Dot6/Tod6 degradation fine-tunes the repression of ribosome biogenesis under nutrient-limited conditions

Summary Ribosome biogenesis (Ribi) is a complex and energy-consuming process, and should therefore be repressed under nutrient-limited conditions to minimize unnecessary cellular energy consumption. In yeast, the transcriptional repressors Dot6 and Tod6 are phosphorylated and inactivated by the TORC1 pathway under nutrient-rich conditions, but are activated and repress ∼200 Ribi genes under nutrient-limited conditions. However, we show that in the presence of rapamycin or under nitrogen starvation conditions, Dot6 and Tod6 were readily degraded by the proteasome in a SCFGrr1 and Tom1 ubiquitin ligase-dependent manner, respectively. Moreover, promiscuous accumulation of Dot6 and Tod6 excessively repressed Ribi gene expression as well as translation activity and caused a growth defect in the presence of rapamycin. Thus, we propose that degradation of Dot6 and Tod6 is a novel mechanism to ensure an appropriate level of Ribi gene expression and thereby fine-tune the repression of Ribi and translation activity for cell survival under nutrient-limited conditions.


INTRODUCTION
Ribosome biogenesis (Ribi) and protein translation are among the most complex processes in the cell. Given the large number of components of the ribosome (e.g., 78 ribosomal proteins [RPs] and four ribosomal RNAs [rRNAs] in budding yeast Saccharomyces cerevisiae), the processing and assembly of ribosomes as well as the production of other cofactors for translation regulation, tRNA biosynthesis, and purine/pyrimidine synthesis are coordinately regulated by a set of genes known as the Ribi regulon (Broach, 2012;Jorgensen et al., 2004;Wade et al., 2006;Warner, 1999). In addition to their inherent complexity, Ribi and protein translation are highly energy-consuming processes in the cell. Therefore, to minimize cellular energy use, they are repressed under nutrient-limited conditions by several mechanisms, including degradation of rRNA (Huang et al., 2015;Pestov and Shcherbik, 2012), ribosome-specific autophagy (ribophagy) (Kraft et al., 2008;Ossareh-Nazari et al., 2010), and repression of the Ribi regulon (Huber et al., 2011).
In yeast, the transcriptional repressors Dot6 and Tod6 negatively control expression of Ribi genes through the target of rapamycin complex 1 (TORC1) and cAMP-dependent protein kinase (PKA) pathways (Broach, 2012;Eltschinger and Loewith, 2016;Huber et al., 2011;Kunkel et al., 2019;Lippman and Broach, 2009;Loewith and Hall, 2011). Under nutrient-rich conditions, Dot6 and Tod6 are inactivated by TORC1-mediated and PKA-mediated phosphorylation, leading to Ribi gene expression. By contrast, under nutrient-limited conditions, TORC1 and PKA signaling is suppressed, leading to dephosphorylation of Dot6 and Tod6, which repress Ribi gene expression by recruiting the RPD3L histone deacetylase complex to Ribi gene promoters (Huber et al., 2011;Kunkel et al., 2019). More specifically, the expression profiles of Ribi genes demonstrated that their repression resulting from TORC1 inactivation is mediated predominantly by Tod6, whereas their repression resulting from PKA inactivation is mediated predominantly by Dot6 (Lippman and Broach, 2009). However, biochemical analysis revealed that simultaneous inhibition of Sch9 and PKA kinases in the cell resulted in pronounced dephosphorylation of both Dot6 and Tod6, suggesting that Sch9 and PKA may act in parallel on Dot6 and Tod6 (Huber et al., 2011). Nonetheless, Dot6 and Tod6 likely represent a convergence point for nutrient sensing mediated by TORC1 and PKA ( Figure 1A).

OPEN ACCESS
During analysis of TORC1 and PKA signaling, attention has focused primarily on phosphorylation-dependent regulation (Broach, 2012;Eltschinger and Loewith, 2016;Loewith and Hall, 2011). The contribution of the ubiquitin-proteasome system (UPS) to control of the TORC1 and PKA pathways is poorly understood. Protein ubiquitination is achieved by a sequence of reactions catalyzed by a ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzymes (E2s), and ubiquitin ligases (E3s) (Finley et al., 2012;Komander and Rape, 2012;Oh et al., 2018). E3s are responsible for specific recognition of substrates. The S. cerevisiae genome contains 1, 11, and 60-100 genes encoding E1s, E2s, and E3s, respectively (Finley et al., 2012). Identification of the specific substrates of each E3 ligase is imperative to establish the role of the UPS in the TORC1 and PKA pathways.
To this end, we searched for short-lived proteins that physically and genetically associate with ribosomes whose biogenesis and functions are major targets of TORC1 and PKA signaling. We found that endogenous Dot6 and Tod6 are rapidly degraded by the proteasome under nitrogen starvation conditions or when TORC1 is inhibited by rapamycin. This observation led us to investigate why Dot6 and Tod6 are degraded when they are required to repress Ribi gene expression. Based on our genetic and biochemical analyses, we propose that derepression of Ribi mediated by Dot6/Tod6 degradation is crucial for cell survival under nutrient-limited conditions.

