Acetic acid induces proteasome storage granule formation and inhibits proteasomal proteolysis: Comparison with other induction conditions

TL;DR

Acetic acid causes yeast cells to form proteasome storage granules more quickly than other conditions, offering new insights into how these granules form.

Contribution

The study identifies acetic acid as a new inducer of proteasome storage granules and reveals condition-specific mechanisms.

Findings

Acetic acid induces PSG formation faster than glucose depletion or mitochondrial stress.

Acetic acid inhibits proteasomal proteolysis, leading to ubiquitinated protein accumulation.

Key factors like Sem1/Dss1 and Rpn10/Rpn13 vary in importance depending on the induction condition.

Abstract

Proteasomes are primarily located within the nuclei of proliferating cells; however, their localization changes dynamically in response to environmental conditions. Saccharomyces cerevisiae forms proteasome storage granules (PSGs) in the cytoplasm upon glucose depletion, mitochondrial stress, or transition to quiescence. Although intracellular acidification drives PSG formation, the detailed mechanisms underlying this process are unclear. As the established PSG induction conditions are limited to the three conditions mentioned above, identifying other induction conditions will elucidate these mechanisms. The current study showed that acetic acid stress induced PSG formation more rapidly than other PSG induction conditions following the formation of nuclear proteasome condensates. Acetic acid inhibited proteasomal proteolysis, leading to the accumulation of ubiquitinated proteins. Of the non-essential proteasome subunits, Sem1/Dss1 was crucial, but Pre9 was less important for PSG formation under all conditions. The necessity of Rpn10 and Rpn13 for PSG formation differed depending on induction conditions. The proteasome shuttle factors, Dsk2 and Rad23, colocalized with all PSGs but were not essential for PSG formation caused by acetic acid or glucose depletion. The contributions of Hul5, an E3 ubiquitin ligase, and new protein synthesis to PSG formation also differed among the conditions, indicating that the key factors for PSG formation varied among the different induction conditions. These findings provide novel insights into the physiological effects of acetic acid on proteasomes and PSG-formation mechanisms.

Figures

Full text

In the proteostasis network, proteasomes are engaged in protein degradation along with autophagy (1, 2). The 26S proteasome is a large ATP-dependent protease machinery (approximately 2.5 MDa) that is ubiquitously found in eukaryotes (3, 4, 5). The 26S proteasome primarily targets and degrades proteins via ubiquitination, known as the ubiquitin–proteasome system, which plays a crucial role in maintaining proteostasis and regulating cellular functions by selectively degrading abnormal, unnecessary, and regulatory proteins, such as cyclins and transcription factors (2, 3, 4).

The 26S proteasome consists of 19S regulatory particle(s) (RP) and a 20S core particle (CP). The 19S RP serves as a “gatekeeper” in an ATP-dependent manner by recognizing, deubiquitinating, unfolding, and translocating substrate proteins to the 20S CP for proteasomal proteolysis (3, 4, 5, 6). The 20S CP consists of four stacked heteroheptameric rings (the two outer α-rings and the two inner β-rings), and the β-rings possess chymotrypsin-like, trypsin-like, and caspase-like proteolytic activities (3, 4, 5, 6). Furthermore, free 20S CP can degrade proteins independently in a ubiquitin-/ATP-independent manner; proteins with unstructured regions due to oxidative stress and intrinsically disordered proteins (IDPs) are susceptible to degradation by free 20S CP (7, 8).

Although proteasomes are predominantly located in the nuclei of proliferating cells, their localization changes dynamically in response to stress and growth conditions (9, 10, 11, 12). Mammalian proteasomes form nuclear condensates in response to diverse stress conditions, including hyperosmotic shock, amino acid starvation, and cellular senescence (11, 12, 13, 14, 15). Mammalian proteasome condensates contain ubiquitinated proteins and the proteasome shuttle factor RAD23B, which contains ubiquitin-like (UBL) and ubiquitin-associated (UBA) domains (11, 14, 16, 17, 18). Yeast proteasomes have also been reported to localize in the nuclear periphery of stationary-phase cells, or of cells treated with MG132, a proteasome inhibitor (19, 20). Regarding cytoplasmic proteasome condensates, the formation of proteasome storage granules (PSGs) is a well-known phenotype unique to yeast cells, but it has not been confirmed in mammalian cells (15). PSGs are also known to contain ubiquitinated proteins and the UBL–UBA shuttle factors, Dsk2 and Rad23 (a yeast homolog of mammalian Rad23B) (21). Changes in proteasome localization may suggest the optimization of protein quality control and stress resistance; however, the physiological significance of this has yet to be fully elucidated.

