Introduction

The plasma membrane (PM) functions as a defensive barrier in all types of cells to protect cellular components from extracellular stimuli. PM frequently experiences physiological and pathological damages, such as eccentric contraction-induced injuries, migration-induced injuries, pore formation by bacterial toxins, or invasion by cancer cells (1, 2). Cells undergo cell lysis if the damaged PM is not repaired. Deficits in PM repair are linked to multiple diseases, including limb-girdle muscular dystrophy (3) and Scott syndrome (4, 5). Therefore, virtually all cells have PM repair machinery.

In eukaryotes, the influx of Ca²⁺ triggers PM repair (1). Upon Ca²⁺ influx, fundamental cellular mechanisms, including exocytosis (6-10), endocytosis (11, 12), membrane shedding by ESCRT complex (13, 14), and the constriction forces by actin cytoskeleton (15-18) are directed to the damage site to reseal the wound. These initial repair processes usually occur within 1 min after PM damage (19). After the resealing of the damaged PM, cells restructure the damaged PM (19-21). Restructuring is defined as the process by which cells modify the damaged PM to restore PM homeostasis and normal cell function (19-21). These repair processes are not mutually exclusive, and they collaborate to repair the damaged PM (22). However, the temporal coordination between the PM repair processes remains poorly understood.

Both resealing and restructuring of the damaged PM are mediated by the proteins, which accumulate at the damage site. Here, we defined them as PM repair proteins. PM repair proteins include ESCRT, annexins, and dysferlin in mammalian cells (13, 23-25). Using budding yeast Saccharomyces cerevisiae as a model, we previously identified evolutionarily conserved repair proteins, including protein kinase C, exocyst, and phospholipid flippases (26, 27). These repair proteins accumulate at the damage site at the appropriate time and proceed with the repair processes. Therefore, identifying repair proteins and observing their movements after PM damage is an important step toward understanding the temporal PM repair processes.

Here, we performed proteome-scale screening of PM repair proteins using yeast GFP-tagged libraries and laser-induced injury assay in budding yeast. We identified 80 repair protein candidates, including 72 previously unreported candidates. Among the observed repair processes, we showed that polarized exocytosis and clathrin-mediated endocytosis (CME) are coupled at the damage site, with exocytosis predominating in the early stage of PM repair and CME predominating in the late stage of PM repair. We also showed that CME delivers repair proteins with transmembrane domains (TMDs) from the growing bud site (bud tip) to the damage site, presumably contributing to the restructuring of the damaged PM. These results provide insights into the temporal dynamics of coordinated cellular responses to PM damage.

Results

Proteome-scale identification of PM/cell wall damage-responsive proteins

To identify the PM repair proteins, we performed a two-step screening. First, we aimed to identify the proteins that change localization in response to SDS treatment, which induces local PM/cell wall damage to budding yeast (28). We used a yeast C-terminally GFP-tagged library (4159 ORFs) (29) and an N-terminally sfGFP-tagged library (N’ Swat library) (5569 ORFs) (30, 31) comprising 86% of the yeast proteome (5718 ORFs in total) (Fig. 1A). We fixed the cells with paraformaldehyde after 1 hour of SDS treatment, in which condition the best-characterized repair protein Pkc1-GFP changes its localization (Fig. 1A). We successfully imaged the signal of 9181 proteins fused with GFP or sfGFP (5609 ORFs in total). We assessed the fluorescence signals of 5609 proteins in normal and SDS-treated conditions by visual inspection of the images. We identified 562 proteins whose fluorescence signal pattern changed after SDS treatment (Fig. 1B and Data S1). The hits included the localization changes of proteins, structural changes of organelles such as mitochondrial fragmentation, and foci/puncta formation in response to SDS treatment. In addition to Pkc1, we identified previously reported repair proteins, such as Rom2 and Dnf1, as screening hits (26, 27) (Data S1). Gene Ontology (GO) analysis of the screening hits revealed enrichment for proteins involved in the actin filament organization (Fig. 1C). This is consistent with the previous study that SDS treatment remodels the cell polarity and actin cytoskeleton (26). Based on the reported subcellular localization of proteins, we categorized the screening hits into six major classes and several minor ones representing less than ten proteins (Fig. 1D and fig. S1, see Materials and Methods). Among the hits, proteins that are localized to the cell periphery showed cytoplasmic dots after SDS treatment (Fig. 1D). This is consistent with the fact that endocytosis of PM proteins contributes to cell adaptation to environmental stress (32). We also found that proteins that form puncta/foci after the SDS treatment include translation factor Gcd7-GFP (Fig. 1D), which forms foci in response to glucose deprivation (33). These observations suggest that SDS treatment induces stress responses, including those associated with PM/cell wall damage responses in budding yeast.