Dot6 and Tod6 are degraded by the proteasome under nutrient-limited conditions
During our analysis of the half-lives of proteins involved in ribosome regulation, we found that endogenous and untagged Dot6 and Tod6 ( Figure S1A) were rapidly degraded with half-lives of 60 min when cells were shifted to nitrogen starvation conditions or treated with rapamycin, a potent inhibitor of TORC1 signaling ( Figure 1B, SD-N medium and Rapamycin). On the contrary, under nutrient-rich conditions and upon treatment with cycloheximide, Dot6 was relatively stable and Tod6 was relatively unstable ( Figure 1B, Cycloheximide). Degradation of Dot6 and Tod6 was dependent on the proteasome because they were stabilized when cells were treated with MG132, a potent proteasome inhibitor ( Figure 1B), or when Rpt6 (Cim3), one of six ATPase regulatory particles in the 19S proteasome (Ghislain et al., 1993), was mutated ( Figure S1B). In these experiments, the band corresponding to Dot6 was shifted downward during the treatment period ( Figures 1B and S1B). These faster migrating bands were most likely dephosphorylated species ( Figure 1C) (Huber et al., 2011) and were particularly evident when their degradation was inhibited by MG132 or the cim3-1 mutation ( Figures 1B and S1B), suggesting that dephosphorylated Dot6 species accumulated when their degradation was inhibited. The levels of Dot6 and Tod6 were restored when nitrogen-starved cells were refilled with nitrogen ( Figure 1D), suggesting they are dynamically regulated. Dot6/ Tod6 degradation did not depend on autophagy because it was unaffected by deletion of PEP4, a vacuole proteinase (Takeshige et al., 1992), and Atg8, a homolog of LC3 (Kirisako et al., 1999) (Figure 1E). These results demonstrate that Dot6 and Tod6 are degraded by the proteasome upon nitrogen starvation or when TORC1 activity is inhibited, conditions where they are required to repress Ribi.
Although TORC1 is primarily activated by nitrogen/amino acids, PKA kinases are primarily activated by glucose (Broach, 2012;Rodkaer and Faergeman, 2014). We therefore analyzed the stabilities of Dot6 Figure 1. Continued (B) Wild-type cells were grown to an OD 600 of 0.4-0.6 in SD medium at 30 C. Cells were treated with the proteasome inhibitor MG132 (100 mM) for 30 min before being shifted to SD-N (-Nitrogen) medium containing the same concentration of MG132 or being treated with 250 nM rapamycin. For cycloheximide chase experiments, cells were treated with MG132 for 30 min before cycloheximide was added to the medium. To allow the action of proteasome inhibitors, a pdr5D null mutation was included in this experiment (Emter et al., 2002). Cells were collected at the indicated time points and lysates were subjected to immunoblotting with anti-Dot6 and anti-Tod6 antibodies. Coomassie Brilliant Blue (CBB) staining of the membrane served as a loading control. Quantified results for Dot6 and Tod6 are plotted in the graph. Values are mean G SD (n = 3). (C) Total yeast lysate was treated with l-phosphatase and subjected to western blotting with an anti-Dot6 antibody. (D) Cells were grown to log phase and collected before (0 0 hr) and after they were shifted to SD-N (-Nitrogen) medium at the indicated time points (0, 1, and 2 h). Cells were returned to nitrogen-containing SD medium and further incubated for up to 5 h. Whole cell lysates were prepared and immunoblotted with anti-Dot6 and anti-Tod6 antibodies. Cdc48 served as a loading control. (E) Wild-type, pep4D, and atg8D cells were grown to log phase in SD medium at 30 C. Cells were collected before (0 0 min) and after they were shifted to SD-N (-Nitrogen) medium at the indicated time points (0, 40, 80, and 120 min). Whole cell lysates were prepared and immunoblotted with anti-Dot6 and anti-Tod6 antibodies. Cdc48 served as a loading control. (F) Cells were grown to log phase and collected before and after they were shifted to S (+Nitrogen, -Glucose) medium at the indicated time points, and analyzed as in (E). . Dot6 and Tod6 are degraded in a SCF Grr1 -dependent and Tom1-dependent manner, respectively (A) The indicated strains were grown at 25 C and shifted to 37 C for 2 h. Cells were then shifted to SD-N (-Nitrogen) medium and collected at the indicated time points. Total cell lysates were prepared and analyzed by western blotting with an anti-Dot6 antibody. Cdc48 served as a loading control. (B) Cells were grown as in (A) and 250 nM rapamycin was added before cells were collected at the indicated time points. Total cell lysates were prepared and analyzed by western blotting with anti-Dot6 and anti-Tod6 antibodies. Cdc48 served as a loading control. (C) The stabilities of Dot6 and Tod6 were analyzed in wild-type and grr1D cells as in Figure 1B. Cells were shifted to SD-N (-Nitrogen) medium and collected at the indicated time points. Cdc48 served as a loading control. Quantified results for Dot6 are plotted in the graph, which shows the average of three independent experiments. Values are mean G SD (n = 3).
(D)The stability of Dot6 was analyzed in rapamycin-treated wild-type and grr1D cells as in Figure 1B. Cdc48 served as a loading control. (E) The stabilities of Dot6 and Dot6S/A (phosphorylation site mutant) in wild-type and grr1D cells were measured by cycloheximide chase experiments. The indicated cells were grown to log phase (OD 600 = 0.4-0.6) in SD medium at 30 C before cycloheximide was added at a final concentration of 100 mg/mL. Cells were collected at the indicated time points. Total cell lysates were subjected to western blotting with an anti-Dot6 antibody. CBB staining of the membrane served as a loading control. Quantified results are plotted in the graph, which shows the average of three independent experiments. Values are mean G SD (n = 3). iScience Article and Tod6 upon carbon starvation. Dot6 and Tod6 were relatively stable over time when cells were shifted to carbon starvation medium ( Figure 1F), suggesting they are specifically degraded upon TORC1 inhibition. When cells were shifted to carbon starvation medium, Dot6 was dephosphorylated at 0 h, but was phosphorylated again between 1 and 2 h before reverting to a dephosphorylated state between 4 and 6 h.
The mechanism underlying such oscillation under carbon starvation conditions is unknown. Our results suggest that Dot6 and Tod6 are specifically degraded upon nitrogen starvation and in the presence of rapamycin, conditions where the TORC1 pathway is inhibited.