PSG formation is induced by acute glucose depletion, quiescence, and mitochondrial stress (9, 15, 21, 22, 23, 24), and these conditions cause decreases in intracellular pH (pHi) and ATP levels (21, 25). Peters et al. (22) demonstrated that intracellular acidification alone is sufficient to drive PSG formation. Furthermore, PSGs behave like liquids, similar to condensates formed by liquid–liquid phase separation (LLPS), and they quickly dissociate upon glucose replenishment (9, 21, 22). Although the involvement of LLPS in PSG formation has been suggested (21), the detailed mechanism is poorly understood. Currently, the established PSG induction conditions are limited to the three conditions mentioned above, and it is of interest to determine whether other induction conditions exist. Even among these three known conditions, differences exist in the factors necessary for PSG formation (21). Therefore, identifying new PSG formation conditions and analyzing their characteristics will be meaningful in clarifying the formation mechanisms and physiological significance of PSGs.

Acetic acid is a carboxylic acid classified as a weak acid, with a dissociation constant (pKa) of 4.76 at 25 °C. As a physiological substrate, acetic acid is produced during metabolism in the human body and by the intestinal bacteria (26). In the budding yeast Saccharomyces cerevisiae, acetic acid is a by-product of alcoholic fermentation. In contrast, exogenous acetic acid often acts as a stressor and fermentation inhibitor in yeast cells, inducing apoptosis and endoplasmic reticulum stress (27, 28, 29, 30, 31) and disturbing the lipid composition and functions of the cellular membrane (31, 32, 33, 34, 35, 36). High concentrations of acetic acid induce apoptosis-like cell death (80–120 mM) and necrosis (160–200 mM) in yeast cells (27, 36).

The toxicity of acetic acid is primarily attributed to its undissociated form due to its ability to enter cells by passive diffusion through the plasma membrane, whereas dissociated acetic acid is mostly nonpermeable (37). Once inside the cell, acetic acid dissociates into protons (H^+^) and acetate anions (CH_3_COO^–^) in the near-neutral cytoplasm, thereby acidifying the cellular interior (37, 38). Intracellular acidification caused by acetic acid leads to ATP depletion and inhibition of cell proliferation in yeast via inhibition of the respiratory chain and glucose uptake or via activation of Pma1, an essential plasma membrane H^+^-ATPase (35, 36, 39, 40, 41).

Based on this background, sublethal concentrations of acetic acid (<80 mM) were hypothesized to serve as novel induction conditions for PSG formation and to be suitable for analyzing the PSG formation process. In the present study, the effects of sublethal acetic acid stress on the localization and activity of yeast proteasomes were examined. We found that 0.3% v/v (52.5 mM) acetic acid caused yeast proteasomes to quickly form condensates in the nucleus, followed by PSG formation. The differences and similarities between acetic acid stress and the previously reported PSG induction conditions were comprehensively compared in terms of the factors necessary for PSG formation. These results elucidate the physiological effects of acetic acid on yeast cells and the mechanisms underlying PSG formation.

Results

Acetic acid causes the formation of proteasome condensates in the nucleus and cytoplasm

To monitor the formation of proteasome condensates, we first used GFP-tagged Rpn5, which is a subunit of the lid substructure of 19S RP and has previously been used as a PSG marker (21, 22, 42). As previously reported (22), upon glucose depletion, Rpn5-GFP formed distinct condensates in the cytoplasm, namely PSGs (Fig. 1A).Figure 1**Acetic acid stress causes the formation of yeast proteasome condensates in the nucleus and cytoplasm.**A, yeast cells expressing Rpn5-GFP (a subunit of 19S RP lid) and Nup116-mRFP (nuclear envelope marker) were treated with acetic acid stress or glucose depletion for the indicated time. The pH values of synthetic defined (SD) media containing 0%, 0.1% v/v (17.5 mM), 0.2% (35 mM), or 0.3% (52.5 mM) acetic acid were 5.30, 3.45, 3.23 and 3.12, respectively. White and yellow arrows indicate the nuclear proteasome condensates (NucPCs) and cytoplasmic proteasomal storage granules (PSG)-like condensates (CytoPCs), respectively. B, propidium iodide (PI) staining was conducted to evaluate cell death. C, cells treated with 0.3% acetic acid for 9 h were transferred into fresh SD medium without acetic acid. D, cells expressing Rpn5-GFP were treated with glucose depletion for 24 h, whereafter glucose replenishment was conducted by transferring the cells to fresh SD medium. Cells were also treated with 0.3% acetic acid for 9 h, then transferred to fresh SD medium containing 0.3% acetic acid and incubated for 60 min. E, quantitative data for cells containing proteasome condensates were classified into three categories. The experiment was repeated three times, with a total of more than 300 living cells observed under each condition. Light blue bars indicate cells without NucPCs and CytoPCs. Light green bars indicate cells containing only NucPCs. Dark green bars refer to cells possessing both NucPCs and CytoPCs. Scale bar, 5 μm.

A concentration of 0.3% v/v (52.5 mM) acetic acid was previously confirmed to be a sublethal stress that inhibited proliferation without causing rapid cell death (30, 41). Therefore, the following experiments were conducted using acetic acid at concentrations of 0.3% or less. Rpn5-GFP also showed localization changes upon acetic acid stress treatment; Rpn5-GFP accumulated at the nuclear membrane periphery and formed condensates (NucPCs, indicated by white arrows) within a few hours of transferring the cells to synthetic defined (SD) medium containing 0.3% acetic acid. Furthermore, after 6 to 9 h of treatment with 0.3% acetic acid, Rpn5-GFP formed PSG-like condensates in the cytoplasm (CytoPCs, yellow arrows) in addition to NucPCs (Fig. 1A). Although not as effective as 0.3% acetic acid, 0.2% acetic acid induced CytoPC formation. No cells containing only CytoPCs without NucPCs were observed after 9 h of treatment with 0.2% or 0.3% acetic acid. In contrast, acetic acid concentrations less than 0.1% did not cause changes in the localization of Rpn5-GFP.

