Thermosensitivity of cellular translation restricts the growth of fission yeast at high temperatures Authors: Yutaka Akikusa, Atsuya Yamaguchi, Yutaka Funahashi, Amelia Cheng Wei Tan, Fontip Mahayot, Taisuke Naka, Ami Matsuo, Yukiko Nakase, Shingo Izawa, Kazuhiro Shiozaki, Yuichi Morozumi Published: 2026 ABSTRACT AbstractLiving organisms have thermal limits above which they are unable to operate and survive. Our previous genetic screen identified proteins that impede the high-temperature growth of fission yeast, including the RNA-binding protein Dri1 and a fission yeast-specific protein termed Rhs1. Here, we show that Dri1 and Rhs1 form a complex and physically interact with the Ccr4-Not complex, a master regulator of mRNA metabolism. Gene expression analysis revealed that the Dri1-Rhs1 and Ccr4-Not complexes negatively regulate a set of genes implicated in ribosome biogenesis (Ribi genes). Loss of the Dri1-Rhs1 complex results in the augmented expression of Ribi genes, thereby suppressing the translation defects and the growth inhibition under high-temperature conditions. The thermosensitivity of the translational processes may be a determinant of the upper limit of the growth temperature in fission yeast. Subject terms: Cell growth, Transcriptomics Introduction Elevated temperatures are one of the major stress conditions that can impair the growth and development of living organisms by causing protein denaturing and heat damage to other cellular macromolecules. To counteract and survive such stressful conditions, cells activate the heat shock response (HSR). This evolutionarily conserved mechanism reprograms gene expression and induces the synthesis of heat shock proteins (HSPs). Most HSPs are known to function as molecular chaperones, which maintain cellular proteostasis by refolding denatured proteins, preventing their aggregation, and sorting unfolded proteins for degradation1,2. Hsp104, an HSP that dissolves protein aggregates and facilitates protein refolding, is necessary for cell survival in response to acute heat stress in Saccharomyces cerevisiae and Candida albicans3–5. In contrast, budding yeast UBI4, a polyubiquitin-encoding gene inducible under high-temperature conditions, is dispensable for the acute heat stress response but is essential for adaptation to chronic, sublethal high temperatures6,7. It was also recently reported that tolerance to short-term and long-term heat stress in Arabidopsis thaliana requires distinct sets of factors8–10. These observations strongly suggest that cells employ different adaptive mechanisms depending on the intensity and duration of heat stress. While pro-survival mechanisms for acute heat shock, including the HSR, have been extensively studied, the molecular mechanisms that contribute to thermotolerance to chronic heat stress remain poorly understood. The optimal growth temperature of the fission yeast Schizosaccharomyces pombe is around 30 ˚C, and its growth is severely impaired at 38 ˚C and above. We previously found that S. pombe cells proliferate even at 39 ˚C when the activity of TOR complex 1 (TORC1), a kinase complex conserved among diverse eukaryotes, is suppressed11. Consistently, the null mutants of Sck1 and Mks1, both of which are the TORC1 substrates, exhibit significant cell growth at 39 ˚C. These results indicate that the TORC1 signaling pathway restricts the high-temperature growth of fission yeast cells. We also conducted genetic screens in S. pombe for additional negative regulators of cell proliferation at high temperatures and identified Dri1 and Rhs111. Dri1 is an RNA-binding protein implicated in heterochromatin assembly and kinesin loading to the mitotic spindle12,13. On the other hand, Rhs1 is a fission yeast-specific protein whose molecular function remains unknown. The high-temperature growth of a fission yeast strain lacking both the dri1 and rhs1 genes is comparable to that of the respective single mutants. Moreover, a physical interaction between Dri1 and Rhs1 was observed, suggesting that Dri1 and Rhs1 function as a complex in the negative regulation of high-temperature cell growth11. The Ccr4-Not complex is a eukaryotic protein complex that serves as a master regulator of mRNA metabolism, from transcription to degradation, thereby modulating gene expression14,15. The fission yeast Ccr4-Not complex contains seven highly conserved subunits, including the scaffold protein Not1 as well as the Ccr4 and Caf1 deadenylases that degrade the poly(A) tail of mRNA. The other subunits are Mot2, Not2, Not3, and Rcd1. Mot2 is a RING finger E3 ubiquitin ligase, and Not2 and Not3 contain the NOT box motif. Little is known about the function of the Rcd1 subunit within the complex. The null mutation of either caf1, ccr4, or mot2 causes growth defects in fission yeast16–18, indicating that the function of Ccr4-Not is important for normal cell growth. On the other hand, the involvement of this complex in the cellular adaptation to temperature changes has never been examined. In this study, we have demonstrated that the complex formation between Dri1 and Rhs1 is essential for their function, and its disassembly results in cell proliferation even at growth-inhibitory, high temperatures. A search for factors physically interacting with the Dri1-Rhs1 complex has identified the Ccr4-Not complex. Indeed, the null mutation of some genes encoding Ccr4-Not subunits confers cellular heat resistance, suggesting that the interacting two complexes, Dri1-Rhs1 and Ccr4-Not, restrict cell proliferation at high temperatures. Interestingly, genes implicated in ribosome biogenesis (Ribi genes) are up-regulated in strains lacking the Dri1-Rhs1 complex or the Ccr4-Not subunit Not3. The augmented expression of Ribi genes in those mutant strains ameliorates the translational defects and growth defects under high-temperature conditions. These results have shed light on protein translation as a thermosensitive process that affects the upper-temperature limit for cell growth. Results Residues 31-50 of Rhs1 mediate its interaction with Dri1 We previously identified fission yeast Dri1 and Rhs1 as suppressors of cell growth at high temperatures; the dri1 and rhs1 null mutants can proliferate at 39 ˚C, where wild-type cells arrest their growth and lose viability11. Although physical interaction between Dri1 and Rhs1 was detected, it remains to be determined whether they function as a protein complex in restricting high-temperature growth. We, therefore, set out to identify the amino acid sequence within Rhs1 required for the interaction with Dri1 by yeast two-hybrid assays. Although autoactivation was observed when Rhs1 was used as bait in the assay, the