Ribosomal protein S7 ubiquitination during ER stress in yeast is associated with selective mRNA translation and stress outcome

eIF2α phosphorylation-mediated translational regulation is crucial for global translation repression by various stresses, including the unfolded protein response (UPR). However, translational control during UPR has not been demonstrated in yeast. This study investigated ribosome ubiquitination-mediated translational controls during UPR. Tunicamycin-induced ER stress enhanced the levels of ubiquitination of the ribosomal proteins uS10, uS3 and eS7. Not4-mediated monoubiquitination of eS7A was required for resistance to tunicamycin, whereas E3 ligase Hel2-mediated ubiquitination of uS10 was not. Ribosome profiling showed that the monoubiquitination of eS7A was crucial for translational regulation, including the upregulation of the spliced form of HAC1 (HAC1i) mRNA and the downregulation of Histidine triad NucleoTide-binding 1 (HNT1) mRNA. Downregulation of the deubiquitinating enzyme complex Upb3-Bre5 increased the levels of ubiquitinated eS7A during UPR in an Ire1-independent manner. These findings suggest that the monoubiquitination of ribosomal protein eS7A plays a crucial role in translational controls during the ER stress response in yeast.


Results
Monoubiquitination of eS7A is required for translational regulation during the UPR in yeast. Ribosome ubiquitination increases significantly upon induction of the UPR in mammalian cells 32 .
To assess the role of ribosome ubiquitination in the UPR in yeast, the levels of ubiquitinated ribosomal proteins were evaluated by affinity purification of ribosomes with FLAG-tagged Rpl25 from cells expressing N-terminal Myc-tagged Ubiquitin protein (Ub), as previously described 22 . The levels of the ubiquitinated ribosomal proteins uS10, uS3 and eS7A were substantially increased under UPR conditions (Fig. 1a). Not4 is an E3 ligase for the ribosomal proteins eS7A and eS7B in yeast 33,34 , and is involved in translation repression 35 . Cells expressing not4∆ mutant showed Tm-sensitive growth (Fig. 1b). In addition, the eS7a-4KR mutant, which involves lysineto-arginine substitutions at all four Not4-specific ubiquitination sites, had the same level of Tm sensitivity as not4∆ mutant cells (Fig. 1b). Monoubiquitinated eS7A was not detected in not4∆ mutant cells (Fig. 1c), confirming that Not4 is responsible for the monoubiquitination of eS7A during the UPR. Because Hel2 was previously reported to form K63-linked polyubiquitin chains on Not4 monoubiquitinated eS7A and to play a crucial role in No-Go Decay (NGD) 24 , we assessed the involvement of Hel2 in the UPR. In contrast to Not4, the deletion of Hel2 (hel2∆) did not affect cell growth in the presence of Tm (Fig. 1b). In addition, neither the uS10 nor the uS3 mutant of Hel2-target lysine residues (uS10-K6/8R and uS3-K212R) affected cell growth in the presence of Tm (Fig. 1b). The uS10-Ub and uS3-Ub signals were not detected in the hel2∆ deletion mutant, whereas the eS7-Ub signal was not detected in the not4∆ deletion mutant ( Supplementary Fig. 1), suggesting that eS7 ubiquitination is dependent on Not4. We previously reported that Hel2 elongates the ubiquitin chain at eS7 after Not4dependent monoubiquitination, which is essential for mRNA quality control in NGD 36 . However, in contrast to not4∆ and the eS7-4KR mutant, the hel2∆ mutant did not show sensitivity to Tm (Fig. 1b). These results suggest that Not4-mediated monoubiquitination of the ribosomal protein eS7A is indispensable for cell survival under ER stress conditions, whereas Hel2-mediated polyubiquitination of eS7A is not. Figure 1. Not4-mediated monoubiquitinated eS7A is required for translational controls in the UPR in yeast. (a) Ubiquitinated proteins in the ribosome