Fig. 1.

Laser-induced PM/cell wall damage assay identified 80 repair protein candidates

Our group previously developed a laser-induced PM/cell wall injury assay under live single-cell conditions (laser damage assay) (26, 34). We performed the laser damage assay to identify the PM repair proteins and observe their movements after PM/cell wall damage (Fig. 2A). The N-terminally sfGFP-tagged library is expressed under exogenous NOP1 promoters, while the C-terminally GFP-tagged library is expressed under endogenous promoters (29, 30). To assess the proteins with endogenous expression levels, we selected the 234 screening hits from the C-terminal library as targets for the laser damage assay. During 25 min of observation at 30-second intervals, we identified 90 proteins whose localization changed in response to laser damage (Fig. 2B). We categorized the proteins into four classes based on their localization changes (Fig. 2C). Three of the puncta-forming proteins, Dna2, Dot6, and Gcd7, form puncta in response to cellular stresses (33, 35). In addition, the transcription factors, Msn2 and Crz1, translocate from the cytoplasm to the nucleus after laser damage (Fig. 2B). Msn2 and Crz1 are activated under various stress conditions (36). Thus, the localization changes of puncta-forming proteins and transcriptional factors suggest common stress responses to laser damage and other cellular stresses. The most frequently observed localization changes were the accumulation at the damage site (Fig. 2B). We defined them as repair protein candidates, given their presumed involvement in the PM/cell wall repair process.

Fig. 2.

To gain further insight into the repair protein candidates, we classified them into four categories based on their subcellular localization (Fig. 2C). The largest category is bud-localized proteins (47 proteins) (Fig. 2C), which is consistent with a previous study showing that bud-localized proteins accumulate at the damage site in budding yeast (26). The repair protein candidates in this category include ten proteins that have transmembrane domains (TMDs), including phospholipid flippases (Dnf1/Dnf2) and osmosensor proteins (Slg1, Sho1) (Fig. 2C). Proteins with TMDs are transported to the PM as cargoes of secretory vesicles (37). Thus, this result suggests that secretory vesicles targeted to the damage site provide repair proteins with TMDs to the damage site. The second largest category is actin-localized proteins (25 proteins). This is consistent with the GO analysis that actin-binding proteins and proteins involved in actin-related biological processes are enriched in the repair protein candidates (Fig. 2, D and E). The proteins in this category include endocytic proteins, suggesting that endocytosis occurs at the damage site. In addition, proteins that bind actin cables, such as Abp140, are included in this category, suggesting that actin cables are formed at the damage site. These results are consistent with previous studies of PM damage repair in budding yeast and human cells (11, 26).

The temporal order of the Pkc1 accumulation, polarized exocytosis, and CME at the damage site

To understand when repair proteins accumulate at the damage site, we defined their accumulation times as the time when the fluorescence intensity at the damage site exceeded the threefold standard deviation (3xSD) above the non-damaged site for at least two consecutive time frames (1 min). We also defined dispersion time when the fluorescence intensity at the damage site becomes less than that of the non-damaged site plus 3xSD for at least 1 min (Data S2).

The defined accumulation times raised the possibility that the Pkc1 pathway proteins accumulated at the damage site earlier than exocytosis regulators and exocytic cargoes with transmembrane domains (Dnf1, Dnf2, and Slg1) (Fig. S2). To verify this observation, we performed the laser damage assay in the cells expressing Exo70-mNeonGreen (mNG) and Pkc1-mScarlet-I (mSc-I), and the cells expressing Pkc1-sfGFP and Dnf1-mSc-I (Fig. 3A and B). Exo70 is one of the subunits of the exocyst complex, serving as a polarized exocytosis marker. The fluorescence-tagged Pkc1 accumulated at the damage site earlier than Exo70-mNG and Dnf1-mSc-I (Fig. 3, A and B). These results suggest that the Pkc1 accumulation at the damage site occurs earlier than the polarized exocytosis of transmembrane proteins.

Fig. 3.

Although not all exocyst subunits (eight subunits) were identified in this work, all the mSc-I-tagged exocyst subunits showed comparable fluorescence intensity changes with Exo70-mNG at the damage site and at the bud tip (Fig. S3, A to G). These results suggest that all subunits of the exocyst complex accumulate at the damage site simultaneously.

