Enhancing Freezing Stress Tolerance through Regulation of the Ubiquitin–Proteasome System in Saccharomyces cerevisiae

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript provides informations about the significant issue of freezing stress tolerance in yeast within the context of the baking industry, outlining a clear research objective focused on elucidating the molecular interplay between the ubiquitin–proteasome system and freezing stress tolerance in Saccharomyces cerevisiae. The experimental approach involving the screening of mutants with enhanced freezing stress tolerance using the proteasome inhibitor MG132 provides a solid foundation for further investigation into the molecular mechanisms involved. The identification of mutations in the ROX1 gene through genomic analysis, a heme-dependent repressor of hypoxic genes, and linking ROX1 deletion to improved freezing stress tolerance, adds depth to the study and offers a clear genetic basis for the observed phenotype. Furthermore, the identification of ANB1 as a potential downstream target of ROX1 and its role in enhancing freezing stress tolerance, along with the connection between ROX1, ANB1, and proteasome activity, contributes significantly to our understanding of the underlying mechanisms. The manuscript effectively discusses the implications of the findings for the baking industry and suggests potential applications in the development of baker's yeast strains, although further elaboration on specific strategies or interventions derived from this research could enhance its impact. Acknowledging any limitations of the study and suggesting avenues for further investigation, such as exploring additional genes or pathways involved in freezing stress tolerance and conducting practical experiments to validate the findings in industrial yeast strains, would strengthen the manuscript. Overall, with minor revisions and additional context, the manuscript has the potential to make a valuable contribution to the field and be a significant publication, as outlined below.

#1 In the abstract, please add more quantitative informations that summarizes your work as a whole. Also, the abstract briefly mentions the potential applications of the study's findings in the development of baker's yeast strains. Consider expanding on this point to highlight the broader significance of the research and its implications for the baking industry.

#2 While the introduction covers relevant background information, it could benefit from clearer organization and flow. Consider breaking down the information into more distinct sections or paragraphs to improve readability and facilitate understanding for readers.

#3 The introduction effectively explains the concept of freezing stress and its impact on yeast cells. However, providing a brief overview of how freezing stress affects yeast physiology and metabolism could enhance the reader's understanding of the problem.

#4 The section introducing the ubiquitin-proteasome system is informative, but it might be too detailed for an introductory section. Consider providing a concise overview of the system's role in protein degradation and cellular maintenance before delving into specific details.

#5 While the introduction briefly mentions previous studies linking the ubiquitin-proteasome system to freezing stress tolerance, it could benefit from a more comprehensive review of relevant literature. Providing a clearer overview of existing research in this area would help situate the current study within the broader scientific context.

#6 The introduction does a good job of outlining the research objectives by mentioning the investigation into the relationship between the ubiquitin-proteasome system and freezing stress tolerance in S. cerevisiae. However, explicitly stating the research hypotheses or questions that the study aims to address would provide clarity on the specific aims of the research.

#7 The discussion effectively explains the role of the Rox1-Anb1 pathway in regulating freezing stress tolerance and proteasome activity. However, some aspects of the findings could be further elaborated to provide a more comprehensive understanding. For example, discussing the potential mechanisms by which Anb1 and Hyp2 isoforms exert different effects on freezing stress tolerance would add depth to the interpretation of the results.

#8 While the discussion mentions previous research on proteasome-related proteins and their role in protein quality control mechanisms, it could be strengthened by more explicitly integrating these findings into the current study's context. Providing a more detailed comparison with previous studies would help readers understand how the current findings contribute to existing knowledge in the field.

#9 The discussion mentions potential contradictions in the observed effects of overexpression of certain proteins on proteasome activity. Addressing these contradictions and proposing explanations or hypotheses for the observed discrepancies would strengthen the discussion and highlight areas for future research.

#10 The discussion briefly touches on potential future directions, such as investigating proteasome structure and identifying denatured proteins under freezing stress conditions. Expanding on these potential avenues for future research and discussing their significance would provide a clearer roadmap for further investigation in this area.