Although cell proliferation is inhibited by 0.3% acetic acid (41), no substantial increase in dead cells was observed by microscopic examination following treatment with 0.3% acetic acid. Additionally, the percentage of propidium iodide (PI)-negative cells remained high (>90%) after 9 h of treatment with 0.3% acetic acid (Fig. 1B). These results confirmed that apoptosis was scarcely induced by 0.3% acetic acid stress in our experimental system.

CytoPCs formed by 0.3% acetic acid disappeared within 10 min of acetic acid elimination, and most NucPCs disappeared within 30 min (Fig. 1C), indicating the rapid reversibility of these condensates. PSGs formed under glucose depletion dissociate upon glucose replenishment (9, 22), and this was confirmed in the current study (Fig. 1D). In contrast, CytoPCs formed after 9 h of treatment with 0.3% acetic acid did not disappear even after glucose replenishment (Fig. 1D), suggesting that glucose depletion was not the reason for their formation.

Acetic acid causes the formation of condensates of various proteasome subunits

The localization of all 19 subunits of 19S RP and six subunits of 20S CP (Pre1, Pre2, Pre4, Pre6, Pre8, and Pup1) under acetic acid stress was examined (Figs. 2 and S1). Because 20S CP is highly stable, both biochemically and structurally (43, 44, 45), analysis of the six subunits is considered sufficient to obtain accurate information regarding the localization of 20S CP. None of the subunits examined showed substantial differences. They initially accumulated at the nuclear periphery, formed NucPCs, and then formed CytoPCs under 0.3% acetic acid stress. Each proteasome subunit formed CytoPCs in more than 60% of the cells after 9 h of treatment with 0.3% acetic acid. In addition, each 19S subunit co-localized with Pre2 or Pre4 of the 20S CP subunit to form condensates (Figs. 2 and S1). No cells containing only CytoPCs without NucPCs were observed under 0.3% acetic acid stress.Figure 219S RP subunits form condensates with 20S CP subunits under acetic acid stress. Cells expressing GFP-tagged Rpn5 (a subunit of 19S RP lid), Rpt3 (a subunit of 19S RP base), Pre1 (β4 of 20S CP), or Pre8 (α2 of 20S CP) and mRFP-tagged Pre2 (β5 of 20S CP) or Pre4 (β7 of 20S CP) were treated with 0.3% acetic acid for the indicated time. White and yellow arrows indicate the nuclear proteasome condensates (NucPCs) and cytoplasmic PSG-like condensates (CytoPCs), respectively. Scale bar, 5 μm. The right panels indicate the condensate formation rate of proteasome subunits examined in this study after 9 h of treatment with 0.3% acetic acid. Light green bars indicate cells containing only NucPCs. Dark green bars refer to cells possessing both NucPCs and CytoPCs. No cells were found to form only CytoPCs. No condensate formation was observed in any subunit under non-stressed conditions. Representative images are also shown in Figure S1.

The rate of PSG formation caused by acute glucose depletion and mitochondrial stress (sodium azide) was previously reported as approximately 60% in all cells (21, 24). Using a formation rate exceeding 60% as an indicator, we concluded that acute acetic acid stress induced PSG formation in yeast cells. Hereafter, the CytoPCs formed under acetic acid stress are referred to as PSGs.

Proteasome activity is repressed, and ubiquitinated proteins accumulate under acetic acid stress

The relationship between PSG formation and proteasome activity under acetic acid stress was examined. To monitor proteasomal proteolysis under stress conditions in vivo, the efficiency of proteasomal proteolysis was investigated using cycloheximide (CHX)-chase analysis with an auxin-inducible degron (AID) system (46). As previously reported (47, 48), proteasomal proteolysis of Paf1–AID∗–6FLAG was rapidly induced by the addition of indole-3-acetic acid (IAA), the most common type of auxin. However, this effect was clearly blocked by 0.3% acetic acid (Fig. 3A–C).Figure 3**Acetic acid stress inhibits proteasomal proteolysis.**A, proteasomal proteolysis was monitored using the auxin-inducible degron (AID) system. The Paf1 AID system (Paf1–AID∗–6FLAG) was activated by treatment with 0.75 mM indole-3-acetic acid (IAA). Cells expressing Paf1–AID∗–6FLAG were treated with 0.3% acetic acid or without acetic acid in the presence of 200 μg/ml cycloheximide (CHX) and 0.75 mM IAA for the indicated times. B, ubiquitin-independent proteasomal proteolysis of Spe1-3HA was induced by spermidine (SPD). Cells expressing Spe1-3HA were treated with 0.3% acetic acid for the indicated time periods in the presence of 200 μg/ml CHX and 1 mM SPD. C, the Paf1–AID∗–6FLAG and Spe1-3HA levels were normalized to the Ponceau S staining results. n.s., statistically not significant; ∗∗p < 0.01 using Dunnett’s test. D, the levels of ubiquitinated proteins were detected using western blotting with an anti-ubiquitin antibody. Ubiquitinated protein levels were normalized to the Ponceau S staining results (right bottom panel). Significant differences were evaluated using Dunnett’s test (n = 3). ∗∗p < 0.01; ∗p < 0.05.