The CME markers, Sla1-GFP and Abp1-GFP (38), exhibited repeated short stays and reaccumulation at the damage site within 5 min to 25 min after laser damage (Fig. S4, A to C). These fluctuating movements of Sla1 and Abp1 are consistent with the previous studies (39, 40). Furthermore, Sla1-mNG and Abp1-mSc-I showed comparable fluctuating accumulation patterns in the same cell (Fig. S4D). These results suggest that CME repeatedly occurs at the damage site from around 5 min to 25 min after laser damage.

To understand the temporal order of polarized exocytosis and CME at the damage site, we performed the laser damage assay in the cells expressing Exo70-mNG/Dnf1-mNG and Ede1-mSc-I (Fig. 3, C and D). Ede1 is one of the earliest proteins recruited to endocytic sites, serving as a CME marker (41, 42). Exo70-mNG/Dnf1-mNG and Ede1-mSc-I simultaneously accumulated at the damage site from 5 min to 35 min (Fig. 3, C and D). The signals of Exo70-mNG/Dnf1-mNG peak within 20 min after the damage, while Ede1-mSc-I peaks 20 min after the damage (Fig. 3, C and D). These results suggest that CME and polarized exocytosis occur simultaneously at the damage site, with polarized exocytosis predominating in the early stage of the PM/cell wall damage response and CME predominating in the late stage of the PM/cell wall damage response.

Knockout mutants of CME are sensitive to PM/cell wall stresses

To identify the biological processes required for cell survival after PM/cell wall stress, we performed growth screening of repair protein knockout mutants in PM/cell wall stress conditions (Fig. S5A). We spotted the same number of yeast cells in YPD media, YPD+0.01% SDS media, YPD+25 μg/ml calcofluor white (CFW), and YPD media in 37 °C (Heat stress) and incubated for 3 days (Fig. S5A). CFW binds to chitin in the cell wall, inducing cell wall damage (43). Heat stress modifies PM structures (44). The growth assay showed that six knockout mutants of CME proteins, rvs167Δ, sla1Δ, sla2Δ, end3Δ, las17Δ, and vrp1Δ, are sensitive to all the stress conditions (Fig. S5B). We knocked out these genes using the WT strain in our laboratory and confirmed that these mutants are sensitive to SDS (Fig. S5C). These results led us to focus further on the CME functions associated with PM/cell wall damage repair.

CME proteins are required for polarized exocytosis at the damage site

Previous studies showed that endocytosis is involved in PM damage repair by directly resealing the PM by removing the damaged pores (45) or by restructuring the PM after resealing in mammalian cells (20, 21, 46). We showed that CME occurs at the damage site from 5 to 45 min after laser damage in budding yeast (Fig. 3, B and C). This supports the possibility that CME restructures the PM after resealing, which usually occurs within 1 min after the damage (20). Another potential mechanism for restructuring the damaged PM is exocytosis, which delivers PM proteins and lipids to the damage site. Our previous work showed that the polarized exocytosis machinery is directed from the bud tip to the damage site in response to laser damage in budding yeast (26). Given that CME and polarized exocytosis constitute a coupled transport cycle in budding yeast (47), we reasoned that CME positively regulates polarized exocytosis at the damage site, thereby restructuring the damaged PM.

To test this idea, we performed the laser damage assay in SDS-sensitive endocytic mutants (sla1Δ, end3Δ, vrp1Δ, and rvs167Δ). We used two exocytosis markers, type V myosin (Myo2) and the exocyst subunit (Exo70), fused to sfGFP and mNG, to assess the exocytosis activity at the damage site. The accumulation of Myo2-sfGFP and Exo70-mNG at the damage site was impaired in sla1Δ, end3Δ, and vrp1Δ (Fig. 4, A and B, fig. S6, A and B). Moreover, we also found that the dispersion of Myo2-sfGFP and Exo70-mNG from the bud tip was partially inhibited in CME mutants (Fig. 4, A and B, fig. S6, A and B). The observed weak phenotype of rvs167Δ for the accumulation and dispersion of Myo2-sfGFP and Exo70-mNG is consistent with the previous study that exocytosis is not impaired in rvs167Δ (47).

Fig. 4.

Our previous work also showed that the actin nucleator formin Bni1 and the exocyst subunit Sec3 are degraded in response to SDS treatment (26). To test the possibility that CME is required for the decrease of Bni1 and Sec3 after SDS treatment, we performed immunoblotting of Bni1-13xMyc and Sec3-GFP before and after SDS treatment (Fig. S7, A and B). Protein levels of Bni1-13xMyc and Sec3-GFP were decreased in WT, rvs167Δ, and end3Δ after SDS treatment (Fig. S7, A and B). Given that end3Δ impaired the accumulation of Myo2-sfGFP and Exo70-mNG at the damage site (Fig. 4, A and B), these results suggest that CME is dispensable for the decrease of Bni1 and Sec3 after SDS treatment. Altogether, these results suggest that CME is involved in the direction of the Exo70-mNG and Myo2-sfGFP to the damage site in a Bni1- and Sec3-degradation-independent manner.