#11 For me it is very strange that an extensive work like this has no conclusion. In my opinion, this work deserves a good conclusion section. Conclusions are an essential component of any scientific manuscript as they serve to succinctly summarize the key findings and implications of the study. While some journals may consider the conclusions section optional (including the Fermentation journal), it is crucial to emphasize the importance of including conclusions for the overall understanding and impact of the manuscript. First and foremost, conclusions provide a cohesive summary of the study's results, allowing readers to quickly grasp the main findings without having to sift through the entire manuscript. This is especially valuable in fields where research articles can be dense and technical, as concise conclusions help readers extract the most relevant information efficiently. Furthermore, conclusions offer the authors an opportunity to contextualize their findings within the broader scientific landscape. By discussing the significance of their results and potential implications for the field, authors can demonstrate the novelty and importance of their research. This not only enhances the credibility of the study but also facilitates knowledge dissemination and contributes to the advancement of scientific knowledge. Additionally, conclusions can serve as a springboard for future research directions (considering the lack of these information in your summary.). By highlighting areas for further investigation or unresolved questions raised by the study, authors can inspire future studies and collaborations. This helps foster a continuous cycle of scientific inquiry and innovation.

#12 The language used in the manuscript is clear, and the scientific concepts are effectively communicated. However, there are a few areas where minor improvements could be made for clarity and conciseness, such as streamlining certain sentences and organizing the content for better readability. Please review the English of your text carefully and pay attention as a whole.

Comments on the Quality of English Language

Please see the report!

Author Response

Reviewer: 1

The manuscript provides informations about the significant issue of freezing stress tolerance in yeast within the context of the baking industry, outlining a clear research objective focused on elucidating the molecular interplay between the ubiquitin–proteasome system and freezing stress tolerance in Saccharomyces cerevisiae. The experimental approach involving the screening of mutants with enhanced freezing stress tolerance using the proteasome inhibitor MG132 provides a solid foundation for further investigation into the molecular mechanisms involved. The identification of mutations in the ROX1 gene through genomic analysis, a heme-dependent repressor of hypoxic genes, and linking ROX1 deletion to improved freezing stress tolerance, adds depth to the study and offers a clear genetic basis for the observed phenotype. Furthermore, the identification of ANB1 as a potential downstream target of ROX1 and its role in enhancing freezing stress tolerance, along with the connection between ROX1, ANB1, and proteasome activity, contributes significantly to our understanding of the underlying mechanisms. The manuscript effectively discusses the implications of the findings for the baking industry and suggests potential applications in the development of baker's yeast strains, although further elaboration on specific strategies or interventions derived from this research could enhance its impact. Acknowledging any limitations of the study and suggesting avenues for further investigation, such as exploring additional genes or pathways involved in freezing stress tolerance and conducting practical experiments to validate the findings in industrial yeast strains, would strengthen the manuscript. Overall, with minor revisions and additional context, the manuscript has the potential to make a valuable contribution to the field and be a significant publication, as outlined below.

We are really grateful for your very positive comments, which indeed were helpful in improving our manuscript.

 

#1 In the abstract, please add more quantitative informations that summarizes your work as a whole. Also, the abstract briefly mentions the potential applications of the study's findings in the development of baker's yeast strains. Consider expanding on this point to highlight the broader significance of the research and its implications for the baking industry.

According to your suggestion, we have included some quantitative information and statements about the impact of our work on the baking industry in the Abstract section of the revised manuscript as follows: “The baking industry is experiencing significant growth, primarily due to the widespread adoption of frozen dough baking. However, this process can negatively impact the fermentation ability of yeast, as freezing can induce stress in yeast cells. This study reports the molecular interplay between the ubiquitin-proteasome system and freezing stress tolerance in the yeast Saccharomyces cerevisiae. Using the proteasome inhibitor MG132, we first screened mutants with enhanced freezing stress tolerance. Three mutants showed the elevated activity of intracellular proteasome, particularly trypsin-like activity (more than 3-fold) and reduced sensitivity to MG132 inhibition of chymotrypsin-like activity (less than 0.125-fold). Genomic analysis with these mutants revealed mutations in the ROX1 gene, a heme-dependent re-pressor of hypoxic genes. Importantly, the ROX1 deletion strain displays slightly improved freezing stress tolerance (about 1.5-fold). Comprehensive transcription analysis identified the ANB1 gene as a potential downstream target of Rox1. Overexpression of ANB1 enhanced freezing stress tolerance (about 1.5-fold) with increased proteasome activity, indicating that Rox1 contributes to changes in proteasome activity and freezing stress tolerance through the function of Anb1. The present data provide new insights into mechanisms of freezing stress tolerance and help us improve the baking of frozen dough to produce higher-quality bread.” (P. 1, L. 17-31).