Degradation of Dot6 and Tod6 depends on the SCF Grr1 and Tom1 ubiquitin ligases, respectively
To investigate the physiological importance of Dot6/Tod6 degradation, we searched for the E3(s) responsible. We first analyzed Dot6/Tod6 degradation in temperature-sensitive mutants of the SCF (Skp1, Cdc53/ Cul1, F box protein) complex, which constitutes a large family of ubiquitin ligases in yeast (Deshaies and Joazeiro, 2009;Harper and Tan, 2012;Willems et al., 2004). Degradation of Dot6, but not of Tod6, was significantly slowed in temperature-sensitive cdc53-1 and cdc34-2 cells (Figures 2A and 2B). We also investigated Dot6 degradation in the skp1 mutant temperature-sensitive strains. The skp1-11 mutant carries two amino acid substitutions (G160E and R167K) in its C-terminal domain (Bai et al., 1996). Both residues are highly conserved from yeast to humans. Another skp1 mutant, skp1-12, has a single amino acid substitution (L8G) in the N-terminal domain (Bai et al., 1996). Previous genetic analyses suggested that the skp1-11 mutation may weaken the interaction of Skp1 with Cdc4, an F box protein involved in cell cycle regulation, more than with Grr1. On the other hand, the skp1-12 mutation may more severely affect the interaction of Skp1 with Grr1 than with Cdc4 (Bai et al., 1996;Jaquenoud et al., 1998;Li and Johnston, 1997). Dot6 degradation was slowed in skp1-12 cells, whereas Dot6 was robustly degraded in skp1-11 cells ( Figures  2A, 2B, and S1C), implying that Grr1 regulates Dot6 stability. Indeed, we tested the stability of Dot6 in cells deleted for genes encoding F box proteins (Jonkers and Rep, 2009;Nakatsukasa et al., 2015b), and found that Dot6 degradation was slowed in grr1D cells under nitrogen starvation conditions or in the presence of rapamycin (Figures S1D and 2C and 2D).
We noted that dephosphorylated Dot6 species appeared during the treatment period in grr1D cells (Figure 2C), as observed when proteasome activity was inhibited ( Figures 1B and S1B). This observation strengthened the idea that Grr1 is responsible for Dot6 degradation. To investigate the connection between dephosphorylation of Dot6 and its degradation, we replaced serine residues that constitute phosphorylation sites (S247, S282, S313, S335, and S368) with alanine (Huber et al., 2011), and analyzed the stability of this protein by a cycloheximide chase experiment. Although wild-type Dot6 was relatively stable (lanes 1-4), Dot6S/A was readily degraded in wild-type cells (lanes 9-12) ( Figure 2E). When GRR1 was deleted, Dot6S/A degradation was blocked (lanes 13-16), suggesting that phosphorylation of Dot6 blocks its Grr1-dependent proteasomal degradation. Furthermore, Dot6 coprecipitated with 1xFLAG-tagged Grr1 ( Figure 2F) and was ubiquitinated in a Grr1-dependent manner ( Figure 2G). These results demonstrate that Dot6 is degraded in a SCF Grr1 -dependent manner under nitrogen starvation conditions or in the presence of rapamycin.
Tod6 degradation did not depend on the SCF complex; therefore, we further searched for the E3 responsible.
Of 50 ubiquitin ligase mutants tested, Tod6 degradation was reproducibly slowed in cells deleted for TOM1 ( Figure 2H), the HECT (homologous to E6-AP carboxyl terminus)-type ubiquitin ligase (Kim and Koepp, 2012;Nakatsukasa et al., 2018;Sung et al., 2016). To investigate the connection between dephosphorylation of Tod6 and its degradation, we replaced serine residues that constitute phosphorylation sites . Continued (F) Cells overexpressing 1xFLAG-tagged Grr1 and Dot6 under the control of the GAL1 and ADH promoters, respectively, were subjected to immunoprecipitation with an anti-FLAG antibody. Immunocomplexes were analyzed by immunoblotting with anti-FLAG and anti-Dot6 antibodies.
(G) cim3-1 and cim3-1grr1D cells expressing HA-tagged Dot6 and harboring plasmids encoding His-tagged Ub were treated with 250 nM rapamycin before being collected as described in Star Methods. Cells were disrupted with glass beads and His-Ub-modified proteins were enriched with Ni-NTA beads before the eluates were subjected to western blotting with an anti-HA antibody.
(H) The stability of Tod6 was analyzed in wild-type and tom1D cells as in Figure 1B. Cells were shifted to SD-N (-Nitrogen) medium and collected at the indicated time points. CBB staining of the membrane served as a loading control.
(I) The stability of Tod6 was analyzed in wild-type, tom1D, and tom1C3235A cells as in Figure 1B. Cells were shifted to SD-N (-Nitrogen) medium and collected at the indicated time points. Cdc48 and CBB staining of the membrane served as loading controls. Quantified results for Tod6 are plotted in the graph, which shows the average of three independent experiments. Values are mean G SD (n = 3).

OPEN ACCESS
iScience 25, 103986, March 18, 2022 5 iScience Article A C B Figure 3. Promiscuous accumulation of Dot6 and Tod6 represses protein translation in the presence of rapamycin (A) Wild-type cells (WT) and cells overexpressing Dot6 and Tod6 (o/e) were grown to log phase (OD 600 = 0.7-1.0) in YPD medium at 30 C. Where indicated (+Rap), cells were treated with 200 ng/mL (219 nM) rapamycin for 50 min before they were collected. Total RNA was extracted from cells and mRNA expression of the NOP4, NOC3, RLP24, FAF1, and NOP15 genes was quantified by real-time PCR using ACT1 mRNA as an internal control. mRNA expression of each gene in wild-type cells in the presence of rapamycin was set to 1.0. The relative levels of mRNA of each gene in wild-type cells in the absence of rapamycin, and in cells overexpressing Dot6 and Tod6 in the absence or presence of rapamycin are shown in the graph. Values are mean G SD (n = 6, triplicate experiments for two independent RNA preparations). p values more than 0.05 are shown as n.s., whereas p values less than 0.05, 0.01, and 0.001 are shown as *, **, and ***, respectively. (B) Polysome profiles of wild-type cells (WT) and cells overexpressing Dot6 and Tod6 (o/e) were analyzed. Cells were grown to log phase in the presence or absence of rapamycin as in (A). Whole cell lysates were prepared and separated on 10-50% sucrose gradients. The gradients were collected from the top. iScience Article (S280, S298, S308, S318, S333, and S346) with alanine (Huber et al., 2011) and analyzed the stability of this protein by a cycloheximide chase experiment. Although wild-type Tod6 was relatively unstable (lanes 1-4) as observed in Figure 1B, Tod6S/A was degraded much faster in wild-type cells (lanes 9-12) ( Figure S2), suggesting that phosphorylation of Tod6 blocks its proteasomal degradation. This result is consistent with the previous suggestion that dephosphorylation of Tod6 may destabilize this protein (Huber et al., 2011). However, in contrast with our assumption that Tod6 is dephosphorylated during starvation and targeted for Tom1-dependent degradation, we failed to detect Tom1-dependent degradation of Tod6S/A. Although the reason for this discrepancy is currently unknown, one possible explanation is that Tod6S/A might be degraded through a Tom1-independent pathway, whereas physiologically dephosphorylated Tod6 can be degraded in a Tom1-dependent manner. Another problem is our failure to obtain convincing evidence for Tod6 ubiquitination (data not shown). A future study is warranted to clarify fully the regulatory mechanisms of Tod6 degradation. Nonetheless, we found that Tod6 degradation was slowed in tom1C3235A mutant cells, in which the catalytic cysteine of Tom1 is replaced by alanine (Kim and Koepp, 2012), under nitrogen starvation conditions, whereas Dot6 was robustly degraded in Tom1 mutant strains ( Figure 2I). These results suggest that Tom1 contributes to Tod6 degradation under nitrogen starvation conditions.