CME at the bud tip directs repair proteins with TMDs to the damage site

Although the requirement of Rvs167 for the Pkc1 accumulation and polarized exocytosis at the damage site is minimal (Fig. 4, A to C), rvs167Δ is sensitive to SDS (Fig. S5C). This raised the possibility that CME plays additional roles, such as membrane protein trafficking, in PM repair. We found that CME proteins, including Apl1, Ede1, and Ent1, changed their localization from the bud neck to the bud tip within 10 min after laser damage (Fig. S8A). The fluorescence intensity changes of CME proteins at the bud tip at 10 min after laser damage were higher than those of other repair protein candidates (Fig. S8B). Consistent with these results, the Ede1-mSc-I signal at the bud tip increased within 10 min after laser damage, and then the Ede1-mSc-I signal at the damage site increased (Fig. 5A). These results suggest that CME occurs at the bud tip within 10 min after laser damage. We also found that Dnf1-mNG disappeared from the bud tip following the recruitment of Ede1-mSc-I to the bud tip (Fig. 5, A and B). The disappearance from the bud tip and the accumulation at the damage site of Dnf1-mNG were impaired in rvs167Δ (Fig. 5, B and C, and fig. S9). Furthermore, the disappearance from the bud tip and the accumulation at the damage site of repair proteins with TMDs (Slg1-sfGFP and Sho1-GFP) and representative endocytic recycling cargo, mNG-Snc1, were impaired in rvs167Δ (Fig. 5, D to F and fig. S10, A to C). These results support our idea that CME directs repair proteins with TMDs from the bud tip to the damage site.

Fig. 5.

mNG-Snc1 is retargeted from the damage site to the bud tip

The fluorescence intensity of accumulated repair proteins with TMDs at the damage site gradually decreases approximately 15 min after laser damage (Fig. 3D and fig. S10, A and C). Using mNG-Snc1 as a model, we investigated the destination of repair proteins after PM repair. To observe the same fraction of mNG-Snc1 before and after laser damage, we transiently expressed mNG-Snc1 under the control of the galactose-inducible promoter (Gal1pr) for 1 hour (Fig. 6A). Then, we transferred the cells to glucose media to stop the expression (Fig. 6A). To minimize the effect of changing the carbon source, we further incubated the cells in glucose media for at least 3 hours prior to laser damage (Fig. 6A). The mNG-Snc1 accumulated at the damage site approximately ~10 min after laser damage, gradually disappearing after the colocalization with Ede1-mSc-I (Fig. 6A and B). The mNG-Snc1 signal at the bud tip decreased to 10-53% 22 min after laser damage (Fig. 6C and D). The normalized mNG-Snc1 signal at the bud tip recovered to 54-88% 50 min after laser damage. (Fig. 6, C and D). These results suggest that mNG-Snc1 is redirected to the bud tip after PM repair.

Fig. 6.

Discussion

There has been a growing interest in the mechanisms underlying PM repair, partly due to their association with human diseases and cellular aging (3-5, 28). PM repair proteins, which accumulate at the damage site, play critical roles in PM repair. In this study, by large-scale identification of PM repair proteins and single- and dual-color live-cell imaging of repair proteins, we analyzed spatiotemporal PM repair processes in budding yeast. We propose a model in which CME at the bud tip and at the damage site delivers repair proteins with TMDs between the bud tip and the damage site, allowing the cell to restructure the damaged PM and to resume growth after PM repair (Fig. 7).

Fig. 7.

Two-step visual screening for PM repair protein identification

By combining proteome-scale visual screening using yeast GFP libraries and the laser damage assay, we identified 80 repair protein candidates (Fig. 2B and Data S2). Strikingly, 72 out of 80 repair protein candidates were not previously reported to accumulate at the damage site in budding yeast (26, 27). The unreported repair protein candidates include uncharacterized proteins, such as Sap1 and Ypr089w (Data S2). Characterizing these proteins may expand our understanding of PM repair mechanisms.