after the addition of tunicamycin. Yeast cells harboring pCUP1p-MYC-UBI and pRPS2(uS5)-FLAG or pRPL25(uL23)-FLAG were cultured in 800 mL of synthetic complete medium. Myc-Ubi expression was induced by culturing the cells in the presence of 0.1 mM Cu 2+ for 2 h. Cell lysates were prepared and FLAG-tagged ribosomes were purified using an M2 FLAG-affinity resin (Sigma), as described 22 . Affinity purified samples were subjected to SDS-PAGE followed by western blotting with an anti-Myc antibody. The arrows indicate proteins previously identified by mass spectrometry. (b) The Not4 ubiquitination of ribosomal protein eS7A is crucial for UPR in yeast. Genetic screening was performed to identify the E3 ubiquitin enzyme NOT4 required for resistance to Tm. (c) Dependence of eS7A mono-and poly-ubiquitination on Not4. (d) Ribosome profiling showing up-and down-regulation of translation by the eS7A ubiquitination. The ribosome profiling and RNA-seq results represent two independent biological replicates. The correlations between replicates are shown in Supplementary Fig. 3a,b. (f) eS7A ubiquitination-dependent up-and downregulation of specific mRNAs in response to UPR. The mRNA most upregulated by eS7A ubiquitination was HAC1, and the mRNA most upregulated was HNT1. (g) UPR does not inhibit bulk translation in wild-type and mutant cells. (h) Phosphorylation of eIF2α in response to amino acid depletion or Tm treatment. Shown are the levels of eIF2α phosphorylation in WT and S52A mutants in response to amino acid starvation and the presence of Tm. (a, c, h) Cropped gels or blots were display. All uncropped images are available in Supplemental Figure S7. www.nature.com/scientificreports/ These results suggested that translational control may play a role in the UPR in yeast and may involve eS7A ubiquitination. To test this, we then performed RNA-seq and ribosome profiling to investigate the regulation of translation in response to ER stress, and estimated translation efficiency (TE) by assessing both mRNA abundance and ribosome occupancy. During the early response, within 1 h after Tm treatment, the translation of 20 mRNAs was statistically and significantly changed (Supplementary Fig. 2a; q-value < 0.01), whereas > 200 mRNAs were up-or down-regulated 4 h after Tm treatment ( Supplementary Fig. 2a). eS7A ubiquitination-dependent translational regulation was therefore monitored at 4 h. A modest translational response was observed in eS7A-4KR mutant cells, with statistically significant changes in the translation of 214 mRNAs (Fig. 1d,e; Supplementary Fig. 2b; q-value < 0.01). To examine the eS7A ubiquitination dependency of the involved mRNAs, TE was compared in eS7A-WT and eS7A-4KR mutant cells (Fig. 1d,e), with 29 and 12 mRNAs categorised as being up-and down-regulated, respectively, in response to eS7A ubiquitination (Fig. 1f). These subsets were identified using the formula "log2 TE fold change (eS7A-WT)-log2 TE fold change (eS7A-4KR)", with mRNAs scored as > 2 and < − 2 defined being up-and down-regulated, respectively, in response to eS7A ubiquitination. Polysome analysis showed that UPR moderately repressed general translation in wild-type and eS7A-4KR mutant cells (Fig. 1g). The levels of phosphorylated eIF2α did not increase during the UPR (Fig. 1h), suggesting that initiation of general translation was not inhibited by phosphorylated eIF2α during the UPR. To rule out the effect of eS7A-4KR mutations on general translation, protein synthesis rates were measured in eS7A-WT and the eS7A-4KR mutant by puromycin labelling. Protein synthesis rates did not differ significantly in eS7A-WT and eS7A-4KR ( Supplementary Fig. 3a,b), indicating that eS7A monoubiquitination is involved in regulating the translation of specific mRNAs in response to ER stress.