The effect of acetic acid on ubiquitin-independent proteasomal proteolysis was examined. Ornithine decarboxylase (ODC), mammalian thymidylate synthase, and yeast Rpn4 and Pih1 are known to be degraded by the proteasome in a ubiquitin-independent manner (49, 50, 51, 52). Proteasomal degradation of Spe1, the yeast ODC, is ubiquitin-independent and is induced by polyamines such as spermidine (48, 53). CHX-chase analysis confirmed that spermidine-induced degradation of Spe1-3HA was blocked in the presence of 0.3% acetic acid (Fig. 3, B and C). Using the same experimental system, the repression of proteasomal proteolysis of Paf1–AID∗–6FLAG and Spe1-3HA under glucose depletion was also observed (Fig. S2). Furthermore, increased levels of ubiquitinated proteins were observed in the yeast cells under acetic acid stress (Figs. 3D and S3). These results clearly indicate that proteasomal proteolysis was quickly inhibited under 0.3% acetic acid stress.

PSG is not formed by mild intracellular acidification caused by lactic acid and severe ethanol stress

As intracellular acidification has been reported to drive PSG formation (22), we investigated whether, as well as by acetic acid, PSG formation was also induced by lactic acid or severe ethanol stress, which are known to acidify the intracellular environment (54, 55, 56). Additionally, severe ethanol stress inhibits proteasome activity (48). Contrary to expectations, neither lactic acid (113 mM or 170 mM) nor severe ethanol stress (10% v/v) triggered PSG formation at all (Fig. 4A). Measurement of intracellular pH (pHi) indicated that the decrease in pHi caused by lactic acid or severe ethanol stress was smaller than that induced by 0.3% acetic acid and glucose depletion (Fig. 4B), suggesting that a sufficient decrease in pHi, comparable to that induced by 0.3% acetic acid and glucose depletion, might be required for PSG formation. In contrast, when pHi was adjusted to 5.5 using 2,4-dinitrophenol (2,4-DNP) (nearly equivalent to the pHi of cells treated with 0.3% acetic acid) (57), PSG formation was induced in more than half of the cells within 3 h (Fig. S4).Figure 4**Lactic acid and severe ethanol stress do not induce PSG formation.**A, cells expressing Rpn5-GFP were treated with lactic acid (113 mM and 170 mM) or 10% v/v ethanol for 9 h. Scale bar, 5 μm. B, intracellular pH was determined by measuring pHluorin2 using flow cytometry. Cells were treated with stress conditions (0.3% acetic acid for 9 h, lactic acid for 9 h, 10% v/v ethanol for 9 h, and glucose depletion for 24 h).

Shuttle factors are included in the PSGs but are not essential for PSG formation under acetic acid stress

The UBL–UBA shuttle factors, Dsk2 and Rad23, are crucial for PSG formation due to quiescence or mitochondrial stress caused by sodium azide but not for PSG formation due to acute glucose depletion (21). The importance of Dsk2 and Rad23 in acetic acid-induced PSG formation was examined using dsk2Δrad23Δ cells. PSG formation was observed in more than half of the dsk2Δrad23Δ cells under 0.3% acetic acid stress, although the formation rate was slightly lower than in wild-type cells (Fig. 5A). These results suggest that Dsk2 and Rad23 are less important for acetic acid-induced PSG formation than for quiescence- and mitochondrial stress-induced PSG formation. Dsk2-GFP and Rad23-GFP co-localized with Pre2-mRFP, forming condensates in the nucleus and cytoplasm (Fig. 5B), suggesting that PSGs caused by acetic acid stress also included these shuttle factors.Figure 5**Dsk2 and Rad23 are included in PSGs but are not essential for PSG formation under acetic acid stress.**A, the formation of PSGs in dsk2Δrad23Δ cells was examined using Rpn5-GFP under 0.3% acetic acid stress. Yellow arrows indicate PSGs. The right panel indicates the PSG formation rates of wild-type and dsk2Δrad23Δ cells after 9 h of treatment with 0.3% acetic acid. ∗p < 0.05 using Student’s t test. B, intracellular localization of Dsk2-GFP or Rad23-GFP was examined using wild-type cells expressing Pre2-mRFP. C, PSG formation was examined in ddi1Δ cells expressing Rpn5-GFP. n.s., statistically not significant using paired Student’s t test. D, intracellular localization of Ddi1-GFP was examined using wild-type cells expressing Pre2-mRFP. White arrowheads indicate Ddi1-GFP condensates that did not co-localize with Pre2-mRFP. Scale bar, 5 μm.