We selected the screening hits from the C-terminally GFP-tagged library as targets for the laser damage assay. Although this is a reasonable approach to evaluating the proteins from the endogenous expression level, it overlooks the potential repair proteins in the N-terminally sfGFP-tagged library. Around 23% of ORFs only exist in the N-terminal library, and 11% of yeast ORFs show different localization from that of the C-terminal library (31). For example, sfGFP-Snc2, but not Snc2-GFP, changes localization in response to SDS treatment because of the mislocalization of Snc2-GFP in the vacuole. Because Snc2 is a homolog of Snc1, Snc2 is a potential repair protein. Screening hits from the N-terminally sfGFP-tagged library can also be a useful resource for further identification of repair proteins.

The screening is unable to identify proteins whose localization remains unaltered in response to SDS treatment. For example, we could not identify Cdc50-Drs2, which accumulates at the damage site induced by a laser (27). A recent study demonstrated the different cellular responses between focal (laser damage) and diffuse (streptolysin-O (SLO) treatment) PM damages (6). Some proteins may be overlooked in the screening due to the intrinsic differences between SDS treatment and the laser damage assay. In addition, the screening did not identify ESCRT proteins as hits. The tagging of fluorescent proteins occasionally interferes with the functions of the tagged proteins (48). Specifically, the exogenous expression of fluorescent-tagged ESCRT subunits can impair their functions (49, 50). The screening may have overlooked some of these proteins, possibly including ESCRT proteins.

We performed live-cell imaging of repair proteins and defined accumulation times of repair proteins (Fig. S2 and Data S2). This dataset provides an overview of repair protein accumulation. However, it should be noted that the accurate relative accumulation timing of repair proteins should be determined by multi-color imaging of repair proteins in the same cells, because of cell-cell variabilities, the GFP-tagging effects on the cells, and small sample size of the screening. These datasets will form the basis for future hypothesis-driven studies.

Coordination of polarized exocytosis and CME

We showed that polarized exocytosis and CME occur simultaneously at the damage site between approximately 5 and 35 min after laser damage (Fig. 3, C and D). Given that the resealing of the damaged PM generally occurs within 1 min (20), this result implies that polarized exocytosis and CME are involved in the restructuring of the damaged PM rather than the resealing of it. The coupling of endocytosis with polarized exocytosis is observed in multiple cell types, including at the growing tip of budding yeast, the growing tip of pollen tubes, and synapses in neurons (51). Polarized exocytosis and CME at the damage site may regulate the PM tensions and amount of lipids and PM proteins in a manner analogous to these cells.

Our results suggest that the activity of CME and polarized exocytosis at the damage site changes over time, with exocytosis predominating within 20 min after laser damage and CME predominating 20 min after laser damage (Fig. 3, C and D). This is consistent with the recent study in human cells (21). A previous study showed that polarized exocytosis activates CME in budding yeast (47). Rab GTPase Sec4, which regulates polarized exocytosis, also activates endocytosis by overriding Sla1 inhibition of Las17 (47). We could not identify Sec4 as a repair protein candidate because C-terminally GFP-tagged Sec4 is not in the library, probably due to the loss of function of C-terminally tagged Sec4. However, we found that Sec2-GFP, the guanine nucleotide exchange factor (GEF) of Sec4, accumulates at the damage site (Data S2). Sec4 and Sec2 are localized to the secretory vesicles (52, 53). Therefore, secretory vesicles that are targeted to the damage site may accumulate Sec2 and Sec4, leading to the activation of CME. The switching of activities between polarized exocytosis and CME may contribute to restoring PM homeostasis after the damage.

Polarized exocytosis at the damage site is inhibited in CME mutants (Fig. 4, A and B, and fig. S6, A and B). Given that CME and polarized exocytosis occur simultaneously (Fig. 3, C and D), CME may activate polarized exocytosis at the damage site, such as by regulating the PM tension (54) around the damage site. Moreover, Myo2-sfGFP and Exo70-mNG were partially retained at the bud tip after laser damage in CME mutants (Fig. 4, A and B, and fig. S6, A and B). These results raise the possibility that CME is involved in targeting Exo70 and Myo2 from the bud tip. At the bud tip, CME may be involved in the dispersion of Exo70 and Myo2 via upstream regulators, such as Rho3.