Not4-mediated monoubiquitination of eS7A plays a crucial role in upregulating translation during the UPR, with the translation of HAC1 most drastically upregulated in response to the UPR. The downregulation of translation of specific mRNAs during the UPR was also abrogated in eS7A-4KR mutant cells, with the translation of mRNA encoding Histidine triad NucleoTide-binding 1 (HNT1) being most drastically repressed (Fig. 1e). Hac1 induces the expression of long un-decoded transcript isoforms, and downregulation of HNT1 translation depends on the uORFs, leading to protein downregulation in response to the UPR 37 . Ribosome profiling confirmed the repression of HNT1 translation, with the TE being downregulated approximately 60-fold 4 h after Tm treatment. Importantly, the UPR-associated downregulation of HNT1 was significantly diminished in eS7A-4KR mutant cells (4-fold; Supplementary Fig. 2d), despite its level of mRNA being slightly increased ( Supplementary Fig. 2d).
Not4-mediated eS7A monoubiquitination is required for translational regulation of HAC1 i and HNT1 mRNA. We identified Hac1 as the most significantly upregulated gene during the UPR (Fig. 1d,e). HAC1 u mRNA is stored in the cytoplasm in the absence of ER stress, and its translation is tightly suppressed by a base-pairing interaction between the intron and the 5′untranslated region (5′UTR) 13,14,38 . Excision of the intron by Ire1-dependent splicing in response to ER stress leads to robust translation of HAC1 i mRNA, with the resulting Hac1 protein upregulating UPR target gene expression. Tm drastically and rapidly increased the TE of HAC1 i mRNA ( Fig. 2a; Supplementary Fig. 12a). This increase was less robust in eS7A-4KR mutant than in eS7A-WT cells (93.05-fold vs. 5.89-fold, respectively; Fig. 2a). Mapping of the footprints throughout HAC1 i mRNA in eS7A-WT and eS7A-4KR mutant cells (Fig. 2b) suggested that this mRNA does not contain a strong translation pause site.
To validate these results by ribosome profiling, the expression of HAC1 was evaluated at the splicing, translation and protein levels. Hac1 protein was detected 0.5 h after the addition of Tm to WT cells, but its signal was much weaker in not4∆ and eS7A-4KR mutant cells, being approximately 50% of the levels in eS7A-WT cells (Fig. 2c). The level of Hac1 protein normalized to that of tubulin also indicated that Hac1 protein was significantly downregulated in the eS7A-4KR mutant, being approximately 50% of the levels in eS7A-WT cells 4 h after Tm treatment (Fig. 2d). In eS7A-4KR mutant cells, induction of HAC1 mRNA by Tm treatment was intact ( Fig. 2e-g), although this induction was moderately delayed by about 30 min after the addition of Tm (Fig. 2f,g). The splicing efficiency (the ratio of HAC1 u + HAC1 i mRNAs to HAC1 i mRNA) was moderately reduced in not4∆ and eS7A-4KR mutant cells (Fig. 2g).
In comparing TE in eS7-WT and eS7-4KR cells 4 h after the addition of Tm based on reads of HAC1, we found that the translation of HAC1 mRNAs was approximately 4-fold lower in eS7-4KR than in wild-type (WT) cells  www.nature.com/scientificreports/ ( Fig. 2i). To validate TE by other methods, we analyzed the distribution of HAC1 mRNA by sucrose gradient ultracentrifugation (Fig. 2j,k). In the absence of Tm, the HAC1 u mRNAs were detected on polysomes, with their distribution and levels being similar in eS7A-4KR and eS7A-WT cells (Fig. 2f,j,k). In the absence of Tm treatment, however, almost no footprints of ribosomes on HAC1 u were observed in eS7A-4KR and eS7A-WT cells (Fig. 2b), consistent with previous findings. These results may have been due to the tightly packed configuration of ribosomes on HAC1 u mRNA. After Tm treatment for 4 h, the levels of HAC1 i mRNA on polysomes were moderately but significantly lower in eS7A-4KR than in eS7A-WT cells (Fig. 2b,k). Overall, the changes in the levels of HAC1 i mRNA on polysomes were consistent with TE calculated based on the levels of HAC1 i mRNA and Hac1 protein, or approximately 2-fold (Fig. 2h). The imperfect correlation of TE with the ratio of protein to mRNA level may be due to a difference in the efficiency of recovery of mRNA reads from ribosomes translating HAC1 i mRNA in eS7A-WT and eS7A-4KR mutant cells.