Promiscuous accumulation of Dot6 and Tod6 represses protein translation in the presence of rapamycin
Next, we examined the physiological significance of Dot6/Tod6 degradation when TORC1 is inhibited by rapamycin. To analyze the phenotype of cells specifically defective in degradation of Dot6 and Tod6, we first attempted to utilize mutant forms of Dot6 and Tod6 in which all lysine residues were substituted with arginine (designated Dot6KR and Tod6KR, respectively). These mutants were not expected to be degraded by SCF Grr1 and Tom1. Surprisingly, however, Dot6KR and Tod6KR were robustly degraded in a Grr1-dependent and Tom1-dependent manner, respectively ( Figure S3). Although some other ubiquitin ligases are known to mediate ubiquitination and proteasomal degradation of lysine-lacking substrates (McClellan et al., 2019;Weber et al., 2016), the mechanisms by which SCF Grr1 and Tom1 mediate degradation of Dot6KR and Tod6KR, respectively, are unknown. These results suggest that Dot6KR and Tod6KR cannot be used to analyze the phenotype of cells specifically defective in Dot6/Tod6 degradation.
As the next best approach to analyze the phenotype of cells defective in Dot6/Tod6 degradation, we constructed cells that overexpressed Dot6 and Tod6 under the control of the constitutive ADH1 promoter from the genomic URA3 and TRP1 loci, respectively ( Figure S4A). We then measured the mRNA levels of specific Ribi genes (NOP4, NOC3, RLP24, FAF1, and NOP15) whose promoters contain an RNA polymerase A (I) and C (III) (PAC) motif (Huber et al., 2011;Wade et al., 2006). In the absence of rapamycin, expression of NOP4 and NOC3 in cells overexpressing Dot6 and Tod6 was nearly identical to that in wild-type cells (Figure 3A). Expression of RLP24, FAF1, and NOP15 was also robustly maintained in cells overexpressing Dot6 and Tod6 in comparison with wild-type cells, although their expression decreased 50-60%. By contrast, when cells were treated with rapamycin, expression of all these genes was strongly repressed (60-fold to 150-fold). Importantly, Dot6 and Tod6 overexpression further decreased their expression by approximately 10-fold in comparison with wild-type cells. These results demonstrate that promiscuous accumulation of Dot6 and Tod6 in the presence of rapamycin excessively represses expression of Ribi genes.
To analyze the effects of promiscuous accumulation of Dot6 and Tod6 on the translational status, we performed polysome analysis. In the absence of rapamycin, overexpression of Dot6 and Tod6 slightly suppressed polysome formation, which was quantified by measuring monosome versus polysome complexes (see bar graph) ( Figure 3B). This was probably because of the appearance of a small amount of dephosphorylated and activated Dot6 and Tod6 caused by their overexpression. In the presence of rapamycin, overexpression of Dot6 and Tod6 clearly suppressed polysome formation, suggesting that promiscuous accumulation of Dot6 and Tod6 excessively repressed active translation. The mechanism by which excess (E) The growth rates of wild-type cells, cells overexpressing Dot6 and Tod6 (Dot6/Tod6 o/e), and cells deleted for DOT6 and TOD6 (dot6Dtod6D) were analyzed in YPD liquid medium. In the left panel, cells were grown to log phase in YPD medium and directly diluted into YPD medium before their growth was analyzed. In the right panel, cells in log phase were transferred to SD-N (-Nitrogen) medium and incubated for 96 h. Subsequently, cells were collected and diluted into YPD medium before their growth was analyzed. See STAR Methods for details. (F) The indicated strains were grown to an OD 600 of 0.4-0.5 in YPD medium at 30 C. Cultures were diluted in sterile water and spotted onto YPD medium supplemented with 25 nM rapamycin. The plate was incubated at 30 C for the indicated numbers of days. iScience Article repression of Ribi genes affects polysome formation is currently unknown. One hypothesis is that each Ribi gene is repressed differently and an imbalance of Ribi regulon levels results in production of aberrant ribosomes containing stoichiometrically imbalanced RPs and/or rRNAs that cannot effectively perform translation.
We next performed puromycin incorporation assays. Puromycin is a chain terminator that is incorporated into growing nascent polypeptide chains and thus can be used to label nascent polypeptides (Schmidt et al., 2009). Cells were treated with puromycin (200 mg/mL), equal amounts of whole cell lysates were separated by SDS-PAGE, and the translation efficiency was evaluated by detecting puromycilated nascent chains using an anti-puromycin antibody by western blotting ( Figure 3C). Under nutrient-rich conditions, the efficiency of puromycin incorporation was almost unchanged when DOT6 and TOD6 were deleted (dot6Dtod6D) (lanes 9-10) or overexpressed (Dot6/Tod6 o/e) (lanes 11-12) in comparison with wild-type cells (lanes 7-8). This is likely because the vast majority of Dot6 and Tod6 species are inactivated by TORC1/Sch9-mediated phosphorylation; therefore, high levels of Dot6 and Tod6 do not affect the translation efficiency in the presence of nutrients. Pretreatment of cells with cycloheximide blocked puromycin incorporation (lanes 1-6), confirming this method allows newly synthesized proteins to be analyzed. Overall incorporation of puromycin was reduced in cells treated with rapamycin (lanes 13-18). This is likely because Ribi and translation initiation were inhibited by inactivation of TORC1. Importantly, the efficiency of puromycin incorporation was substantially lower in cells overexpressing Dot6 and Tod6 (lanes 17-18) than in wild-type cells (lanes 13-14). The amount of protein in the whole cell lysate in each lane was almost identical, suggesting that the rate of protein translation was significantly lowered transiently. Intriguingly, the efficiency of puromycin incorporation was comparable in dot6Dtod6D and wild-type cells (lanes 15-16), although deletion of DOT6 and TOD6 was expected to derepress Ribi gene transcription and increase polyribosome association in comparison with wild-type cells. One possible reason for this finding is that it takes longer for an obvious phenotype of Ribi gene derepression to materialize (see discussion). Nevertheless, these results suggest that promiscuous accumulation of Dot6 and Tod6 leads to an excessive reduction of translation activity in cells treated with rapamycin. In other words, proteasome-mediated degradation of Dot6 and Tod6 derepresses Ribi gene expression, thereby ensuring an appropriate level of the Ribi regulon and translation activity.