S. cerevisiae contains another UBL–UBA shuttle factor, Ddi1, which has a viral protease domain that cleaves substrates containing long poly-ubiquitin chains (21, 58). Thus, ddi1Δ cells accumulate poly-ubiquitinated substrates and form PSGs even during logarithmic growth without stress (21, 52). The formation of PSGs in ddi1Δ cells was observed under non-stressed conditions (Fig. 5C). Treatment with 0.3% acetic acid had little effect on the PSG formation rate in ddi1Δ cells. Ddi1 forms condensates in quiescent cells but does not necessarily co-localize with PSGs (21). The formation of Ddi1 condensates in wild-type cells was induced in the cytoplasm after a brief acetic acid stress treatment (1 h), during which PSGs did not form (white arrowheads in Fig. 5D). Ddi1 condensate formation also preceded PSG formation during glucose depletion (Fig. S5). Even under acetic acid stress and glucose depletion, Ddi1 condensates localized independent of PSG (Figs. 5D and S5).

Cycloheximide does not inhibit the formation of proteasome condensates induced by acetic acid

In addition to the requirements for Dsk2 and Rad23, differential requirements for new protein synthesis in PSG formation have been reported. PSG formation induced by sodium azide or transition to quiescence is blocked by the addition of cycloheximide (CHX), whereas PSG formation due to glucose depletion is not prevented by CHX (21). Similar to the PSG formation induced by glucose depletion, acetic acid-induced PSG formation was not blocked by the addition of CHX (Fig. 6).Figure 6Cycloheximide does not disrupt PSG formation under acetic acid stress. Cells expressing Rpn5-GFP were treated with 0.3% acetic acid for 9 h in the presence or absence of 200 μg/ml cycloheximide (CHX). Yellow arrows indicate PSGs. N.D., not detected. n.s., statistically not significant using Student’s t test. Scale bar, 5 μm.

PSG formation is repressed in rpn13Δ and sem1Δ cells

In addition to dsk2Δrad23Δ cells, the formation of PSGs in various mutant cells deficient in proteasome-related factors was examined. Of the genes encoding the proteasome subunits, PRE9, RPN10, RPN13, and SEM1 (also known as DSS1) are not essential for yeast survival. We examined PSG formation in pre9Δ, rpn10Δ, rpn13Δ, and sem1Δ cells and found that PSGs were formed in pre9Δ and rpn10Δ cells, as well as in wild-type cells, under acetic acid stress (Fig. 7A). However, distinct PSGs were less likely to form in rpn13Δ and sem1Δ cells under acetic acid stress. The introduction of pRS313-RPN13 and pRS313-SEM1 into rpn13Δ and sem1Δ cells, respectively, restored the formation of PSGs, suggesting that Rpn13 and Sem1/Dss1 play a crucial role in acetic acid-induced PSG formation. The sem1 deletion also markedly repressed PSG formation induced by acute glucose depletion, sodium azide, and transition to quiescence (Fig. S6), indicating that Sem1/Dss1 is crucial for PSG formation under all four PSG induction conditions. In contrast, it has been reported that PSG formation is induced by acute glucose depletion in rpn13Δ cells (21). Dsk2 and Rad23 formed cytoplasmic condensates even in rpn13Δ and sem1Δ cells under acetic acid stress (Fig. 7B), suggesting that the formation of the shuttle factor condensates might not be dependent on PSG formation under acetic acid stress.Figure 7**PSG formation is mitigated in rpn13Δ and sem1Δ cells under acetic acid stress.**A, formation of the proteasome condensates in pre9Δ, rpn10Δ, rpn13Δ, and sem1Δ cells was examined using Rpn5-GFP. Cells were treated with 0.3% acetic acid for the indicated period. pRS313-RPN13 and pRS313-SEM1 were introduced into rpn13Δ and sem1Δ cells, respectively. The empty vector (pRS313) was also introduced as a control. White and yellow arrows indicate NucPCs and PSGs, respectively. The PSG formation rates under non-stress conditions and after 9 h of treatment with 0.3% acetic acid are shown in the bottom left panel. N.D., not detected. Significant differences were evaluated using Student’s t test. ∗∗∗p < 0.005; ∗∗p < 0.01; ∗p < 0.05; n.s., statistically not significant. B, condensate formation of shuttle factors (Dsk2-GFP and Rad23-GFP) was examined in rpn13Δ and sem1Δ cells. Yellow arrowheads indicate Dsk2-GFP and Rad23-GFP condensates that did not colocalize with Pre2-mRFP. Scale bar, 5 μm.