Dri1-Rhs1 interaction was successfully detected by utilizing Dri1 and Rhs1 as bait and prey, respectively (Supplementary Fig. 1). Dri1 interacted with the C-terminally truncated fragment of Rhs1 (1-240) but not the N-terminally truncated one (121-368) (Fig. 1a). Consistently, the N-terminal 120 amino residues of Rhs1 (1-120) were sufficient for the interaction with Dri1. Moreover, these observations were confirmed by immunoprecipitation experiments with fission yeast strains expressing a series of truncated Rhs1; Dri1 was co-purified with the FLAG-tagged Rhs1 fragments containing residues 1-120 (Fig. 1b). Thus, the N-terminal 120 residues of Rhs1 include the Dri1-binding region. We also noticed that, in the anti-FLAG blotting, multiple bands are detected in the strain expressing Rhs1-FLAG (Fig. 1b). Because no such bands were observed in cells expressing untagged Rhs1, these bands appear to represent the degradation products of Rhs1-FLAG. Consistently, the FLAG-tagged Rhs1 protein was specifically detected as a single band when cell lysates were prepared under denaturing conditions using TCA (Supplementary Fig. 2), indicating that Rhs1 is highly susceptible to degradation under non-denaturing conditions. The residues 31-50 of Rhs1 are indispensable for the interaction with Dri1. a Summary of the interaction between Dri1 and various Rhs1 fragments in yeast two-hybrid assays. Interaction was assessed by ADE2 and HIS3 reporter gene expression in the budding yeast Y2HGold strain as shown in Supplementary Fig. 1. +, interaction; -, no interaction. An orange box indicates the Rhs1 region required for the Dri1-Rhs1 interaction. b, c Physical interaction between Dri1 and the various truncated Rhs1 fragments. dri1:myc cells expressing various FLAG-tagged Rhs1 fragments were cultured in YES medium at 30 ˚C. Cell lysates were subjected to immunoprecipitation using anti-FLAG antibody (IP: FLAG), and co-purified Dri1-myc was analyzed by immunoblotting. Asterisks indicate immunoglobulin bands recognized by the anti-FLAG antibody. d Predicted structure of Rhs1 (light blue) in complex with Dri1 (light orange) generated by AlphaFold 3. The RRM and RanBP2-type zinc finger of Dri1 are shown in pale green and light pink, respectively. Disordered regions of Dri1 (364-596) and Rhs1 (96-368), whose structures could not be reliably predicted, are not shown. A dashed oval (red) indicates the region of Rhs1 (31-50) required for the interaction with Dri1, as determined by yeast two-hybrid assays and immunoprecipitation experiments. e A schematic diagram of the Dri1 fragments. RRM, RNA recognition motif; ZF1-3, RanBP2-type zinc fingers. f Physical interaction between Rhs1 and various truncated Dri1 fragments. rhs1:FLAG cells expressing various GFP-tagged Dri1 fragments were cultured in YES medium at 30˚ C. Cell lysates were subjected to immunoprecipitation as in (b). Asterisks indicate immunoglobulin heavy chains. We repeated the yeast two-hybrid assay to further narrow down the Dri1-binding region. The Dri1 interaction was detected with the Rhs1 N-terminus of 50 residues or longer (Fig. 1a). In addition, while the Rhs1 fragment lacking the N-terminal 30 residues (31-368) interacted with Dri1, further truncation of N-terminal 40 residues or more abrogated the association in the yeast two-hybrid and immunoprecipitation assays (Fig. 1c). Indeed, the Rhs1 fragment of only residues 31-50 was sufficient for the interaction with Dri1 (Fig. 1a). Notably, in a structure prediction of the Dri1-Rhs1 complex by AlphaFold 319 (Fig. 1d), this region of Rhs1 forms an α-helix and is inserted into a groove formed by the N-terminal region and RNA recognition motif (RRM) of Dri1 (see below). Together, these results strongly suggest that residues 31-50 of Rhs1 contain the sequence to bind Dri1. Dri1 harbors a single RRM and three repeats of the RanBP2-type zinc finger domain (ZF1-3), both of which are known as RNA-binding domains (Fig. 1e). As residues 31–50 of Rhs1 are predicted to bind the groove formed between the N-terminal region and RRM of Dri1 (Fig. 1d), we next tested this interaction model by immunoprecipitation using strains expressing a series of truncated Dri1 proteins (Fig. 1f). Like full-length Dri1, the Dri1 fragment lacking the two zinc fingers from the C-terminus (1-402) was co-purified with Rhs1 (Fig. 1f). The Rhs1 interaction was partially compromised with Dri1(101-596), which lacks the N-terminal 100 residues that are not part of the groove in the predicted structure (Fig. 1d). On the other hand, the N-terminal fragment without RRM (1-217) and Dri1(218-596) lacking the entire N-terminal region failed to bind Rhs1, confirming the contribution of both the N-terminal region and RRM to the Dri1-Rhs1 interaction. However, the Dri1(1-322) fragment, consisting only of the N-terminal domain and RRM, failed to bind to Rhs1 (Fig. 1f). Although this observation could imply a contribution of the ZF1 domain to the interaction with Rhs1, it is also likely that such a minimal fragment might not be properly folded to form the Rhs1-binding groove. Formation of the Dri1-Rhs1 complex is essential for restricting high-temperature growth As shown above, the Dri1-binding region within Rhs1 was successfully narrowed down to its residues 31-50 (Fig. 1a). To identify the amino acid residues contributing to Dri1 binding, a series of Rhs1 mutants were constructed by replacing two consecutive residues in this Rhs1 sequence with alanine (Fig. 2a). Among the Rhs1 mutant proteins constructed, only Rhs1-AF33.34AA and -QF37.38AA failed to interact with Dri1 in the yeast two-hybrid assay (Fig. 2a). Consistently, the amount of co-purified Dri1 was significantly decreased when FLAG-tagged Rhs1-FQF34.37.38AAA, -FQ34.37AA, and -FF34.38AA were immunoprecipitated (Fig. 2b). These results were further assessed by constructing the alanine mutants of the individual residues. Single substitutions at Phe-34 and Gln-37 in Rhs1 drastically abrogated the copurification of Dri1, while that of Phe-38 only partially compromised the Dri1-Rhs1 interaction (Fig. 2b). It was noted that the protein level of Rhs1-FLAG is decreased in the rhs1 mutant strains where the amount of co-precipitated Dri1-myc was reduced. In contrast, comparable amounts of FLAG-tagged Rhs1 were detected in cell lysates prepared under denaturing conditions (Supplementary Fig. 3). Therefore, the Phe-34 and Gln-37 of Rhs1 appear to be important residues for the interaction with Dri1, and disruption of the complex formation may cause destabilization and degradation of Rhs1 under non-denaturing conditions. Formation of the Dri1-Rhs1 complex is indispensable for its function in the negative