Promiscuous accumulation of Dot6 and Tod6 causes severe growth defects in the presence of rapamycin
We next examined whether promiscuous accumulation of Dot6/Tod6 and subsequent reduction of translation affect cell growth in the presence of rapamycin. When grown in nutrient-rich medium in the absence of rapamycin, dot6Dtod6D cells and cells overexpressing Dot6 and Tod6 grew as efficiently as wild-type cells. However, cells overexpressing Dot6 and Tod6 grew more poorly than wild-type and dot6Dtod6D cells in the presence of rapamycin ( Figures 4A, S4B, and S4C). To confirm this result, we measured cellular growth in liquid medium. These three strains grew similarly in the absence of rapamycin ( Figures 4B-4D). However, cells overexpressing Dot6 and Tod6 grew more poorly than wild-type and dot6Dtod6D cells in the presence of rapamycin. Taken together, these results demonstrate that promiscuous accumulation of Dot6 and Tod6 causes severe growth defects specifically when TORC1 activity is inhibited.
Several previous studies suggested that Dot6 and Tod6 are important for adaption to environmental changes (e.g., high salt) (Ho et al., 2018;Lee et al., 2011). Therefore, we reasoned that fine-tuning ribosome levels under starvation conditions to maintain the energy balance is important for recovery of cellular growth after replenishment of nutrients. We therefore analyzed the growth of wild-type cells, dot6Dtod6D cells, and cells overexpressing Dot6 and Tod6 under nutrient-rich conditions after they had been incubated under nitrogen starvation conditions for a prolonged period of time. When log-phase cells in nutrient-rich medium were directly diluted into nutrient-rich medium, they entered log phase again after a short ''lag'' phase (100 min) regardless of the absence or overexpression of Dot6 and Tod6 (left panel) ( Figure 4E). In marked contrast, when cells were diluted into nutrient-rich medium after they had been incubated in SD-N medium for 4 days, wild-type cells entered log phase after a lag phase of 400 min, whereas dot6Dtod6D iScience Article cells and cells overexpressing Dot6 and Tod6 entered log phase after a lag phase of 500 and 600 min, respectively (right panel). These results suggest that fine-tuning of ribosome levels mediated by appropriate levels of Dot6 and Tod6 during prolonged starvation is crucial for regrowth of cells after replenishment of nutrients.
These results were based on experiments using cells artificially, but mildly, overexpressing Dot6 and Tod6 at higher than physiological levels. Cells lacking GRR1 or TOM1 were sensitive to rapamycin and these phenotypes were partially rescued by additional deletion of DOT6 or TOD6, respectively ( Figure 4F). This result strongly suggests that the rapamycin sensitivity of these cells is at least partly because of accumulation of Dot6 and Tod6. In other words, Dot6 and Tod6 should be kept at optimal levels by SCF Grr1 -mediated and Tom1-mediated proteolysis, respectively, when TORC1 signaling is inactivated.