Not4 monoubiquitinates eS7 at four lysine residues ( Supplementary Fig. 4a), with K83 ubiquitination being primarily responsible for mRNA quality control 24 . To identify the ubiquitination site(s) of eS7A required for the upregulation of HAC1 i , HAC1 u mRNA splicing and Hac1 protein levels were examined in four mutants containing a single lysine residue, eS7A, susceptible to Not4-mediated monoubiquitination, eS7A-3KR-K72, eS7A-3KR-K76, eS7A-3KR-K83 and eS7A-3KR-K84 24 . Hac1 protein levels were significantly lower in eS7A-3KR-K72 and eS7A-3KR-K76 single-lysine and eS7A-4KR mutant cells ( Supplementary Fig. 4b,d), but not in eS7A-3KR-K83 and eS7A-3KR-K84 single-lysine cells ( Supplementary Fig. 4c,e). These results, which indicate that monoubiquitination of eS7A at lysine 83 or 84 is sufficient for the production of Hac1 protein, are consistent with a model in which translation of HAC1 i mRNA is facilitated by monoubiquitinated eS7A at lysine residue 83 or 84. However, the possibility that mutation of eS7A could cause structural changes to the ribosome in addition to the ubiquitination defect cannot be excluded.
After Tm treatment, the translation of HNT1 mRNA is suppressed in an uORF-dependent manner 37,39 . We also found that HNT1 was most markedly repressed upon UPR (Fig. 1e). Hac1 is reported to be required to synthesize uORF-containing HNT1 mRNAs, making it essential for the downregulation of HNT1. Interestingly,  www.nature.com/scientificreports/ assessment of ribosome occupancy on HNT1 mRNA (Fig. 3a) showed that the ribosomes efficiently read through the uORF in the eS7A-4KR mutant after Tm treatment. The reduction of Hnt1 protein after Tm treatment was significantly restored in the eS7A-4KR mutant (Fig. 3b), but the level of long transcripts containing uORFs was not reduced in the mutant (Fig. 3c). These findings indicate that, although the level of Hac1 was lower in eS7A-4KR mutant than in es7A-WT cells, it was sufficient in the former to induce uORF-containing HNT1 mRNA under UPR conditions. Thus, independently of Hac1, eS7 ubiquitination may facilitate the translation of uORFs, thereby repressing the translation of HNT1 ORF (Fig. 3d).  www.nature.com/scientificreports/ Deubiquitinating enzyme complex Upb3-Bre5 is involved in the regulation of eS7A ubiquitination during UPR. Our results demonstrated that eS7A ubiquitination is required for translational controls during UPR. We therefore assessed whether the upregulation of eS7A ubiquitination is dependent on Ire1 and Hac1. We found that eS7A ubiquitination was increased in ire1∆ and hac1∆ mutant cells 4 h after Tm addition (Fig. 4a), indicating that the Ire1-Hac1 pathway is not required for the upregulation of eS7A ubiquitination.