Hul5-mediated ubiquitination is involved in PSG formation due to quiescence (20). Hul5 is a proteasome-interacting E3 ubiquitin ligase (59), and hul5Δ cells show remarkable nuclear retention of the proteasomes without PSG formation in quiescent cells (20, 60). The failure of PSG formation in quiescent hul5Δ cells was confirmed (Fig. 8). Additionally, mitochondrial stress did not induce PSG formation in hul5Δ cells. In contrast, PSG formation was observed in hul5Δ cells under acetic acid stress and glucose depletion. The survival rates of the null mutants used in the current study, including hul5Δ, remained high after the treatment with 0.3% acetic acid for 9 h (Fig. S7).Figure 8PSG formation in hul5Δ cells under stress conditions. The hul5Δ and wild-type cells expressing Rpn5-GFP were treated with the stress conditions (0.3% acetic acid for 9 h, glucose depletion for 24 h, mitochondrial stress with 0.5 mM NaN_3_ for 24 h, and transition to quiescence by culturing more than 24 h in YPD medium). Yellow arrows indicate PSGs. Scale bar, 5 μm. The ratio of cells that formed PSGs was compared. ∗∗∗p < 0.005; ∗∗p < 0.01; ∗p < 0.05 using Student’s t test.

No distinct insoluble protein deposits are formed by acetic acid stress

PSGs caused by glucose depletion are transiently associated with cytoplasmic insoluble protein deposits (IPODs) during their formation process (42, 61). Whether a similar phenomenon occurs under acetic acid stress was determined using Hsp42-GFP, an IPOD marker (61). Unlike glucose depletion and severe ethanol stress, the expression level of Hsp42-GFP was low, and no distinct IPOD formation was observed (Fig. 9). Hsp104, another marker of denatured protein deposits (62), also failed to form distinct foci, indicating that IPODs were rarely formed under 0.3% acetic acid stress.Figure 9Acetic acid does not cause the formation of distinct insoluble protein deposits (IPODs). Cells expressing an IPOD marker (Hsp42-GFP or Hsp104-mRFP) and a proteasomal subunit marker (Rpn5-GFP or Pre2-mRFP) were treated with 0.3% acetic acid stress or glucose depletion for the indicated period. Scale bar, 5 μm.

Discussion

This study highlights the novel physiological effects of acetic acid on yeast proteasomes. A concentration of 0.3% acetic acid (52.5 mM) induced the formation of proteasome condensates in the nucleus and cytoplasm. Additionally, acetic acid stress repressed proteasomal proteolysis. Cell-free extracts of yeast cells treated with 80 mM acetic acid, which induces apoptosis, have been shown to retain high proteasomal protease activity in in vitro assays (63). Thus, it appears that proteasomes are in an inactive state rather than irreversibly disrupted in yeast cells under acetic acid stress. Furthermore, ATP depletion destabilizes proteasomes to an inactive state (15, 64, 65), and proteasome activity is repressed under glucose depletion, mitochondrial stress, and during the stationary phase (64, 65, 66, 67, 68) (Fig. S2). Therefore, the acetic acid-induced depletion of ATP and PSG formation seems to support the inactivation of proteasome holo-enzyme activity under acetic acid stress. This inactive state of the proteasome may not be a critical problem because 0.3% acetic acid inhibits the proliferation of yeast cells (41), reducing the need to precisely degrade cell cycle regulators such as cyclins. Furthermore, proteasome inactivation contributes to reduced ATP consumption under acetic acid stress.

The properties of PSG formation induced by acetic acid were compared with those induced under other known conditions (Fig. 10). Acetic acid stress induced PSG formation in a shorter time than the other three induction conditions, offering the advantage of reduced processing time. Similar to the PSGs induced by other conditions, PSG formation due to acetic acid was reversible, as the PSGs quickly disassembled upon the elimination of acetic acid. This reversibility fits well with the model in which PSGs are formed by LLPS. Waite et al. (21) recently proposed a model in which PSG formation is driven by the induction of LLPS via multivalent interactions between proteasomes, shuttle factors, and ubiquitinated proteins. Under all four PSG induction conditions, the intracellular environment appears to be conducive to inducing LLPS. In addition to acetic acid stress, other previously-reported PSG induction conditions cause a decrease in intracellular ATP levels and pHi (22, 25, 69, 70, 71, 72). Furthermore, decreases in intracellular ATP levels and pHi lead to increased macromolecular crowding within the cytoplasm, which has been implicated in LLPS induction (57, 73). Although the detailed mechanisms are unclear (74, 75), increased macromolecular crowding may be one of the driving forces for PSG formation via LLPS.Figure 10Comparison of factors involved in PSG formation under four different PSG induction conditions. Properties of the PSG-formation-related factors examined in this study were added to the summary of Waite et al. (2024) ^1)^ along with the results of Li et al. (2019) ^2)^, van Deventer et al. (2015) ^3)^, and Ohigashi et al. (2025) ^4)^. PSG formation rate: ++++, comparable to wild-type cells (WT); +++, less than 70% of WT; ++, less than 50% of WT; +, less than 30% of WT; –, almost no formation (less than 5% of WT).