DISCUSSION
Cells exposed to nutrient-limited conditions must adjust their metabolism to minimize their energy use.
Ribi and protein synthesis are major energy-consuming processes in the cell and therefore must be carefully controlled. In this study, we discovered that Dot6 and Tod6 are rapidly degraded by the proteasome in a SCF Grr1 and Tom1 ubiquitin ligase-dependent manner, respectively, in the presence of rapamycin or under nitrogen starvation conditions. This observation led us to investigate why Dot6 and Tod6 are degraded when they are required to repress Ribi gene expression under nutrient-limited conditions. Based on our biochemical and genetic analyses, we propose that regulated degradation of Dot6 and Tod6 ensures an appropriate level of Ribi gene expression, thereby fine-tuning the repression of Ribi and translation activity for cell survival under starvation conditions ( Figure 4G).
A previous study reported that neither the dot6Dtod6D double mutant nor either single mutant exhibits any apparent change in sensitivity to rapamycin treatment at different concentrations or for different durations, although Dot6 and/or Tod6 are required to repress Ribi gene expression (Lippman and Broach, 2009). Indeed, we did not observe any growth defect of dot6Dtod6D cells in the presence of rapamycin ( Figures 4A-4D and S4B). However, instead of such short-term exposure to rapamycin (several hours to a day), the Broach group performed a chemostat-based competition experiment, which is a very sensitive method to detect a difference in fitness between two strains, under nutrient-limited conditions. They found that Dot6 and Tod6 are necessary for successful adaptation to nitrogen starvation conditions. Moreover, the viability of the dot6Dtod6D strain is modestly, but reproducibly, decreased upon carbon or nitrogen starvation for 30 days. They also found that deletion of DOT6 and TOD6 induces petite formation, which is caused by mutation of mtDNA, upon nitrogen starvation, but not upon carbon starvation for 5 days (Lippman and Broach, 2009). Likewise, derepression of Ribi by deletion of DOT6 and TOD6 genes may elicit adverse effects specifically after a long period of starvation, whereas the phenotype induced by excessive repression of Ribi upon overexpression of Dot6 and Tod6 may materialize faster.
Previous studies suggested that accumulation of misfolded proteins in the ER not only induces the unfolded protein response (UPR) but also leads to deactivation of PKA in yeast (Chantranupong and Sabatini, 2016;Pincus et al., 2014;Walter and Ron, 2011). Although the mechanism by which ER stress affects PKA activity remains to be elucidated, the interplay between the UPR and PKA signaling is considered to be an important feature of the ER stress response in budding yeast. As reported previously, growth of dot6Dtod6D cells was modestly impaired in the presence of tunicamycin compared with wild-type cells (Pincus et al., 2014) ( Figure S4D). In marked contrast, cells overexpressing Dot6 and Tod6 were resistant to tunicamycin, which is consistent with the previous observation that ribosome deficiency protects against ER stress (Steffen et al., 2012). Thus, global reduction of protein synthesis via repression of Ribi is critical to protect cells against ER stress even when the repression is excessive. On the contrary, excessive repression of Ribi is detrimental to cells under nutrient-limited conditions, possibly because a minimum level of protein synthesis is required for cell survival.
Although our data strongly suggest the importance of Ribi gene derepression mediated by Dot6/Tod6 degradation under nutrient-limited conditions, it can be argued that its effect might be subtle and overwhelmed by other ribosome repression systems such as rRNA degradation, ribophagy, and intron-mediated repression of RP genes (Huber et al., 2011;Kraft et al., 2008;Kunkel et al., 2019;Parenteau et al., 2019;Pestov and Shcherbik, 2012). However, we speculate that these repression and derepression systems work together and fine-tune Ribi to control not only the quantity of ribosomes but also the quality of each iScience Article ribosome in the cell. In contrast with the historical view of the ribosome, which was considered to be a passive and indiscriminate machine, recent quantitative mass spectrometry measurements revealed the composition of the ribosome changes at the RNA and protein levels (Shi et al., 2017;Slavov et al., 2015). Moreover, individual ribosomal components likely play specific roles in translation of specific mRNAs (Genuth and Barna, 2018; Gingras et al., 2007). Such ribosome heterogeneity is thought to be important for specialized translation of individual transcripts, and regulation and expression of key gene regulatory networks. In mammals, heterogeneity and specialization of ribosomes is thought to be important for proteome regulation during multicellular organismal development. Even if Ribi gene derepression mediated by Dot6/Tod6 degradation only subtly affects the quantity of ribosomes, it may result in different levels and/or combinations of Ribi proteins, and hence significant qualitative differences (i.e., heterogeneity), in ribosomal particles, leading to a different proteome in the cell. In higher eukaryotes, we suspect there is a similar system to fine-tune Ribi. Continued analysis of the yeast system will be important to further elucidate the significance of the balance between repression and derepression of Ribi, which may enable cells to adjust their metabolism according to fluctuations in nutrient levels.

Limitations of the study
To analyze the phenotype of cells specifically defective in Dot6/Tod6 degradation, we attempted to utilize Dot6KR and Tod6KR. However, in contrast to our expectation, these mutants were robustly degraded in a Grr1-dependent and Tom1-dependent manner, respectively. It will be important to investigate the mechanisms of their degradation to further analyze the phenotype of cells specifically defective in Dot6/Tod6 degradation.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Kunio Nakatsukasa (nakatsukasa@nsc.nagoya-cu.ac.jp).

Materials availability
All unique/stable reagents generated in this study are available from the lead contact.

Data and code availability
Any information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND SUBJECT DETAILS Strains
The yeast strains used in this study are listed in the key resources table. All gene deletion and tagged strains were constructed using standard homologous recombination methods (Knop et al., 1999). The tod6D:: KanR fragment was amplified by PCR using the tod6D::KanR strain (Open Biosystems: Thermo Scientific) as a template and primers OKN1043 and OKN1044. The dot6D::natR fragment was amplified by PCR using pFA6a-natMX6 (Euroscarf: P30437) as a template and primers OKN980 and OKN981. The grr1D::CgHIS and grr1D::CgTRP fragments were amplified by PCR using p1804 and p1805 (National BioResource Project) as templates, respectively, and primers OKN1969 and OKN1970 or OYS15 and OYS16, respectively. The tom1D::CgHIS and tom1D::CgTRP fragments were amplified by PCR using p1804 and p1805 as templates, respectively, and primers OYS48 and OYS49. Yeast strains expressing HA-Dot6 were constructed by transforming cells with pYS8 that had been linearized with NcoI. KKY91 was constructed by transforming TKY3380 with pKK3 and pKK4 that had been linearized with HindIII and NcoI, respectively. iScience Article was confirmed by western blotting with anti-Dot6 and anti-Tod6 antibodies. KKY111 was constructed by transforming TKY3227 with pYS18 that had been linearized with NheI. KKY112 was constructed by transforming TKY3173 with pYS17 that had been linearized with NcoI. KKY139 and KKY140 were constructed by transforming TKY3227 and SUY58 with pKN476 that had been linearized with NcoI, respectively. KKY118 and KKY119 were constructed by transforming TKY3173 and SUY119 with pKN477 that had been linearized with NcoI, respectively. In Figure S1D, yeast strains deleted for genes encoding F-box proteins were derived from the Yeast Knockout (Mata) collection (Open Biosystems: Thermo Scientific).