regulation of high-temperature growth. a Summary of the interaction between Dri1 and the Rhs1 mutant proteins in yeast two-hybrid assays. Interaction between Dri1 and the Rhs1 mutants carrying alanine substitutions in residues 31-50 was examined as in Supplementary Fig. 1. The amino acid sequence of Rhs1(31-50) is shown. +, interaction; -, no interaction. b Phe-34 and Gln-37 of Rhs1 are important for its interaction with Dri1. dri1:myc cells expressing the indicated alanine mutants of Rhs1 were cultured in YES medium at 30 ˚C. Cell lysates were subjected to immunoprecipitation as in Fig. 1b. Asterisk indicates immunoglobulin heavy chains recognized by the anti-FLAG antibody. c The F34A and Q37A substitutions in Rhs1 confer heat resistance to fission yeast. The indicated rhs1 mutant strains were grown in YES medium at 30 ˚C, and their growth was tested at the indicated temperatures by spotting serial culture dilutions on YES agar plates. d A schematic representation of the Dri1-Rhs1 complex formation mediated by the GFP-GBP interaction. e Dri1 and Rhs1 function as a complex for suppressing cell growth at high temperatures. The indicated strains were grown in YES medium at 30˚C, and their growth was tested as in (c). The functional significance of the Dri1-Rhs1 complex formation was evaluated by the high-temperature growth assay of strains expressing the alanine-substitution mutants of Rhs1 described above. The rhs1 mutants carrying F34A and/or Q37A, both of which abrogate the Dri1-Rhs1 interaction (Fig. 2b), exhibited heat-resistant growth similar to that of rhs1∆ (Fig. 2c). On the other hand, a strain expressing Rhs1-F38A, which still interacts with Dri1 (Fig. 2b), failed to grow at 39 ˚C, like the wild type. Thus, the loss of the Dri1-Rhs1 association results in heat resistance, as observed with the dri1∆ and rhs1∆ mutants. Moreover, we found that the rhs1F34A and rhs1Q37A defects were suppressed by artificially tethering the Rhs1 mutant proteins to Dri1 through the GFP-GFP binding protein (GBP) targeting system20 (Fig. 2d). When Rhs1 with F34A or Q37A mutations was expressed as GBP-fusions together with GFP-fused Dri1, the cells exhibited the wild-type phenotype and were sensitive to high temperatures (Fig. 2e). Together, these results strongly suggest that the association between Dri1 and Rhs1 is indispensable for their functionality in restricting high-temperature cell growth. Dri1, but not Rhs1, forms cytoplasmic aggregates upon high-temperature stress We examined the cellular localization of Dri1 and Rhs1 by fluorescence microscopy of strains expressing these proteins tagged with GFP from their chromosomal loci. Dri1 was found to be distributed throughout the cytoplasm, with a small, dot-like signal also detected within the nucleus (Fig. 3a). Interestingly, the rhs1∆ mutation resulted in the distribution of Dri1 throughout the nucleus and cytoplasm, and brighter, multiple Dri1 puncta were detected in the nucleus (Fig. 3a). Dri1 and Rhs1 regulate their cellular localization by one another. a, b The indicated strains expressing GFP-tagged Dri1 or Rhs1 from their chromosomal locus were grown in EMM at 30 °C, and then shifted to 39 ˚C. After 4 h incubation at 39 °C, cells were analyzed by fluorescence microscopy. Nuclei were visualized by Hoechst staining. Z-stack images were collected and mid-section images after deconvolution are shown. BF bright-field image. Scale bars, 5 µm. As shown in Fig. 3b, Rhs1-GFP was diffused in the cytoplasm but excluded from the nucleus; unlike Dri1-GFP, no nuclear puncta were observed. In the dri1∆ background, some diffused Rhs1 signals were detectable in the nucleus. One possible interpretation for these observations is that Dri1 and Rhs1 are excluded from the nucleus once they are assembled into a complex; however, a fraction of Dri1 forms nuclear foci independently of Rhs1. The cellular localization of Dri1 and Rhs1 was also examined under high-temperature conditions. After a temperature shift to 39 ˚C, Dri1 formed cytoplasmic aggregates both in the rhs1+ and rhs1∆ backgrounds (Fig. 3a). A recent study13 reported that Dri1 is incorporated into Protein Aggregate Centers (PACs), cytoplasmic protein aggregates believed to sequester misfolded proteins upon mild heat stress to protect them from degradation21,22. Consistently, we confirmed the co-localization of cytoplasmic Dri1 aggregates with Hsp16, a known component of PACs21 (Supplementary Fig. 4). On the other hand, unlike Dri1, Rhs1 aggregation was not observed even at 39 ˚C both in the wild-type and dri1∆ strains (Fig. 3b). Thus, it appears that, although Dri1 and Rhs1 function as a complex, only Dri1 is assembled into PACs independently of Rhs1 upon heat stress. The Dri1-Rhs1 complex interacts with Ccr4-Not, a eukaryotic protein complex regulating mRNA metabolism Aiming to identify proteins physically interacting with the Dri1-Rhs1 complex, its immunoprecipitation was performed using a strain that expresses Dri1 and Rhs1 from their chromosomal loci as fusions with the myc and FLAG epitope tags, respectively. A tandem affinity purification of the Dri1-Rhs1 complex, followed by mass spectrometry, identified the components of the Ccr4-Not complex, Not1, Not3, and Mot223 (see Introduction), as well as Sum3, a member of the DEAD box RNA helicase family24 (Fig. 4a, b, Supplementary Table 1). Indeed, immunoprecipitation of Dri1-myc or Rhs1-FLAG copurified GFP-tagged Not1 and Not3 (Fig. 4c–f), confirming the physical interaction between the Dri1-Rhs1 complex and Ccr4-Not. A previous mass-spectrometry study also reported Dri1 and Rhs1 as co-purified proteins with Mmi1, which physically associates with Ccr4-Not16. Ccr4-Not physically interacts with the Dri1-Rhs1 complex. a Affinity purification of the Dri1-Rhs1 complex from S. pombe cells. The complex was purified from the cell lysate of the dri1:myc rhs1:FLAG strain by sequential immunoprecipitation using anti-FLAG and anti-myc antibodies. The recovered samples were separated by SDS-PAGE followed by silver staining. A wild-type strain was used as a negative control (No tag). b Mass spectrometric analysis of proteins co-purified with the Dri1-Rhs1 complex shown in (a). c–f Physical interaction between the Dri1-Rhs1 complex and either Not1 or Not3. dri1:myc (c, e) or rhs1:FLAG (d, f) cells expressing GFP-tagged Not1 (c, d) or Not3 (e, f) were cultured in YES medium at 30 ˚C. Cell lysates were subjected to immunoprecipitation using anti-myc (IP: myc) or anti-FLAG antibody (IP: FLAG), and co-purified Not1-GFP and Not3-GFP were analyzed by immunoblotting. g, h Interaction between the Dri1-Rhs1 complex and Not1 at high temperatures. The dri1:myc (g) or