Co-localization of shuttle factors, Dsk2 and Rad23, to PSGs was another commonality among the four PSG induction conditions. However, the requirements for these shuttle factors in PSG formation differ depending on the PSG induction conditions. PSG formation is induced in dsk2Δrad23Δ cells by acute glucose depletion and acetic acid stress, but rarely induced by mitochondrial stress or transition to quiescence (21). The promotion of multivalent interactions between ubiquitinated substrates and the proteasome via Dsk2 and Rad23 appears to be less important for PSG induction due to glucose starvation or acetic acid stress than that due to mitochondrial stress and quiescence. In mammalian cells, the shuttle factor, Rad23B, is essential for the formation of proteasome condensates and LLPS in ubiquitinated substrates (11, 14, 18). Collectively, these findings suggest that acute glucose depletion and acetic acid stress are unique conditions that induce PSG formation. Coincidentally, PSG formation induced by glucose depletion and acetic acid stress showed little requirement for the E3 ubiquitin ligase, Hul5. Acute glucose depletion and acetic acid stress may cause sufficient multivalent interactions between ubiquitinated substrates and proteasomes, even in the absence of shuttle factors or Hul5.

Different effects of CHX on PSG formation have been observed under different PSG induction conditions (21) (Fig. 6). Because yeast cells retain translational activity during treatment with sodium azide and during the culture process until the stationary phase, the effect of translation inhibition by CHX under these conditions could be substantial. In contrast, the inhibitory effect of CHX is limited under conditions of glucose depletion and acetic acid stress, which quickly lead to the global impairment of translation (31, 76, 77). The differences in translational activity under each condition may reflect the differential effects of CHX on PSG formation.

There are also differences in the importance of the intrinsic ubiquitin receptors, Rpn10 and Rpn13, in PSG formation. Rpn13, but not Rpn10, is essential for acetic acid-induced PSG formation, whereas Rpn10 is more important than Rpn13 for PSG formation induced by glucose depletion (21). Rpn10 is less important than Rpn13 for sodium azide-induced PSG formation, whereas both intrinsic ubiquitin receptors are important for PSG formation caused by quiescence (21). Rpn10 and Rpn13 are located away from each other at the edge of the 19S RP (78) and have been suggested to exhibit distinct substrate specificities, covering a broad range of ubiquitinated substrates (79). Rpn10 is inactivated via mono-ubiquitination (80), whereas mammalian Rpn13 is inactivated via polyubiquitination (81). Although it remains unclear whether yeast Rpn13 is polyubiquitinated, Rpn10 and Rpn13 are known to undergo various posttranslational modifications (82, 83). Differences in the modifications of Rpn10 and Rpn13 may lead to variations in the requirements of both subunits for PSG formation under the four PSG induction conditions.

Of the non-essential proteasome subunits, Pre9 (α3) appears to be less important for PSG formation. The rates of PSG formation caused by transition to quiescence, mitochondrial stress, and acetic acid were comparable in pre9Δ cells and wild-type cells (Fig. S6); PSG formation is also induced in pre9Δ cells by acute glucose depletion, albeit at a lower frequency (61). In contrast, Sem1/Dss1 appears to be essential for PSG formation under all four PSG induction conditions. PSG formation due to acute glucose depletion does not occur in sem1Δ cells (61). In the present study, PSG formation induced by acetic acid stress, mitochondrial stress, and transition to quiescence was also rarely observed in sem1Δ cells (Fig. S6). Sem1/Dss1 and its human homologue, DSS1, are involved in many different protein complexes, such as TREX-2, Thp3-Csn12, and human BRCA2 complexes, and the 26S proteasome (84, 85, 86, 87). During proteasome biogenesis, Sem1/Dss1 plays a role in the incorporation of Rpn3 and Rpn7 into 19S RP (88). Sem1/Dss1 is a small IDP, and IDPs and intrinsically disordered regions (IDRs) are well known to be deeply involved in LLPS (89, 90, 91, 92). The involvement of pH changes in LLPS of IDPs has also been reported (93). Because Rpn13 also has an IDR-like region, it would be interesting to determine whether the IDRs of Sem1/Dss1 and Rpn13 are involved in the driving force of PSG formation upon a drop in pHi.