To further elucidate the mechanism underlying the regulation of eS7A ubiquitination, we measured the expression of the E3 ligase Not4, which is responsible for the monoubiquitination of eS7A. We found that Not4 expression remained unchanged after the addition of Tm (Fig. 4b). We next hypothesized that the increase in monoubiquitinated eS7A in response to Tm is caused by a decrease in deubiquitinating activity. To test this possibility, we first performed a genetic screen to identify the enzyme responsible for deubiquitinating ubiquitinated eS7A. The levels of ubiquitinated eS7A were significantly and specifically increased in ubp3∆ mutant cells 2 h after the addition of Tm (Fig. 4c, lane 4). Deletion of UBP3 conferred sensitivity to Tm (Fig. 4d), suggesting that Ubp3 is the enzyme responsible for deubiquitinating ubiquitinated eS7A and that it contributes to resistance to ER stress. To measure the levels of ubiquitinated eS7A during UPR, we used N-terminal Myc-tagged ubiquitin, followed by ribosome affinity purification 22 . Tm addition significantly upregulated monoubiquitinated eS7A in a time-dependent manner (Fig. 4e). Western blot analysis of sucrose gradient fractions showed that monosomes or polysomes, but not free 40S subunits, contained ubiquitinated eS7A in both WT and ubp3∆ mutant cells (Fig. 4f), suggesting that translating ribosomes contain polyubiquitinated eS7A. In ubp3∆ mutant cells, the levels of mono-and di-ubiquitinated eS7A were increased in mono-and polysome fractions, but not in the 40S subunit (Fig. 4f), indicating that eS7A is the substrate of Ubp3 in monosomes and polysomes but not in the 40S subunit. Assessments of the levels of expression of HAC1 mRNA and Hac1 protein in the ubp3∆ deletion mutant showed that Hac1 protein expression was significantly decreased (Fig. 4g,h).
Next, we assessed whether the UPR results in the downregulation of the deubiquitinating enzyme. Ubp3 forms a complex with its cofactor Bre5 in vivo, with complex formation required for Ubp3 function 40,41 . Western blot analysis of Ubp3-3HA and Bre5-3HA expressed from endogenous promoters showed that the levels of the deubiquitinating enzyme Ubp3 and its cofactor Bre5 decreased significantly and gradually during the UPR (Fig. 4i). The downregulation of Ubp3 was impaired in ire1∆ but not in hac1∆ mutant cells, although neither of www.nature.com/scientificreports/ these UPR factors was required for the downregulation of Bre5 (Fig. 4j). These results indicate that activated Ire1 induces the downregulation of Ubp3, but not of Bre5, during the UPR. Taken together, these findings indicate that downregulation of deubiquitination in response to UPR increases the levels of monoubiquitinated eS7A, a downregulation that is affected by, but not completely dependent on, the Ire1-Hac1 pathway (Fig. 5).

Discussion
Ubiquitination of ribosomal proteins plays an essential role in quality control induced by ribosomal stalling [22][23][24]26,28,37 . However, the physiological functions of ribosome modifications remain unclear. The findings of this study indicate that ribosome eS7 monoubiquitination is required for translational controls during ER stress responses in yeast. Ribosome profiling revealed that eS7A monoubiquitination was necessary for translational up-and down-regulation of specific mRNAs. Monoubiquitination of eS7A facilitated the translational upregulation of HAC1 i mRNA, a master transcription factor in the UPR, and the downregulation of HNT1 during ER stress. The mechanisms underlying the roles of eS7A ubiquitination in translational regulation of specific mRNAs during ER stress, however, remain unclear. Hac1-mediated production of long transcripts containing uORFs was shown to repress the translation of histidine triad nucleotide-binding 1 (HNT1) mRNA 37 . We recently reported that uORF3 is required for HNT1 expression, and that translation of HNT1 is efficiently repressed by a strong Kozak sequence uORF3 during UPR 39 . These findings suggest that initiation of translation at the AUG codon of uORF3 is inefficient, and that leaky scanning of uORF3 is responsible for translation of HNT1. Although Tm treatment reduced the production of Hnt1 protein in eS7A-WT, it did not alter Hnt1 protein production in the eS7A-4KR mutant (Fig. 3b), despite the level of long transcripts containing uORFs not being changed in this mutant (Fig. 3c). These findings suggested that initiation of translation at the AUG codon of uORF3 is repressed in the eS7A-4KR mutant, and that initiation of translation at the AUG codon of the HNT1 ORF is stimulated by the increase in leaky scanning of uORF3. Thus, eS7 ubiquitination may facilitate