rhs1:FLAG (h) strain expressing Not1-GFP was cultured in YES medium at 30 ˚C and shifted to 39 ˚C. After 4 h incubation at 39 ˚C, cell lysates were recovered and subjected to immunoprecipitation as in (c, d). i, j Formation of the Dri1-Rhs1 complex is required for stable interaction with Ccr4-Not. Physical interaction between Not1-GFP and Dri1-myc (i) or Rhs1-FLAG (j) in the rhs1∆ (i) or dri1∆ (j) background was analyzed as in (c, d). k, l Physical interaction between the Dri1-Rhs1 complex and Not1 in the absence of Ccr4-Not subunits. The indicated null mutants of the Ccr4-Not subunits expressing Not1-GFP and either Dri1-myc (k) or Rhs1-FLAG (l) were cultured in YES medium at 30 ˚C. Cell lysates were subjected to immunoprecipitation as in (c, d). Band intensities of Not1-GFP in IP fractions were quantified using ImageJ and normalized to those in input fractions, and relative Not1 levels bound to Dri1 (k) or Rhs1 (l) are shown. Data represent the mean ± SD (n = 3 independent experiments). *P < 0.05; **P < 0.01; ***P < 0.001; n.s. not significant, compared to the wild-type control (WT) using a two-tailed Student’s t-test. Asterisks in (d, f, h, j, l) indicate immunoglobulin heavy chains recognized by the anti-FLAG antibody. Further immunoprecipitation experiments found that the association between the Dri1-Rhs1 complex and Ccr4-Not is not significantly affected by the temperature shift from 30 ˚C to 39 ˚C (Fig. 4g, h). On the other hand, the interaction between Dri1 and Not1 was notably compromised by the rhs1∆ mutation (Fig. 4i). Similarly, the Rhs1-Not1 interaction was reduced in the dri1∆ background (Fig. 4j), indicating that the formation of the Dri1-Rhs1 complex promotes its interaction with Ccr4-Not. The association of the Dri1-Rhs1 and Ccr4-Not complexes was also examined in a series of mutants lacking individual subunits of the Ccr4-Not complex. The Not1-GFP signal in the input fraction of the mot2∆ mutant was higher than that in the other mutants, suggesting increased Not1 expression in this strain (Fig. 4k, l). Importantly, quantitative analysis revealed that all tested null mutants except rcd1∆ exhibited impaired interaction of Not1-GFP with Dri1-myc and Rhs1-FLAG (Fig. 4k, l). Thus, most subunits of Ccr4-Not appear to contribute to the interaction between Ccr4-Not and the Dri1-Rhs1 complex. The Ccr4-Not subunits contribute to suppressing high-temperature cell growth Because of its interaction with the Dri1-Rhs1 complex, which suppresses cellular heat resistance, Ccr4-Not was further characterized for its physiological role under high-temperature conditions. The mot2∆, caf1∆, ccr4∆, and rcd1∆ mutants showed compromised growth even at temperatures that allow wild-type cells to grow (Fig. 5a–d). On the other hand, the not2∆ mutant strain exhibited better growth than the wild-type strain at 38 ˚C (Fig. 5e). In addition, the not3∆ mutation significantly improved the cellular heat resistance (Fig. 5f), implying that, like the Dri1-Rhs1 complex, Ccr4-Not contributes to restricting high-temperature growth of fission yeast. To examine the functional relationships between the Dri1-Rhs1 complex and Not3, double mutants of these proteins were constructed for epistasis analyses. The high-temperature growth of the dri1∆ not3∆ and rhs1∆ not3∆ double mutants was no better than that of the dri1∆ and rhs1∆ single mutants (Fig. 5f); therefore, the heat resistance conferred by the not3∆ mutation is not additive to that by the dri1∆ and rhs1∆ mutations. The growth of the dri1∆ not3∆ and rhs1∆ not3∆ double mutants was slightly slower than that of the dri1∆ and rhs1∆ single mutants at any temperature, an observation attributable to the modest growth retardation caused by the not3∆ mutation (Fig. 5g). Ccr4-Not suppresses cell growth at high temperatures. a–f The indicated mutant strains lacking the components of Ccr4-Not were grown in YES medium at 30 ˚C, and their growth at the indicated temperatures was tested by spotting serial culture dilutions on YES agar plates. g The not3∆ mutation impaired cell proliferation at 30 ˚C. Wild-type and not3∆ cells were grown in YES liquid medium at 30 ˚C. At 0 hour, cultures were adjusted to OD600 = 0.115 ± 0.005, and cell growth at 30 °C was monitored by OD600 measurement at the indicated time points. Data are presented as means ± SD (n = 3 independent experiments). **P < 0.01; ***P < 0.001, compared to the wild-type control (WT) using a two-tailed Student’s t-test. Together, these results suggest that the Dri1-Rhs1 and Ccr4-Not complexes may cooperate in suppressing cell growth at high temperatures. Among the Ccr4-Not subunits, Not2 and Not3 appear to be particularly important for this cellular function, as their absence compromises the interaction between the two complexes (Fig. 4k, l) and improves cellular heat resistance (Fig. 5e, f). Dri1-Rhs1 and Ccr4-Not suppress the expression of ribosome biogenesis genes Ccr4-Not controls gene expression by regulating various steps of mRNA metabolism, from synthesis in the nucleus to degradation in the cytoplasm14,15. In addition, Dri1 contains two types of RNA-binding domains, RRM and RanBP2-type zinc finger motifs. Therefore, we hypothesized that the Dri1-Rhs1 complex suppresses cellular growth at high temperatures by modulating gene expression together with Ccr4-Not. We performed RNA-sequencing (RNA-seq) on two biological replicates of the wild-type dri1∆, and rhs1∆ strains cultured at 30˚C to identify genes whose expression is regulated by the Dri1-Rhs1 complex. The quality control (QC) metrics summarized in Supplementary Table 2 confirmed the high quality of the RNA-seq data, with strong correlations between gene expression levels in the two biological replicates for each strain (R²: WT = 0.993, dri1∆ = 0.989, rhs1∆ = 0.992) (Supplementary Fig. 5). Differentially expressed gene (DEG) analysis revealed that 101 and 67 genes are up-regulated, and 45 and 15 genes are down-regulated in the dri1∆ and rhs1∆ mutants, respectively (Fig. 6a, and Supplementary Data 1, 2). Heat shock proteins (HSPs), which are molecular chaperones important for cellular thermotolerance, were not among the genes that were up-regulated in the mutants. Moreover, heat-induced accumulation of Hsp16, a member of the HSP family, expressed as a GFP fusion from its chromosomal locus was not significantly affected by the dri1∆ or rhs1∆ mutations (Supplementary Fig. 6). Thus, the high-temperature growth of those mutant strains does not appear to be due to the enhanced expression of HSPs. The Dri1-Rhs1 complex negatively regulates the expression of Ribi genes. a Venn diagram depicting the overlap of the up- and down-regulated genes between the dri1∆ and