Plasmids
The plasmids and oligonucleotide primers used in this study are listed in the key resources table. The individual plasmids were constructed as follows. pYS3 The DNA fragment encoding P DOT6 -DOT6-T DOT6 was amplified by PCR with primers OKN1413 and OYS23. The resultant fragment was excised with SacI/XhoI and inserted into the same sites of pRS306 to generate pYS3.

pYS7
The DNA fragment encoding P TOD6 -TOD6 was amplified by PCR with primers OKN1404 and OKN1405. The resultant fragment was digested with SacI/XhoI and inserted into the same sites of pRS306 to generate pKN335. The DNA fragment encoding T TOD6 was amplified by PCR with primers OYS28 and OYS29. The resultant fragment was digested with XhoI/KpnI and inserted into the same sites of pKN335 to generate pYS7.

pYS8
The promoter region of the DOT6 gene was amplified from genomic DNA by PCR with primers OKN1413 and OYS25. The open reading frame and terminator region of the DOT6 gene were amplified from genomic DNA by PCR with primers OYS24 and OYS23. The two resultant fragments were mixed and used as a template for amplification of P DOT6 -HA-DOT6-T DOT6 by PCR with primers OKN1413 and OYS23. The resultant fragment was excised with SacI/XhoI and inserted into the same sites of pRS306.

pYS13
The DNA fragment encoding DOT6 was amplified by PCR with primers OYS10 and OKN1414, digested with XbaI/XhoI, and inserted into the same sites of p415GPD.

pYS14
The open reading frame of TOD6 was amplified by PCR with primers OKN1485 and OKN1405, digested with SpeI/XhoI, and inserted into the same sites of p416GPD.

pKN435
The DNA fragment encoding 1xFLAG-GGGS (linker)-GRR1 was amplified by PCR with primers OKN1875 and OKN1678. The resultant fragment was digested with NheI/XhoI and inserted into the SpeI/XhoI sites of pKN16.

pKN476
The DNA fragment encoding a ClaI-AflII region of the DOT6 gene containing phosphorylation site mutations (S247A, S282A, S313A, S335A, and S368A) (Huber et al., 2011) was synthesized by GENEWIZ (pKN398). This fragment was amplified by PCR using primers OKN1981 and OKN1982. The resultant fragment was digested with ClaI/AflII and inserted into the ClaI/AflII sites of pYS3 to create pKN476.

pKN477
The DNA fragment encoding a BglII-EcoRI region of the TOD6 gene containing phosphorylation site mutations (S280A, S298A, S308A, S318A, S333A, and S346A) (Huber et al., 2011)  The DNA fragment encoding DOT6 was amplified by PCR with primers OYS10 and OKN1414. The amplified fragment was digested with XbaI/XhoI and inserted into the same sites of p415ADH to generate pKK1. pKK1 was digested with SacI/KpnI. The resultant fragment encoding P ADH1 -DOT6-T CYC1 was inserted into the SacI/KpnI sites of pRS304 (Sikorski and Hieter, 1989) to generate pKK3.

pKK4
The DNA fragment encoding TOD6 was amplified by PCR with primers OKN1485 and OKN1405. The resultant fragment was digested with SpeI/XhoI and inserted into the same sites of p416ADH to generate pKK2. pKK2 was digested with SacI/KpnI. The resultant fragment encoding P ADH1 -TOD6-T CYC1 was inserted into the SacI/KpnI sites of pRS306 (Sikorski and Hieter, 1989) to generate pKK4.

pYS17
The HindIII-PartialP TOD6 -tod6KR-PartialT TOD6 -EcoRI gene was synthesized by GENEWIZ and inserted into the same sites of pYS7. The resultant plasmid was digested with SacI/KpnI. The SacI-P TOD6 -tod6KR-T TOD6 -KpnI fragment was inserted into the same sites of pRS306.

Antibodies
Polyclonal antisera to Dot6 and Tod6 were generated in a rabbit (Sigma-Aldrich) using recombinant proteins. Briefly, the open reading frames of DOT6 and TOD6 were amplified by PCR with primers OKN986/ OKN987 and OKN1031/OKN1032, respectively. The resultant fragments were digested with NheI/XhoI and inserted into the same sites of pRSET-b (Thermo Fisher Scientific) to generate pKN229 and pKN245, respectively. These two plasmids were transformed into BL21 (DE3). Expression of His-tagged Dot6 and Tod6 proteins was induced, and these proteins were purified with Ni-NTA agarose beads. Antibodies against Dot6 and Tod6 were purified from each antiserum using recombinant proteins immobilized to NHS-activated Sepharose beads. Other antibodies used in this study are listed in the key resources table.
The purified anti-Dot6 and anti-Tod6 antibodies were diluted in Can Get Signal Solution 1 (#NKB-201, TOYOBO) at 1/5000 (typically 3 mL was added to 15 mL of Can Get Signal Solution supplemented with 30 mM azide) and used several times. Although the anti-Dot6 antibody was highly specific, several weak cross-reacting bands overlapped with the band corresponding to Dot6 when a whole cell lysate was separated on a 6% SDS-PAGE gel. The cross-reacting bands no longer overlapped with the band corresponding to Dot6 when the lysate was separated on a 6% SDS-PAGE gel supplemented with 8 M urea. These bands also disappeared after the diluted antibodies had been used several times. Except where otherwise indicated, a 6% gel containing 8 M urea was routinely used for western blotting with the anti-Dot6 antibody.

Culture conditions
Yeast cells were grown in YP medium (YPD: 1% yeast extract, 1% peptone, and 100 mg/L adenine hydrochloride) supplemented with 2% glucose or 2% galactose, synthetic complete medium (SD medium: 0.67% yeast nitrogen base without amino acids, 100 mg/L adenine hydrochloride, all standard amino acids, and 2% glucose), or nitrogen starvation medium (SD-N medium: 0.17% yeast nitrogen base without amino acids and ammonium sulfate, 2% glucose, and 0.95% MES, pH 6.2 adjusted by KOH). Where indicated, galactose was used as a carbon source in SD medium instead of glucose (SGal medium). Appropriate amino acids were excluded from synthetic medium to maintain plasmids.