the translation of uORF3, thereby repressing the translation of HNT1 ORF. Understanding translational regulation in response to the accumulation of unfolded proteins in the ER can be improved by determining the molecular mechanisms underlying eS7A monoubiquitination-mediated regulation of translation during the UPR. Our results suggested that initiation of translation of specific mRNAs, including HAC1 i and uORF3 of HNT1 mRNA, depends on eS7 ubiquitination by an as yet unknown mechanism, and that reduced translation from these initiation codons resulted in defects in the upregulation of HAC1 i mRNA and the downregulation of HNT1 mRNA upon UPR (Fig. 5). These interactions between eS7A and translation initiation factors may be critical for initiating translation at specific sites. eIF3 binding to ribosomes elongating and terminating on short upstream ORFs has been shown to promote the re-initiation of GCN4 translation [42][43][44][45] . Moreover, eIF3-dependent translation initiation mechanism contributes to translational recovery in chronic ER stress response 7 .The recently resolved cryo-EM structure of eIF3 in the context of the human 43S pre-initiation complex 44,46,47 , and the proximity of eIF3 to eS7A in a yeast 48S pre-initiation complex model suggest that eS7A is associated with specific initiation factors 47 (Supplementary Fig. 5). Modification of eS7A, including ubiquitination, may affect the interaction of the 40S subunit with translation initiation factors in the pre-initiation complex, modulating the initiation of translation of specific mRNAs.
The level of monoubiquitinated eS7A in response to UPR was also upregulated in ire1∆ and hac1∆ mutant cells, indicating that the Ire-Hac1 pathway is not necessary for the regulation of eS7A ubiquitination. These results strongly suggest that the deubiquitinating enzyme complex Upb3-Bre5 was involved in the regulation of eS7A ubiquitination. The level of monoubiquitinated eS7A was upregulated in ubp3∆ mutant cells. Ubp3 is the enzyme that deubiquitinates ubiquitinated eS7A and contributes to cell resistance to ER stress. Bre5 is a regulatory subunit that is downregulated upon UPR, even in ire1∆ and hac1∆ mutant cells, indicating that the downregulation of the deubiquitinating enzyme complex Ubp3-Bre5 is independent of the Ire1-Hac1 pathway. These findings suggest that ER stress reduced the levels of the deubiquitinating enzyme complex Ubp3-Bre5, leading to an increase in the ubiquitinated form of eS7A. Monoubiquitinated eS7A facilitates the translation of HAC1 i mRNA, resulting in efficient induction of Hac1-target genes and downregulation of HNT1 by as yet unknown mechanisms. Further investigations are needed to determine the mechanism underlying the ribosome ubiquitination-mediated regulation of translation initiation upon UPR.

Materials and methods
Yeast strains and genetic methods. The S. cerevisiae strains used in this study are listed in Table 1.
Plasmid constructs. Specific DNA sequences were PCR amplified using gene specific primers and cloned into vectors using PrimeSTAR HS DNA polymerase (#R010A, Takara-bio) and T4 DNA ligase (#M0202S, NEB). The sequences of all cloned DNAs amplified by PCR were verified by sequencing. Plasmids and primers used in this study are listed in Tables 2 and 3, respectively.
Yeast culture and media. All yeast cells were cultured at 30 °C in YPD or synthetic complete (SC) medium containing 2% glucose and harvested during log phase by centrifugation. To induce ER stress, yeast cells were grown at 30 °C until their absorbance at 600 nm was 0.2, treated with 1 µg/mL tunicamycin (Tm, #208-08243,  www.nature.com/scientificreports/ again mixed by vortexing for 10 s and chilled on ice for 5 min. After centrifugation at 16,000 × g for 5 min at room temperature, 190 µL of each water layer was transferred to a new 1.5 mL RNase free tube. A 250 µL aliquot of water-saturated phenol was added and the procedure was repeated. After centrifugation, 170 µL of each water layer was transferred to a new 1.5 mL RNase free tube; 200 µL of water-saturated phenol/chloroform (1:1) was added; and the tubes were vortexed for 10 s and centrifuged at 16,000 × g for 5 min at room temperature. A 150 µL aliquot of each water layer was transferred to a new 1.5 mL RNase free tube, to which was added 200 µL of water-saturated phenol/chloroform/isoamylalchol (25:24:1), followed by vortexing for 10 s and centrifugation at 16,000 × g for 5 min at room temperature. Finally, 130 µL of each water layer was transferred to a new 1.5 mL RNase free tube and subjected to ethanol precipitation. Each RNA pellet was dissolved in 20-30 µL of DEPCtreated water.
