ctu1 effect) or Ctu1 iKO ± 5 h mTORC1 inhibitor (inhibitor effect). mTORC1 inhibitors were a) torin 1 [50 nM] and (b) rapamycin [50 nM]. Newly synthesized proteomes were quantified using the QuaNPA workflow (n = 3 experimental replicates). c–h Expression of HA-tagged ribosomal proteins (c, d) Rpl29, (e, f) Rps25 and (g, h) Rps3 in Ctu1 iKO EPP2 cells. Expression of wild type (WT) and VAA-to-VAG mutant (VAG Mut.) ribosomal protein variants was analyzed after 40 h + rapamycin [50 nM], MG132 [200 nM] by quantitative immunoblotting. Codon usage deviation of wild type proteins is indicated. Data are normalized to WT sgRNA controls and represented as mean ± SEM (n = 4 independent experiments); p values were calculated by two-tailed one sample t-test with a hypothetical mean of 1. Source data are provided as a Source Data file.

Fig. 7. Concerted suppression of mTORC1 and tRNA wobble enzymes depletes cells of ribosomal proteins.

a–c Changes in global and ribosomal protein abundance in the steady-state proteome of EPP2 cells. a Ctu1 iKO and Elp3 iKO; (b) control after 40 h ± rapamycin [50 nM]; (c) Ctu1 iKO and Elp3 iKO after ± 40 h rapamycin [50 nM]. Proteomes were quantified by label-free mass spectrometry (n = 5 independent experiments). In box plots, centre line represents median, upper and lower bounds of the box 75th and 25th percentiles, whiskers 10th to 90th percentiles. d, e Levels of ribosomal proteins Rpl29 or Rps3 in Ctu1 iKO and Elp3 iKO EPP2 cells after 40 h ± torin 1 [300 nM] or rapamycin [50 nM], analyzed by quantitative immunoblotting. f–m Changes in Rpl29 and Rps3 abundance in different U₃₄ enzyme-deficient cell lines after 40 h ± rapamycin [50 nM], analyzed by quantitative immunoblotting. f, g Ctu1/Elp3 iDKO EPP2 cells; (h, i) Ctu1 KO T24 cells; (j, k) Ctu1/Elp3 iDKO KRas^G12D MEFs; (l, m) Ctu1/Elp3 iDKO KPC cells. e–m Data are normalized to sgRNA controls (dashed line) and represented as mean ± SEM (e) n = 8, (g) n = 3, (i) n = 4, k), m n = 5 independent experiments); p values were calculated by two-tailed one sample t-test with a hypothetical mean of 1. Source data are provided as a Source Data file.

Fig. 8. mTORC1 cooperates with tRNA wobble enzymes to sustain translational capacity.

a–g Puromycin (puro) incorporation assay in U₃₄-enzyme-deficient EPP2 cells after 16 h inhibitor treatment, analyzed by quantitative immunoblotting. Asterisk denotes an unspecific band. a, b Ctu1 iKO ± torin 1 [300 nM]; (c) Ctu1 iKO ± GDC-0941 [1 µM], MK-2206 [2 µM], rapamycin [12.5 nM]; (d, e) Elp3 iKO ± torin 1 [300 nM]; (f, g) Ctu1/Elp3 iDKO ± torin 1 [300 nM]. Data are normalized to sgRNA controls (dashed line) and represented as mean ± SEM (b, e) DMSO n = 8, torin 1 n = 4, (c, g) n = 4 independent experiments). p values were calculated by two-tailed one sample t-test with a hypothetical mean of 1. h Puromycin incorporation assay in Elp3 iKO EPP2 cells after torin 1 [300 nM] treatment for the indicated periods, analyzed by immunoblotting. i, j Puromycin incorporation assay in Elp3 iKO EPP2 cells after 40 h torin 1 [300 nM] followed by 1 h washout of torin 1, analyzed by quantitative immunoblotting. Data are normalized to sgRNA control + torin 1 and represented as mean ± SEM (n = 5 independent experiments); p values were calculated by two-tailed unpaired t-test with Welch correction. Source data are provided as a Source Data file.