CME functions for PM repair in budding yeast

Previous studies showed that endocytosis actively occurs at the damage site to repair the damaged PM (11, 12, 20). Surprisingly, our results suggest that CME occurs not only at the damage site but at the bud tip within 10 min after laser damage. In rvs167Δ, in which exocytosis inhibition at the damage site is minimal (Fig. 4, B and C), repair proteins with TMDs at the bud tip fail to accumulate at the damage site (Fig. 5, B to F). These results are consistent with our idea that CME at the bud tip directs repair proteins with TMDs to the damage site (Fig. 7). Consistent with our results, dysferlin-containing vesicles increase in response to PM damage in muscle cells (25, 55). Dysferlin has a TMD and functions as a PM repair protein (25, 55, 56). Delivery of repair proteins with TMDs from non-damaged sites to the damage site by endocytosis may occur in a wide range of eukaryotic species.

We showed that 54-88% of the normalized mNG-Snc1 signal at the bud tip is recovered 50 min after the damage (Fig. 6, C and D). Because Snc1 is a CME cargo and it colocalizes with Ede1-mSc-I at the damage site, it is presumably recovered from the damage site via CME. Previous studies showed that macropinocytosis and CME restructure the damaged PM in human cells (20, 21). However, it is not clear whether the cargo proteins are recycled or degraded. We propose that CME terminates the PM damage responses by removing the repair proteins with TMDs from the damage site and resuming the growth by retargeting them to the bud tip (Fig. 7).

Here, we showed spatiotemporal cellular responses after PM damage in budding yeast by large-scale identification of repair proteins and their live-cell imaging. Despite the limitations mentioned above, our datasets provide the first functional catalog for PM repair proteins. Because some of the PM repair proteins identified in this study and CME mechanisms are evolutionarily conserved, this work may serve as a basis for future studies, including those to be conducted in mammalian cells.

Materials and Methods Media and strains

Standard procedures were used for DNA, E. coli, and yeast genetic manipulation. Yeast transformations were performed using the lithium acetate method. A PCR-based procedure was used for gene deletion. The deletion of the expected locus was confirmed by colony PCR (57). Yeast cells were cultured in YPD media (1% yeast extract, 2% bacto peptone, and 2% glucose) unless otherwise indicated. SD media (yeast nitrogen base (6.7 g/liter) without amino acids, L-adenine (550 mg/liter), L-arginine (280 mg/liter), L-alanine (280 mg/liter), L-asparagine (280 mg/liter), L-aspartic acids (280 mg/liter), L-cysteine (280 mg/liter), glycine (280 mg/liter), L– glutamic acids (280 mg/liter), L-glutamine (280 mg/liter), L-isoleucine (280 mg/liter), L-lysine (280 mg/liter), L-phenylalanine (280 mg/liter), L-proline (280 mg/liter), L-serine (280 mg/liter), L-threonine (280 mg/liter), L-tyrosine (280 mg/liter), L-valine (280 mg/liter), leucine (530 mg/liter), methionine (86 mg/liter), histidine (86 mg/liter), uracil (22 mg/liter), myo-inositol (100 mg/liter), and p-aminobenzoic acid (3 mg/liter) (pH 5.5)) were used for the laser damage assay. Yeast culture was performed at 25°C unless otherwise indicated. The yeast strains, yeast libraries, and plasmids used in this study are listed in Tables S1 to S3.

For the mNG-Snc1 expression experiments in Fig. 6, we grow cells in SD media containing 2% raffinose instead of glucose overnight. Then, we induced the expression of mNG-Snc1 by adding 3% galactose. After one hour of growth in galactose media, the cells were centrifuged at 5000xg for 5 min, and the spun-down cells were washed twice with SD media containing 2% glucose. We transferred the spun-down cells to SD media containing 2% glucose. The cells were incubated for at least 3 hours before the laser damage assay.

GFP screen for SDS damage response

GFP strains were spotted onto the YPD plates from 96-well plates using a pin replicator and incubated at 25°C. After 3 days, the colonies were inoculated into 200 μl of YPD media and incubated overnight to saturation. After mixing, 3 μl of the saturated culture was transferred to 1 ml of YPD media in deep well plates and incubated for 4-5 hours. 500 μl of the cultures were transferred to the new deep well plates. 10 μl of 1% SDS was added to 500 μl of the cultures and incubated at 25°C for 1 hour. The cells were centrifuged at 2000xg for 2 min. The cells were fixed by 300 μl of 4% PFA in YPD for 30 min at room temperature. The fixed cells were centrifuged at 2000xg for 2 min. Cells were washed twice with PBS and centrifuged at 2000xg for 2 min. The supernatant was removed between each wash. Cells were kept at 4°C until imaging. Cells were imaged with an LSM880 confocal microscope using a 20x air objective lens with an airyscan detector for GFP fluorescence (Ex 488/Em522). Maximum intensity projections of z-stack images are shown.