RNA electrophoresis and northern blotting. RNA electrophoresis and northern blotting were performed as described 24 with the following modifications. Trichloroacetic acid (TCA) precipitation for protein preparation. Yeast cell pellets in 1.5 mL tubes were resuspended in 500 µL ice-cold TCA buffer (20 mM Tris-HCl pH 8.0, 50 mM NH 4 OAc, 2 mM EDTA, and 1 mM PMSF) and transferred to new 1.5 mL tubes containing 500 μL 20% TCA and 500 μL 0.5 mm Zirconia/Silica Beads (BioSpec). The mixtures were vortexed three times for 30 s each, and the supernatants were transferred to new 1.5 mL tubes. A 500 µL aliquot of ice-cold TCA buffer was added to each tube, followed by vortexing for Table 3. Primers used in this study.

Name Description Sequence Use
OIT 4182 NOT4-F2  5′-CTG ATT TAC TAA ATC AAC TAA TCA ACG GAA GGA AAA TTA TCG CCG GTA ATC GGA TCC  CCG GGT TAA TTAA-3′  Genomic tagging of NOT4 C-terminus   OIT 4183  NOT4-S2  5′-AAT AGA TAA AAT TAT GGT TAA TGC AAA CAA GAA AAA TAT TTA GAG TCG GAA TCG ATG  AAT TCG AGC TCG-3′  Genomic tagging of NOT4 C-terminus   OIT 3451  UBP3-F2  5′-AAG CTT CTG ATT CGA GGA CTG CCT ATA TTT TAA TGT ATC AAA AGA GAA ATC GGA TCC  CCG GGT TAA TTAA-3′  Genomic tagging of UBP3 C-terminus   OIT 3452  UBP3-S2  5′-TAT TGC TAT ATT ATT TTT TAT GTA TTT TGT CTA TAA TAC CAC CCC CCG TCA TCG ATG AAT  TCG AGC TCG-  Ribosome purification to observe ribosome ubiquitination. To assess ribosome ubiquitination during ER stress, ribosomes were purified with Myc-tagged ubiquitin (Myc-Ubi) and the FLAG-tagged ribosomal protein uL23 (uL23-FLAG), as described previously 22 . Yeast cells harbouring pCUP1p-MYC-UBI and pRPL25(uL23)-FLAG were cultured in 800 mL of synthetic complete medium. To induce the expression of Myc-Ubi, the cells were cultured in the presence of 0.1 mM Cu 2+ for 1 h, followed by the addition of Tm to a concentration of 1 µg/mL and harvesting at the indicated time points. Cell lysates were prepared, and FLAG-tagged ribosomes were purified using M2 FLAG-affinity resin (Sigma), as described previously 22 . Affinity purified samples were subjected to SDS-PAGE followed by staining with Coomassie brilliant blue (CBB) or western blotting with an anti-Myc antibody.
Protein electrophoresis and western blotting. Protein electrophoresis and western blotting were performed as described previously 24 with the following modifications. Protein samples were separated by SDS-PAGE or Nu-PAGE, and stained with CBB or transferred to PVDF membranes (Immobilon-P, Millipore). After blocking with 5% skim milk in PBST (10 mM Na 2 HPO 4 /NaH 2 PO 4 pH 7.5, 0.9% NaCl, and 0.1% Tween-20), the membranes were incubated with primary antibodies (Table 4) for 1 h at room temperature followed by three washes with PBST and further incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 h at room temperature. If detecting HA-tagged proteins, the membranes were incubated with HRP-conjugated antibodies. After three washes with PBST, chemiluminescence was detected by LAS4000 (GE Healthcare).