Experimental procedures

Yeast strains and stress treatment

The parental wild-type strain BY4742 (MATα ura3Δ0 his3Δ1 leu2Δ0 lys2Δ0) and its isogenic knockout mutants (ddi1Δ, hul5Δ, pre9Δ, rad23Δ, rpn10Δ, rpn13Δ, and sem1Δ) were purchased from Open Biosystems (Huntsville). The AID system strain, W303 (PADH1-409-TIR1) Paf1–AID∗–6FLAG Lea1-3HA (47), was obtained from the National BioResource Project (Yeast). To construct a dsk2Δrad23Δ double knockout mutant, a DNA fragment (1.3 kb) encoding dsk2Δ::CgHIS3 was amplified by PCR using the primer set, dsk2Δ-F2/R2 (Table S1), and pCgHIS3 as the DNA template (94). The amplicon was introduced into rad23Δ cells to construct the dsk2Δrad23Δ mutant. The pCgHIS3 plasmid was provided by the National BioResource Project (Yeast). C-terminal tagging of Pre1, Rpn8, Rpn9, and Rpt3 with GFP was also introduced using the method of Longtine et al. (95) with pFA6a-GFP (S65T)-His3MX6 as the DNA template and primer sets of FA-F1/R1 (Table S1). The C-terminal tagging of Spe1 with 3HA has been described previously (48). fry320 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 ho::pTEF1-pHluorin2-tCYC1::HphMX) was used for the pHi assay (72). Yeast cells were cultured in synthetic defined (SD) medium (2% glucose, 0.67% yeast nitrogen base without amino acids, 20 mg/L uracil, 30 mg/L L-lysine HCl, 100 mg/L L-leucine, and 20 mg/L L-histidine HCl) with reciprocal shaking (120 rpm) at 28 °C, and exponentially growing cells were harvested at an optical density at 600 nm (OD_600_) of 0.5. The acetic acid stress treatment was conducted by resuspending the harvested cells in fresh SD medium containing acetic acid. Glucose replenishment after acetic acid stress treatment was achieved by transferring the cells to fresh SD medium containing glucose and acetic acid. The decrease in pHi by 2,4-DNP was achieved by transferring the cells to a pH-adjusted SD medium containing 2,4-DNP (57). Instead of distilled water, 100 mM potassium phosphate buffer was used to prepare the pH-adjusted SD media (pH 5.5 and 7.0) to which 2,4-DNP was added to achieve a final concentration of 2 mM. Other stress treatments, such as glucose depletion, mitochondrial stress using sodium azide, and transition to quiescence, were performed using the method of Waite et al. (21). The cell death rate was assessed using propidium iodide (PI) staining (96).

Plasmids

A portion of the open reading frame (ORF) of each gene was cloned into pJK67 (97) or YIp-HSP104-mRFP (98) to construct integrative plasmids for expressing GFP- or mRFP-tagged proteins, respectively. The ORFs of HSP42, NUP116, and the proteasome subunit genes were amplified by PCR using the genomic DNA of BY4742 as the template and primer sets of F1/R1 (Table S1). YIp*-PRE2*-mRFP-LEU2 was constructed by cloning the regions encoding the PRE2-mRFP into pRS305 (99). YIp*-NUP116*-mRFP-HIS3 was constructed by cloning the regions encoding the NUP116-mRFP into pRS303 (99). Although Pre2 tagged with GFP has been reported to cause a loss of proteasome activity (100), no loss of proteasome activity was observed when tagged with mRFP (Fig. S8). YIp-RPN1-GFP, YIp-RPN7-GFP, and YIp-PRE6-GFP were previously constructed (48). Each plasmid was linearized and introduced into yeast cells to be integrated into the chromosomal locus of each gene. The pRS313-RPN13 and pRS313-SEM1 were constructed by cloning the RPN13 and SEM1 genes, which were amplified using the primer sets, *RPN13-*F2/R2 and *SEM1-*F2/R2, respectively, into the SacI/XhoI sites of pRS313 (99).

CHX-chase analysis and western blotting

As described previously (48), CHX-chase analysis was performed to monitor the efficiency of proteasomal degradation using 200 μg/ml CHX (Nacalai Tesque). Ubiquitinated proteins were detected using an anti-ubiquitin antibody (P4D1; Santa Cruz Biotechnology). An anti-HA monoclonal antibody (M180–3; Molecular and Biological Laboratories Co., Ltd, Tokyo, Japan), anti-FLAG antibody (F1804; Sigma-Aldrich), and anti-mouse IgG, HRP-linked antibody (7076S; Cell Signaling Technology) were also used for western blotting. Western blotting bands were detected using Chemi-Lumi One L (Nacalai Tesque) and Multi Imager II Chemi Box (BioTools Inc.). Equal loading and transfer of all proteins were confirmed using Ponceau S staining. Protein levels were quantified using the ImageJ software and were normalized based on the Ponceau S staining results.

Fluorescent microscopic analysis

An Olympus IX83 microscope (Tokyo, Japan) was used to analyze proteasome condensate formation. Ten sequential images per field were captured by gradually changing the focal plane along the z-axis, and proteasome condensates were counted using the entire z-stack. The cells containing proteasome condensates were classified into three categories: no condensates, nuclear proteasome condensates only (NucPCs), and nuclear and cytoplasmic condensates (NucPCs and CytoPCs). To obtain quantitative data, more than 100 living cells were examined under each condition, and the experiments were independently repeated three times (more than 300 cells in total). Microscopic cell images were classified using ImageJ, visually reconfirmed, and quantitative data were calculated.

Measurement of intracellular pH

Intracellular pH (pHi) was measured as previously described using the genetically encoded pH sensor pHluorin2 (72, 101).

Statistical analysis

Statistical significance was evaluated using one-way ANOVA with unpaired Student’s t test or Dunnett’s post hoc test using RStudio (https://posit.co/products/open-source/rstudio/). In Figure 5C, however, paired Student's t test was used. Graphs were prepared using RStudio and Microsoft Excel. Data are expressed as the mean ± standard deviation (n = 3).

Data availability

All relevant data can be found within the article and its supporting information. The images, yeast strains, and plasmids are available upon request.

Supporting information

This article contains supporting information (Figs. S1–S8, Table S1) (57, 96).

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.