rhs1∆ mutants. Wild-type, dri1∆ and rhs1∆ cells were grown to exponential phase at 30 ˚C, then harvested for RNA-seq experiments. The up-regulated and down-regulated genes in the dri1∆ and rhs1∆ mutants, relative to the wild-type strain, were determined using DESeq2 with thresholds of |log2FC | > 0.8 and adjusted p-value < 0.05. Two biological replicates were used for each strain. b Volcano plots representing differentially expressed genes in the dri1∆ and rhs1∆ mutants compared to wild-type cells. Ribi and RP genes are shown in orange and blue, respectively. Significantly differentially expressed Ribi genes are separately illustrated in red. c Box plots showing fold change in the Ribi and RP genes in the dri1∆ and rhs1∆ mutants. The center line represents the median; box limits indicate the first and third quartiles; whiskers extend to 1.5 × the interquartile range. Individual points represent genes in each category. d Cells were cultured in YES medium at 30 ˚C and shifted to 39 ˚C. After 16 hours at 39˚C, cells were harvested and total RNA was extracted. Expression levels of representative Ribi genes (imp4+, mrd1+, cdh1+ and icp5+) in the indicated strains were measured by RT-qPCR relative to wild-type cells. Data represent the mean ± SD (n = 3 independent experiments). *P < 0.05; **P < 0.01; ***P < 0.001, compared to the wild-type control using a two-tailed Student’s t-test. e High-temperature growth of fission yeast in the presence of a ribosome biogenesis inhibitor. The indicated mutant strains were grown in YES medium at 30 ˚C, and their growth in the presence of Rbin-1 (1 µM) was tested at 30 ˚C and 39 ˚C by spotting serial culture dilutions on YES agar plates. f Polysome profiles of fission yeast cells under high temperature conditions. The indicated cells were grown in EMM liquid medium at 30 ˚C and shifted to 39 ˚C. After 24 h at 39 ˚C, cells were harvested for polysome analysis. The polysome-to-monosome (P/M) peak ratios shown in each profile represent mean ± SD (n = 3 independent experiments). p-values calculated using a two-tailed Student’s t-test with the wild-type control (WT) as the reference are also shown. g In vivo puromycin incorporation at high temperatures. The indicated cells were grown in YES liquid medium at 30 ˚C and shifted to 39 ˚C. After 24 h at 39 ˚C, cells were treated with puromycin and their cell lysates were subjected to immunoblotting. Immunoblotting results using the anti-puromycin antibody are shown at different exposure times. As Dri1 and Rhs1 function as a complex in restricting cellular heat resistance, we focused on the 59 up-regulated and 11 down-regulated genes shared between the dri1∆ and rhs1∆ mutants (Fig. 6a) and conducted gene ontology (GO) analysis for biological processes. While no significant enrichment of GO terms was observed in the down-regulated genes, the most up-regulated genes in the dri1∆ and rhs1∆ mutants were classified as ribosome biogenesis (Ribi) genes, which encode proteins that assist in rRNA maturation and ribosome assembly (Supplementary Fig. 7). The volcano and box plots further illustrate the significant up-regulation of Ribi genes but not ribosomal protein (RP) genes encoding ribosomal subunits (Fig. 6b, c). We performed RT-qPCR analysis and confirmed the increased expression of the Ribi genes imp4+, mrd1+, cdh1+, and icp5+ in the dri1∆ and rhs1∆ mutants not only at 30 ˚C but also at 39 ˚C (Fig. 6d). Moreover, a similar up-regulation of the Ribi genes was observed in the not3∆ mutant (Fig. 6d). Together, these results suggest that the Dri1-Rhs1 complex negatively regulates Ribi gene expression in conjunction with Ccr4-Not irrespective of growth temperatures. As the physical interaction between the Dri1-Rhs1 complex and Ccr4-Not was also detected (Fig. 4), the Dri1-Rhs1 complex may recruit Ccr4-Not to Ribi gene mRNAs for their degradation (see Discussion). Next, we examined whether the augmented expression of Ribi genes in the dri1∆ and rhs1∆ mutants indeed contributes to their heat-resistant growth, using the ribosome biogenesis inhibitor Rbin-125. The growth of fission yeast cells at 30 ˚C was not significantly affected by Rbin-1 at the concentration tested (Fig. 6e). On the other hand, the compound impaired the heat resistance conferred by the deletion of the sck1+ and mks1+ genes, whose products restrict high-temperature growth independently of Dri1 and Rhs111. Thus, ribosome biogenesis appears to be a critical cellular process for growth at elevated temperatures. Interestingly, the growth of dri1∆ and rhs1∆ cells at high temperatures was only marginally affected by Rbin-1 (Fig. 6e). The limited sensitivity of these mutants to Rbin-1 may be due to their induced expression of Ribi genes, which is likely to counteract the action of the drug. Having found that the Dri1-Rhs1 and Ccr4-Not complexes modulate Ribi gene expression, we next performed polysome analysis to compare the translational process in the wild-type, dri1∆, and rhs1∆ strains. The polysome profile in wild-type cells was comparable to that in the dri1∆ and rhs1∆ mutants at 30 ˚C (Fig. 6f). After a temperature shift of wild-type cells to 39 ˚C, the 80S monosome peak was significantly elevated while polysome peaks were lessened, resulting in a decreased polysome-to-monosome (P/M) peak ratio. On the other hand, such a change in the P/M ratio was not apparent in the dri1∆ and rhs1∆ mutants upon the temperature shift (Fig. 6f). Consistent results were also obtained by puromycin labeling of nascent polypeptides, which is indicative of global protein synthesis26,27. As shown in Fig. 6g, the levels of puromycin-labeled polypeptides at 39 ˚C were significantly decreased compared to those at 30 ˚C in wild-type cells, suggesting defects in protein translation at high temperatures. Notably, this compromised translation at high temperatures was partially rescued by the dri1∆ and rhs1∆ mutations. Thus, these mutants likely maintain more translational activities than wild-type cells under high-temperature conditions. Taken together, these results suggest that the elevated expression of Ribi genes in these mutants complements translational defects, thereby enabling fission yeast cells to proliferate at high temperatures. Discussion Recent climate warming has increased the risk of living organisms being exposed to high temperatures that impact their growth and survival28. Our study, which aimed to discover the molecular mechanisms that determine cellular heat resistance, previously found that a fission yeast strain lacking either Dri1 or Rhs1 can proliferate at high temperatures inhibitory to the growth of wild-type cells11. In this study, we have delved into the