Sucrose density gradient (SDG) centrifugation. SDG was performed as described 22 with the following modifications. Yeast cells were grown exponentially at 30 °C and treated with 0.1 mg/mL cycloheximide for 5 min before harvesting by centrifugation. The cell pellets were frozen and ground in liquid nitrogen using a mortar and pestle. The cell powder was resuspended in lysis buffer ( Ribosome profiling and RNA-seq. To induce ER stress, yeast cells were grown at 30 °C until reaching an optical density at 600 nm of 0.2. The cells were treated with 1 µg/mL Tm for ~ 4 h, harvested by vacuum filtration, and lysed by cryogenic grinding in a mixer mill (Retsch MM400). Whole cell lysates containing 10 µg of total RNA were each treated with 12.5 units of RNase I (Epicentre) at 23 °C for 45 min, and the ribosome fraction was sedimented through a 1 M sucrose cushion. The ribosome protected mRNA fragments were extracted with TRIzole regent (Life Technologies) and used for library preparation. Library preparation was performed as described 50 with the following modifications. For ribosome profiling analysis, whole cell lysates containing 20 μg of total RNA were each treated with ten units of RNase I (Epicentre) at 24 °C for 45 min. Linker DNA consisted of 5′-(Phos)NNNNNIIIIITGA TCG GAA GAG CAC ACG TCT GAA www.nature.com/scientificreports/ (ddC)-3′, with (Phos) indicating 5′ phosphorylation; (ddC) indicating a terminal 2′, 3′-dideoxycytidine; and Ns and Is indicate random barcodes for eliminating PCR duplication and multiplexing barcodes, respectively. The linkers were pre-adenylated with a 5′ DNA Adenylation kit (NEB), and then used for the ligation reaction. Un-reacted linkers were digested with 5′ deadenylase (NEB) and RecJ exonuclease (epicentre) at 30 °C for 45 min. RNA was reverse transcribed using the oligonucleotide primer, 5′- (Phos)NNAGA TCG GAA GAG  CGT CGT GTA GGG AAA GAG (iSp18)GTG ACT GGA GTT CAG ACG TGT GCT C-3′. PCR was performed with  the primers, 5′-AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT CTT TCC CTA CAC GAC GCT C-3′ and 5′-CAA  GCA GAA GAC GGC ATA CGA GAT JJJJJJGTG ACT GGA GTT CAG ACG TGTG-3′, where Js indicate reverse complement of the index sequence determined during Illumina sequencing. For RNA-seq analysis, total RNA was extracted from lysate using Trizol reagent (Life Technologies); rRNAs were removed from the total RNA using the Ribo-Zero Gold rRNA Removal Kit (Yeast) (Illumina); and the cDNA libraries were prepared using a TruSeq Stranded mRNA Library Prep Kit (Illumina). The libraries were sequenced on a HiSeq 2000/4000 (Illumina). The reads were mapped to yeast transcriptome, removing duplicated reads based on random barcode sequences. The analyses for ribosome profiling were restricted to read lengths of 30-33 nt for eS7-WT (0 h after Tm treatment), 29-33 nt for eS7-WT (4 h after Tm treatment), and 28-32 nt for eS7A-4KR datasets. The position of the A-site from the 5′-end of the reads was estimated based on the length of each footprint. The offsets using for analysis of ribosome profiling were 17 for 32-33 nt, 16 for 29-31 nt and 15 for 28 nt reads. For analysis of RNA-seq, an offset 15 was used for all mRNA fragments. The DESeq package was used to calculate the fold change of mRNA expression and TE 51 .

Data availability
The sequencing data for ribosome profiling experiments have been deposited in NCBI's Gene Expression Omnibus (GEO) and is accessible through GEO series accession number GSE128578. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.