Co-immunoprecipitation
Cells overexpressing Dot6 under the control of the GPD promoter and cells overexpressing 1xFLAGtagged Grr1 under the control of the GAL1 promoter were inoculated into 50 mL of SD medium and grown at 30 C overnight. The next morning, 100 OD 600 equivalent of cells were collected, washed twice with 50 mL of water, and further grown in 250 mL of YPGalactose medium for 6 hr at 30 C to induce expression of ll iScience Article

Rapamycin treatment
The stabilities of proteins in cells after rapamycin administration were analyzed essentially the same as the cycloheximide chase assay, except that rapamycin (R-5000, LC Laboratories) was added at a final concentration of 250 nM instead of cycloheximide. The rapamycin stock solution in DMSO was stored at -30 C and used within 1 month. iScience Article Nitrogen starvation assay The stabilities of proteins in cells cultured in SD-N medium were analyzed as follows. Yeast cultures in log phase were prepared as described for the cycloheximide chase assay. After the initial sample (time=0) was collected, cells in the culture (20-30 mL) were collected and washed once with 50 mL of RO water and once with 50 mL of SD-N medium. Cells were then resuspended in an equal volume of SD-N (-Nitrogen) or S (+Nitrogen, -Glucose) medium, further incubated at 30 C, and collected at the indicated time points. Whole cell lysates were prepared and analyzed by western blotting as described above.
In vivo ubiquitination assay Cells (cim3-1) expressing HA-tagged Dot6 under the control of its own promoter from the URA3 locus were transformed with the plasmid pKN366 (2m, HIS3, P CUP1 -8xHis tagged Ub) and grown to an OD 600 of 0.4-0.5 in SD medium at 25 C. The medium was changed to YPD medium, cells were further grown at 25 C for 2 hr in the presence of 100 mM CuSO 4 , the temperature was shifted to 37 C, and cells were further grown for 2 hr. Rapamycin was added at a final concentration of 250 nM and cells were further incubated for 45 min, treated with 30 mM NaN 3 on ice, collected by centrifugation, and stored at -80 C until use. Cells were thawed in Wash buffer (4 M urea, 50 mM NaCl, 20 mM imidazole, and 40 mM HEPES-NaOH, pH 7.5) and disrupted with a multibead shocker (2700 rpm, nine cycles of 'on' for 60 sec and 'off' for 60 sec) (Yasui Kikai). Cell debris and unbroken cells were removed by centrifugation. Fifty microliters of the supernatant was precipitated with TCA and used as a whole cell lysate control. The remaining supernatant was added to 30 mL of Ni-NTA agarose beads (FUJIFILM Wako) and rotated at 4 C for more than 1 hr. The beads were washed four times with Wash buffer and bound proteins were eluted with 20 mL of Elution buffer (4 M urea, 50 mM NaCl, 300 mM imidazole, and 40 mM HEPES-NaOH, pH 7.5). The eluate was collected by centrifugation, added to SDS-PAGE sample buffer, and incubated at 88 C for 5 min.

Spot assay (serial dilution assay)
Yeast cells were grown in SD medium overnight at 30 C until OD 600 reached 0.7-0.8. Cells (0.75 OD 600 equivalent) were collected and suspended in 1.5 mL of sterile water. Ten-fold serial dilutions were generated as follows. Two hundred microliters of the cell suspension was transferred to the first lane of a 96-well plate (EVERGREEN). One hundred and eighty microliters of sterile water was placed in the second, third, and fourth lanes. Then, 20 mL of the cell suspension in the first lane was transferred to the second lane (lane 2) and mixed by pipetting ten times. Similarly, 20 mL of the cell suspension was transferred from the second lane to the third lane, and from the third lane to the fourth lane, as described above. An aliquot of the cell suspension (2.5 mL) was spotted on agar plates, which were then incubated at the indicated temperature.

Yeast cell growth assay
Cells were grown in YPD medium overnight at 30 C. The next morning, cells were diluted to an OD 600 of 0.1 in YPD medium and grown at 30 C for 8 hr. During this period, OD 600 was kept below 0.5 by diluting cells with YPD medium several times. Subsequently, cells were diluted 1:10,000 in 50 mL of YPD medium and grown overnight at 30 C until OD 600 reached 0.5. Cells were then diluted to an OD 600 of 0.1 in YPD medium before rapamycin was added to the culture at the indicated concentration (0, 5, 10, 15, 20, or 25 nM). One milliliter of the cell culture was transferred to a 24-well plate (Cell Culture Plate TR5002 24-well, TrueLine). Cell growth was analyzed by measuring OD 660 . Plate reader software (i-controlä, TECAN), which was setup to measure OD 660 for 50 hr at 10-12 min intervals, was used. The plate was agitated for 30-45 sec immediately before OD 660 was measured. For each well, OD 660 was measured at 4-9 spots and the mean value was used for analysis.
In Figure 4E, a saturated preculture was diluted 1/50,000 in YPD medium and grown overnight at 30 C until OD 600 reached 0.3-0.5. One portion of cells was diluted to an OD 600 of 0.1 in YPD medium, and their growth was analyzed using a plate reader as described above. Another portion of cells was washed once and diluted to an OD 600 of 0.2 in SD-N medium. After cells had been incubated at 30 C for 96 hr (4 days), they were diluted to an OD 600 of 0.1 in YPD medium and their growth was analyzed using a plate reader as described above.

Puromycin incorporation assay
Yeast cultures were grown in 10 mL of YPD medium at 26 C overnight. The next morning, the culture was diluted to an OD 600 of 0.1 in YPD, grown at 30 C until OD 600 reached 0.6, diluted 25,000-fold in 50 mL of YPD, and grown at 30 C overnight. The next day, for samples treated with rapamycin, the culture was diluted 10-fold in 50 mL of YPD several times before OD 600 reached 0.4. After 5-6 hr, the culture was adjusted to an OD 600 ll OPEN ACCESS