cellular functions of these proteins to understand how they regulate the upper thermal limit of cell growth in fission yeast. As shown in Fig. 2, the formation of the Dri1-Rhs1 complex is essential for restricting cell growth at high temperatures. It was previously reported that Dri1 physically interacts with Dpb4, a subunit of DNA polymerase epsilon, and regulates heterochromatin assembly in fission yeast12. In our fluorescence microscopy analysis (Fig. 3), a single, dot-like signal of Dri1 was observed within the nucleus. Although the specific nuclear region where Dri1 localizes remains unclear, this punctate signal may represent a fraction of Dri1 involved in heterochromatin assembly. On the other hand, most of the Dri1 protein appears diffused throughout the cytoplasm, as observed also with Rhs1. Thus, Dri1 is likely to regulate the formation of heterochromatin independently of Rhs1, and this function does not appear to contribute to the suppression of the high-temperature growth of S. pombe. A previous study showed that Dri1 is exported from the nucleus by the mRNA export pathway13. In the rhs1∆ mutant, Dri1 is detectable in the nucleoplasm (Fig. 3a). This result implies that Dri1 bound to Rhs1 is preferentially retained in the cytoplasm, highlighting a potential role of Rhs1 in the cytoplasmic anchoring of Dri1. While Dri1 forms a complex with Rhs1, only Dri1 is incorporated into PACs, cytoplasmic aggregates observed upon mild heat stress21,22. Thus, the Dri1-Rhs1 association may not be very stable, allowing Dri1 to function with another partner protein, Dpb4, in the nucleus. The significance of such dual functionality of Dri1 remains unknown. The affinity purification of the Dri1-Rhs1 complex from fission yeast cell lysate led to the identification of Ccr4-Not as its interacting factor (Fig. 4). In addition, our gene expression analysis revealed that Dri1-Rhs1 as well as Ccr4-Not negatively regulates the expression of Ribi genes (Fig. 6). Ccr4-Not is known to control gene expression by regulating mRNA metabolism at various steps; it is involved in mRNA decay by degrading the poly(A) tail at the 3’ end of mRNA in the cytoplasm, while also regulating transcription of mRNA and its export in the nucleus15,29. As the Dri1-Rhs1 complex was localized in the cytoplasm (Fig. 3), we speculate that this complex downregulates the mRNA levels of Ribi genes by facilitating Ccr4-Not-dependent deadenylation (Fig. 7). The recruitment of Ccr4-Not to particular mRNA targets for degradation is thought to be mediated by RNA-binding proteins that recognize specific sequence elements, predominantly in the untranslated regions30,31. For instance, human tristetraprolin (TTP), an RNA-binding protein that regulates the inflammatory response, binds to AU-rich elements in target mRNAs and recruits the Ccr4-Not complex for mRNA turnover32,33. Dri1 contains the RNA-binding domains, and thus, the Dri1-Rhs1 complex may mediate the binding of Ccr4-Not to the mRNAs of Ribi genes (Fig. 7). Furthermore, the physical interaction between the Dri1-Rhs1 complex and Ccr4-Not was significantly compromised by the not3∆ mutation (Fig. 4k, l), suggesting that Not3 contributes to the physical interaction of Ccr4-Not with the Dri1-Rhs1 complex. A previous study revealed that, in Drosophila, the RNA-binding protein Bicaudal-C (Bic-C), which is required for embryo patterning, directly binds to Not3/5, an ortholog of the fission yeast Not3, and recruits the Ccr4-Not complex to the Bic-C mRNA, promoting its deadenylation34. Therefore, the molecular function of Not3 as a module mediating the association of Ccr4-Not with RNA-binding proteins may be evolutionarily conserved across eukaryotes. Proposed mechanism underlying high-temperature growth in the dri1∆ and rhs1∆ mutants. In wild-type cells, the Dri1-Rhs1 complex binds Ribi gene mRNAs and recruits Ccr4-Not for their degradation. This negative regulation may lead to insufficient ribosome biogenesis, thereby limiting protein translation required to sustain cell growth under high-temperature conditions. On the other hand, Ccr4-Not fails to target Ribi gene mRNAs in the dri1∆ and rhs1∆ mutants, resulting in increased expression of Ribi genes. This up-regulation, in turn, enhances ribosome biogenesis and allows sufficient protein translation to sustain growth at high temperatures. Our polysome analysis found the accumulation of 80S monosome accompanied by a reduction in polysomes at high temperatures in wild-type cells (Fig. 6f). Consistent with this, global protein synthesis was significantly decreased under such conditions (Fig. 6g). On the other hand, in the dri1∆ and rhs1∆ strains, the polysome peaks appear to be largely unaffected by heat stress, though protein translation was impaired. One possible explanation for these observations is that the polysome peaks in these mutants include half-mer polysomes, which are indicative of stalled 40S subunits at the initiation stage caused by a limiting supply of 60S subunits35. Indeed, the first polysome peak immediately following the 80S monosome peak was less pronounced at high temperatures compared with that at 30 ˚C in these mutants, suggesting the accumulation of partially assembled ribosomal particles. Thus, the augmented expression of Ribi genes by the dri1∆ and rhs1∆ mutations may only partially improve ribosome assembly, leading to the accumulation of half-mer polysomes and a modest restoration in protein synthesis at high temperatures (Fig. 6g). However, because the dri1∆ and rhs1∆ mutants are able to grow at high temperatures, the partial rescue of translational defects appears to be sufficient to sustain cell proliferation under such conditions (Fig. 7). Given the inherent sensitivity of proteins to heat, maintaining such a minimal level of protein synthesis may be beneficial to mitigate proteotoxic stress at elevated temperatures. We have demonstrated that the Dri1-Rhs1 complex negatively regulates Ribi gene expression in S. pombe. While orthologs of Rhs1 are found only in fission yeast species, Dri1 is highly conserved among eukaryotes from yeast to humans. Interestingly, the expression of budding yeast NRP1, a gene encoding Dri1 ortholog, is co-regulated with genes involved in the ribosome biosynthesis pathway36. Moreover, a previous study reported that the RanBP2-type zinc finger of TEX13A, a human ortholog of Dri1, can bind single-stranded RNA containing a GGU motif37. It is of great interest to understand whether those Dri1 orthologs regulate the expression of Ribi genes, as observed in fission yeast. Another intriguing question is whether the thermosensitivity of the translation process is also a determinant of the