Categorization of localization changes in response to SDS treatment

We manually reviewed the acquired images using Zen Blue edition (Zeiss) or Fiji software (58) and compared the fluorescence signals of GFP-tagged proteins in normal and SDS treatment conditions. We identified the proteins whose fluorescence signal pattern changes between the two conditions. We categorized the screening hits based on reported protein localization (29). We first categorized the bud tip- or the bud neck-localizing proteins as “From bud tip/neck”. Among the uncategorized remaining screening hits, the actin-localizing proteins were categorized as “Actin”. Among the remaining uncategorized screening hits, cell periphery-localizing proteins were categorized as “From cell periphery.” Among the remaining uncategorized screening hits, nucleus-localizing proteins were categorized as “Nucleus” or “Nucleus to the cytoplasm”. We categorized Urc1 as “Nucleus to cytoplasm” because its signal dispersed to the cytoplasm in an SDS treatment condition. Other proteins in the “Nucleus” showed stronger nucleus localization in the SDS treatment conditions. Among the remaining uncategorized screening hits, we categorized mitochondria-localizing proteins as “Mitochondria”. Among the remaining uncategorized screening hits, we categorized spindle pole-localizing proteins as “Spindle pole.” All proteins in the “Spindle pole” showed stronger dot-like structures in the SDS treatment condition. Among the remaining uncategorized screening hits, we categorized proteins that show puncta or foci in SDS treatment conditions as “Puncta/foci.” We categorized Gcn2, Ato3, Sun4, Csi2, Sps4, and Ypr089w as “Puncta/foci dispersed” because they showed weaker foci or puncta signals in SDS treatment conditions. We categorized Tul1, Emp24, Hor7, Scw10, and Ynl019c as “From vacuole” because they showed decreased vacuole signal in the SDS treatment conditions. We categorized Prm5 as “to vacuole” because its vacuole signal increases. We categorized Yps1 and Msc1 as “cytoplasm to ER” because they localize to the ER, and their ER signal increased in SDS treatment conditions. In this study, we make use of the protein localization data from the Saccharomyces Genome Database (SGD) and (29). All results of this screening are shown in Data S1.

Laser damage assay

Yeast cells were grown in SD medium at 25°C until the culture reached an OD600 of 0.1-0.4. We diluted the yeast culture and further incubated the yeast cells in SD medium for 3-8 hours at 25°C. We took 1 mL of the culture and centrifuged it at 500xg for 5 min to spin the cells down. We took 5-10 μL of the cell suspension and placed it onto SD medium + 2.2% agarose bed. A Concanavalin A (ConA) (Nacalai Tesque) coated glass slip was placed on the cells prior to imaging.

Cells were observed with A1R (Nikon). A1R was equipped with a CFI Apochromat TIRF 60×/1.49 oil objective lens (Nikon). GFP, mNG, and sfGFP were excited by a 488 nm laser, and the fluorescence that passed a 525/50 nm band-pass filter was detected with a GaAsP detector. mSc-I was excited by a 561 nm laser, and the fluorescence that passed a 595/50 nm band-pass filter was detected with a GaAsP detector. For the laser damage assay, the 405 nm laser was irradiated to a circle of 0.5 μm diameter in a cell periphery. The laser power was set between 30% and 70%, depending on the fluorescence-tagged proteins. We determined the laser power sufficient for repair protein recruitment at the damage site in control cells without resulting in cell lysis during the experiments. At least ~80% of cells are viable during the experiments, otherwise indicated. Only the cells that survive after laser damage are quantified. To compare the accumulation of repair proteins between different strains, we set the exact same laser power and microscopy.

Categorization of localization changes in response to laser damage

We categorized proteins that accumulate at the damage site as “Damage site”. These proteins are defined as repair protein candidates. Among the remaining uncategorized proteins, we categorized cell periphery-localizing proteins as “PM to cytoplasm”. Among the remaining uncategorized proteins, we categorized Msn2 and Crz1 as “Nucleus” because they changed localization from the cytoplasm to the nucleus. Among the remaining uncategorized proteins, we categorized Dot6, Dna2, Msc1, and Gcd7 as “Puncta” because they formed punctate structures in response to laser damage. The repair protein candidates identified in this study are listed in Data S2.

Quantification of fluorescence signal for laser damage assay

The quantification of the fluorescence signal from the laser damage assay was performed as described previously (27). For the large-scale identification of repair protein candidates, we selected the region of interest (ROI) around the whole cell, the damage site, and the bud tip of the cells. We manually move the ROIs as the cell moves so that the ROIs remain at the same position in the cell. We set the ROIs and manually selected them for quantification. The fluorescence intensity at the damage site and the bud tip were normalized by the fluorescence intensity of a whole cell to minimize the effect of photobleaching during imaging.

We defined the accumulation time as the time when the fluorescence intensity at the damaged site became greater than the fluorescence intensity at the non-damaged site by three times the standard deviation (3xSD) for at least two consecutive time frames. We defined the dispersion time when the fluorescence intensity at the damage site becomes less than that of the non-damaged site plus 3xSD for at least two consecutive time frames after the accumulation time. We defined the retention time as the difference between the accumulation time and the dispersion time. We measured at least three cells. The median values of accumulation times and retention times across replicates for all repair proteins are listed in Data S2.

Growth screening of repair protein knockout mutants

Yeast cultures grown overnight in YPD at 25°C were diluted to OD600 = 0.1. The diluted cultures were spotted onto YPD, YPD + 0.01% SDS, and YPD + 25 μg/ml CFW plates. After 3 days of incubation at 25°C or 37°C (Heat stress), the growth of the YPD plate and other plates was compared. We performed the screening of two independent colonies. The strains were from the yeast deletion collection library (59) except for smi1Δ, las17Δ, skg6Δ, end3Δ, vrp1Δ, and sla2Δ. Only the mutants that showed sensitivity to the stress in two independent colonies were defined as sensitive to the stress. The results of the screening are shown in Data S2.

Spot assay

Yeast cultures grown overnight in YPD at 25°C were diluted to OD600 = 0.1. The four-fold serial dilutions of cultures were spotted onto the indicated plates.

Immunoblotting

Yeast cells in the early log phase (OD600 = 0.1-0.3) were treated with 0.02% SDS. 5 ml of yeast cells were collected before SDS treatment, 1 hr after SDS treatment, and 2 hr after SDS treatment. Cells were flash frozen by liquid nitrogen and stored at −80°C. Cells were resuspended in 120μl of cold lysis buffer (0.25 M NaOH and 1% β-mercaptoethanol) and incubated on ice for 10 min. 20 μl of trichloroacetic acid was added to the lysates. After 10 min of incubation on ice, the lysates were spun down, and the supernatant was discarded. The precipitates were washed with 500μl ice-cold acetone twice and dried at room temperature. The precipitates are resuspended in SDS polyacrylamide gel electrophoresis (PAGE) sample buffer [63 mM tris-Cl (pH 6.8), 2% SDS, 1% β-mercaptoethanol, 0.01% Bromophenol Blue, and 10% glycerol]. Immunoblotting was performed with anti-mini c-Myc (SANTA CRUZ, sc-40), anti-GFP (Roche Merck, 11814460001), or anti-tubulin α (Bio-Rad, MCA78G) antibodies.

Statistical Analysis

Dunnett’s multiple comparison test, Welch’s t-test, and Mann–Whitney U test were performed with Python software (v. 3.11). Gene enrichment analysis and Fisher’s exact test were performed with R software (v. 4.2.2).

Data and materials availability

All data needed to evaluate the conclusions in the paper are present in the main text and/or the supplementary materials.

Acknowledgements

We thank Dr. Maya Schuldiner (Weizmann Institute) for providing the N’ SWAT library. We thank Dr. S. Sugiyama for providing yeast strains and technical advice. We also thank the lab members of the Membranology unit for the discussion and critical reading of the manuscript. We are grateful for the help and support provided by the Imaging Section and Sequence Section of the Core Facilities at Okinawa Institute of Science and Technology Graduate University. We thank Dr. P. Barzaghi, Dr. K. Koizumi, Dr. S. Komoto, and Dr. T. Mochizuki for their technical assistance in microscopy experiments.

Additional information

Funding

This study was supported by MEXT/JSPS KAKENHI Grant Number 23KJ2138 to Y.Y., JSPS grant-in-aid for scientific research (B) 20H03440, 24K02233, and JST-COI-NEXT JPMJPF2205 to K.K..

Author contributions

Conceptualization, Y.Y. and K.K.; Investigation, Y.Y.; Writing – Original Draft, Y.Y.; Writing – Review & Editing, Y.Y. and K.K.; Funding Acquisition, Y.Y. and K.K.; Supervision, K.K..

Funding

Japan Society for the Promotion of Science (23KJ2138)

Japan Society for the Promotion of Science (20H03440)

Japan Society for the Promotion of Science (24K02233)

Japan Science and Technology Agency

https://doi.org/10.52926/jpmjpf2205

Additional files

Yamazaki_and_Kono_2025_Supplementary_materials

Data_S1_S2