upper-temperature limit for cell growth in other eukaryotic species. Further studies in diverse eukaryotes, including functional characterization of Dri1 orthologs, will provide a better understanding of the relationship between translation and cellular thermosensitivity. Methods Fission yeast strains and general methods S. pombe strains used in this study are listed in Supplementary Table 3. Cells were grown in YES or EMM medium containing 3% or 1% glucose, respectively, at 30 ˚C unless indicated otherwise 38,39. For strain construction, a PCR-based method was used for gene disruption as well as for introducing amino acid substitutions, truncations, fluorescent proteins and epitope tags into chromosomal genes40,41. The DNA fragments approximately 500 bp upstream and downstream of the target regions were amplified using S. pombe genomic DNA as the template. The reverse primer for the upstream fragment and forward primer for the downstream fragment contain the homologous sequence to the upstream (5’-TTAATTAACCCGGGGATCCG-3’) and downstream (5’-CAGATCCACTAGTGGCCTAT-3’) of a pFA6a plasmid cassette, respectively. A second round of PCR was performed using the amplified fragments and a pFA6a-derived plasmid42 containing the desired fluorescent protein, epitope tag, and selection marker as templates, generating two fragments with overlapping sequences within the selection marker region. The resulting DNA fragments were co-transformed into S. pombe cells to replace the endogenous locus via homologous recombination. Transformants were selected on appropriate selection plates and verified by PCR. Introduced amino acid substitutions and truncations were confirmed by genomic PCR followed by Sanger sequencing. Yeast two-hybrid assay In the yeast two-hybrid assay, the budding yeast Y2HGold strain (Clontech Laboratories) was used as a host. Cells were co-transformed with pGADGH and pGBT plasmids expressing various Rhs1 fragments (prey) and full-length Dri1 (bait), respectively. The Dri1-Rhs1 interaction was assessed by cell growth on SD medium lacking adenine and histidine (SD -Ade/-His), indicating activation of the reporter genes (ADE2 and HIS3). S. pombe growth assay Fission yeast cells were grown in YES liquid medium, and the cultures were adjusted to the cell concentration equivalent to an optical density at 600 nm (OD600) of 1.0. Serial dilutions of the adjusted cultures were spotted onto agar solid media. After an incubation at 30 ˚C, 36 ˚C, 37 ˚C, and 38 ˚C for 2–3 days, and 39 ˚C for 3–4 days, images of agar plates were captured by the LAS-4000 system (FUJIFILM, Japan). For the growth curve assay, cells were grown in YES liquid media at 30 ˚C. Overnight cultures were adjusted to initial OD600 = 0.115 ± 0.005 in fresh media, and cell density was measured at indicated time intervals. Immunoblotting Crude cell lysates were prepared using trichloroacetic acid (TCA)43. Cells were mixed with TCA at a final concentration of 10% and then harvested. Cell pellets were suspended in 10% TCA and vigorously vortexed with 0.5 mm zirconia beads for 5 min at 4 ˚C. Following cell disruption, the suspensions were centrifuged at 800 × g for 10 min, and the supernatant was removed. The recovered pellets were then suspended in SDS sample buffer containing 0.5 M Tris-HCl (pH 8.0) and boiled for 5 min. After centrifugation at 17,700 × g for 15 min, the supernatant was collected as cell lysates. Proteins were separated by SDS-PAGE, transferred to nitrocellulose membranes, and probed with primary antibodies as follows; anti-DYKDDDDK (1:5000; 1E6, FUJIFILM Wako) for detection of FLAG-tagged proteins, anti-myc (1:5000; 9E10, Covance), anti-Spc143 (1:10000), anti-HA (1:2000; 12CA5, Roche), anti-GFP (1:2500; Cat. no. 04404, Nacalai Tesque), and anti-puromycin (1:1000; EPR27218-173, Abcam). Anti-rabbit IgG (H + L) HRP-conjugated (1:10000; Cat. no. W4011, Promega), anti-mouse IgG (H + L) HRP-conjugated (1:10000; Cat. no. W4021, Promega), and anti-rat IgG (H + L) HRP-conjugated (1:10000; Cat. no. 112-035-003, Jackson Immuno Research) were used as secondary antibodies. Immunoprecipitation and mass spectrometry Fission yeast cells were disrupted in lysis buffer (20 mM HEPES-NaOH [pH 7.5], 150 mM sodium glutamate, 10% glycerol, 0.25% Tween-20, 10 mM sodium fluoride, 10 mM p-nitrophenylphosphate, 10 mM sodium pyrophosphate, 10 mM β-glycerophosphate and 0.1 mM sodium orthovanadate), containing 1 mM PMSF and protease inhibitor cocktail (P8849, Sigma-Aldrich) with glass beads using Multi-beads Shocker (Yasui Kikai). The cell lysate was recovered by centrifugation for 15 min at 17,700 × g and the total protein concentrations of cell lysates were determined by Bradford assay. For interaction between myc-tagged Dri1 and FLAG-tagged Rhs1, the recovered cell lysates were incubated with anti-FLAG M2-affinity gel (Sigma-Aldrich) for 2 h at 4 ˚C, followed by extensive washes with lysis buffer. Resultant samples were subjected to immunoblotting. For interaction of GFP-tagged proteins, the cell lysates were incubated with anti-DYKDDDDK (FUJIFILM Wako) or anti-myc (Covance) antibodies for 1 h at 4 ˚C, followed by incubation with Dynabeads Protein G (Thermo Fisher Scientific) for 2 h at 4 ˚C. For identification of proteins interacting with the Dri1-Rhs1 complex, cell lysates were first incubated with anti-FLAG M2-affinity gel for 2 h at 4 ˚C, followed by elution using 3X FLAG peptides (Sigma-Aldrich). The recovered samples were then further incubated with EZviewTM Red anti-c-Myc affinity gel for 2 h at 4 ˚C. After extensive washes with lysis buffer, the resultant samples were resolved in a 12% Mini-PROTEAN TGX precast gel (Bio-Rad). Each lane was sliced into four pieces and digested with trypsin. Mass spectrometric analysis was performed using the LTQ-Orbitrap XL-HTC-PAL system. MS/MS spectra were analyzed by Mascot server (Matrix Science) and compared against NCBInr protein database (Taxonomy: S. pombe). Microscopy Fluorescence microscopy was performed using a DeltaVision Elite microscopy system (GE Healthcare) as described in our previous study44–46. Details are outlined below. Cells grown exponentially in EMM liquid were stained with Hoechst 33342 (FUJIFILM Wako) for nuclear visualization and mounted on a thin layer of EMM agar. Z-axial images were taken at 0.4 μm with a 60× objective lens. Deconvolution of images was performed using DeltaVision SoftWoRx software. RNA extraction, quantitative RT-PCR (qPCR) and RNA-seq For RNA-seq, cells were grown to exponential phase at 30 ˚C, then harvested and flash-frozen using liquid nitrogen, and total RNA was purified using FastGeneTM Premium Kit (Nippon Genetics, Japan). mRNA was enriched using oligo dT beads, and mRNA-sequencing was performed by BGI JAPAN (Kobe, Japan) using a DNBseq platform. Clean reads were mapped to reference genome using Bowtie247, and gene expression level was calculated with RSEM48. The up-regulated and down-regulated genes were determined by DEseq249, with thresholds of |log2FC | > 0.8 and adjusted P-value < 0.05. The reference genome sequence and gene annotation information of S. pombe were obtained from the PomBase database (http://www.pombase.org)50. GO categories of the target genes were determined using GO Term Finder (https://go.princeton.edu/cgi-bin/GOTermFinder), and the volcano and box plots were generated using R studio using ggplot2 package51. For RT-qPCR, cells were grown to exponential phase at 30 ˚C and shifted to 39 ˚C for heat treatment experiment. After a 16 h incubation at 39 ˚C, cells were harvested and total RNA was purified as described above. The SuperScript II reverse transcriptase (Thermo Fisher Scientific) was used to synthesize cDNA. RT-qPCR was carried out on a LightCycler 96 Real-Time PCR System using PowerUp SYBR Green Master mix (Thermo Fisher Scientific). The results were normalized against the housekeeping gene act1+. PCR primer sets used in RT-qPCR are listed in Supplementary Table 4. Polysome analysis Cells were grown in EMM at 30 ˚C and shifted to 39 ˚C for heat treatment experiment. After 24 h incubation at 39 ˚C, cell cultures were mixed with 0.1 mg mL−1 of cycloheximide and incubated 30 ˚C for 5 min. Cells were harvested and washed with polysome lysis buffer (20 mM Tris-HCI [pH 7.5], 150 mM KCl, 10 mM MgCl2, 100 µg mL−1 of cycloheximide, 1 mM dithiothreitol, 1 mM PMSF, and protease inhibitor cocktail), followed by flash-freezing with liquid nitrogen. Polysome analysis was performed as previously described in refs. 52,53, with details provided below. Frozen cell pellets were pulverized using a mortar and pestle and resuspended in 3 mL of extraction buffer (20 mM HEPES-KOH [pH 7.4], 2 mM magnesium acetate, and 100 mM potassium acetate). After centrifugation at 15,500 × g for 10 min, the supernatant was recovered. This procedure was repeated three times to obtain total RNA samples. 1 mg of total RNA was then loaded onto a 10–50% sucrose density gradient (10 mM Tris-acetate [pH 7.4], 70 mM ammonium acetate, and 40 mM magnesium acetate) prepared in centrifuge tubes using a Gradient Master (107–201 M; BioComp Instruments), and centrifuged at 130,000 × g for 3 h at 4 ˚C. Gradients were fractionated using a Gradient Fractionator (152-002; BioComp Instruments), and absorbance at 254 nm was continuously monitored with a Bio-mini UV monitor AC-5100 (ATTO) to generate polysome profiles. Polysome-to-Monosome (P/M) ratios were calculated by the areas of monosome and polysome peaks using Adobe Photoshop Elements 7. Puromycin assay In vivo incorporation of puromycin into nascent polypeptides was assessed by immunoblotting as a measure of overall protein synthesis26,27. Cells were grown in YES at 30 ˚C and shifted to 39 ˚C for heat treatment experiment. After 24 h incubation at 39 ˚C, cells were mixed with 1 mM puromycin and further incubated for 30 min. Cells were then harvested, and crude lysates were prepared for immunoblotting. As a negative control, cells were treated with DMSO instead of puromycin. Statistics and reproducibility Statistical analyses were performed using a two-tailed Student’s t-test, and p-values are indicated as * (P < 0.05), ** (P < 0.01), and *** (P < 0.001). Data are presented as mean ± standard deviation (SD). Sample sizes for each experiment are specified in the figure legends. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Transparent Peer Review file Supplementary information Description of Additional Supplementary files Supplementary Data 1 Supplementary Data 2 Supplementary Data 3 Reporting Summary Acknowledgements We thank A Higuchi and the NAIST Life Science Collaboration Center (LiSCo) for technical assistance. We are also grateful to T Toda and M Yukawa for sharing unpublished data and strains. Author contributions Y.M. and K.S. designed research; Y.A., A.Y., A.C.W.T., T.N., Y.F., Y.N., F.M., S.I. and Y.M. performed research; Y.A., A.Y., A.C.W.T., T.N., Y.N., F.M., S.I. and Y.M. analyzed data; and K.S. and Y.M. wrote the paper. Peer review Peer review information Communications Biology thanks Abhrajyoti Ghosh, Udita Acharya and Margarita Cabrera for their contribution to the peer review of this work. Primary Handling Editors: Dr. Nilanjan Banerjee. A peer review file is available. Funding This study was supported by research grants to Y.M. from the Sumitomo Foundation and the Noda Institute for Scientific Research, and to Y.N. from the Tojuro Iijima Foundation for Food Science and Technology. This work was also supported by the Takeda Science Foundation to K.S. and Y.M., the Ohsumi Frontier Science Foundation to S.I. and K.S., and the Institute for Fermentation, Osaka (IFO) to S.I. and Y.M, and the Japan Society for the Promotion of Science (JSPS) KAKENHI grants (19K06564, 22K06145 and 25K09615 to Y.M., 20K06485 and 25K09040 to Y.N., and 19H03224 to K.S.). Data availability The RNA sequence data in this article have been deposited in the DDBJ Sequence Read Archive (DRA) under accession codes DRR654225-DRR654230. The mass spectrometry data have been deposited in the Japan Proteome Standard Repository (jPOSTrepo)54 under accession code JPST004115. Source data for the graphs and uncropped immunoblot images are provided in Supplementary Data 3 and Supplementary Fig. 8, respectively. All other data that support the findings of this study are available from the corresponding author upon reasonable request. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Yutaka Akikusa, Atsuya Yamaguchi. Supplementary information The online version contains supplementary material available at 10.1038/s42003-026-10292-y. References Associated Data Supplementary Materials Transparent Peer Review file Supplementary information Description of Additional Supplementary files Supplementary Data 1 Supplementary Data 2 Supplementary Data 3 Reporting Summary Data Availability Statement The RNA sequence data in this article have been deposited in the DDBJ Sequence Read Archive (DRA) under accession codes DRR654225-DRR654230. The mass spectrometry data have been deposited in the Japan Proteome Standard Repository (jPOSTrepo)54 under accession code JPST004115. Source data for the graphs and uncropped immunoblot images are provided in Supplementary Data 3 and Supplementary Fig. 8, respectively. All other data that support the findings of this study are available from the corresponding author upon reasonable request.