 

#2 While the introduction covers relevant background information, it could benefit from clearer organization and flow. Consider breaking down the information into more distinct sections or paragraphs to improve readability and facilitate understanding for readers.

Due to the changes requested by you and other reviewers, the current text effectively provides the necessary information in a clear and understandable form. We hope that you will consider the current version of the manuscript.

 

#3 The introduction effectively explains the concept of freezing stress and its impact on yeast cells. However, providing a brief overview of how freezing stress affects yeast physiology and metabolism could enhance the reader's understanding of the problem.

Yeast cells possess inherent mechanisms to respond and adapt to many environmental stresses. Under freezing stress conditions, yeast cells exhibit various responses to maintain cellular homeostasis, including the expression of low-temperature chaperones, accumulation of compatible solutes such as trehalose, and changes in cell membrane composition involving unsaturated fatty acids. The intracellular trehalose and glycerol could play an important role in protecting yeast cells from ice crystal formation. Additionally, extracellular amino acids such as proline have cryoprotective activity. These statements have been included in the revised manuscript (P. 2, L. 49-56).

 

#4 The section introducing the ubiquitin-proteasome system is informative, but it might be too detailed for an introductory section. Consider providing a concise overview of the system's role in protein degradation and cellular maintenance before delving into specific details.

Thank you for providing us the nice advice. Protein quality control mechanisms consist of two systems: the degradation system mediated by ubiquitin-proteasome and the repair system involving chaperones. Chaperones can refold misfolded proteins into their native conformation, but they are ineffective against severe protein misfolding (denaturation). Freezing stress can cause severe protein damage, and the degradation system may be the primary mechanism for protein quality control under freezing stress. These statements have now been included in the revised manuscript (P. 2, L. 64-69).

 

#5 While the introduction briefly mentions previous studies linking the ubiquitin-proteasome system to freezing stress tolerance, it could benefit from a more comprehensive review of relevant literature. Providing a clearer overview of existing research in this area would help situate the current study within the broader scientific context.

This is a similar comment as the above #3. Please see our response described above.

 

#6 The introduction does a good job of outlining the research objectives by mentioning the investigation into the relationship between the ubiquitin-proteasome system and freezing stress tolerance in S. cerevisiae. However, explicitly stating the research hypotheses or questions that the study aims to address would provide clarity on the specific aims of the research.

Frozen dough baking has many advantages over conventional baking methods. However, baker’s yeast cells are exposed to severe freezing stress, which limits their fermentation ability. Although the proteasome system has been well studied, the molecular relationship between the proteasome system and freezing stress tolerance has not been fully understood. We have now included these statements in the revised manuscript (P. 3, L. 113-117).

 

#7 The discussion effectively explains the role of the Rox1-Anb1 pathway in regulating freezing stress tolerance and proteasome activity. However, some aspects of the findings could be further elaborated to provide a more comprehensive understanding. For example, discussing the potential mechanisms by which Anb1 and Hyp2 isoforms exert different effects on freezing stress tolerance would add depth to the interpretation of the results.

We have already discussed the different effects between Anb1 and Hyp2 (please see the below sentences) in the Discussion section. Anb1 and Hyp2 (encoded by the ANB1 paralog gene) are isoforms of the translation elongation factor eIF-5A, which supports the synthesis of nascent proline-rich proteins by mitigating translational stalling. Anb1 and Hyp2 share over 90% similarity in their amino acid sequences, suggesting that they recognize the same proteins. However, when overexpressed, Anb1 and Hyp2 have different effects on freezing stress tolerance, suggesting that they have distinct substrate specificities. To better understand these differences, we need to investigate the expression of proline-rich proteins in the ROX1-deleted and ANB1-overexpressing strains. This will help us determine the precise substrate specificities of Anb1 and Hyp2. We hope to report the details in the near future.

 

#8 While the discussion mentions previous research on proteasome-related proteins and their role in protein quality control mechanisms, it could be strengthened by more explicitly integrating these findings into the current study's context. Providing a more detailed comparison with previous studies would help readers understand how the current findings contribute to existing knowledge in the field.

According to your suggestion, we have now modified and added new sentences (please see the below sentence) in the Discussion section. “The frozen-dough baking is a promising technology for future advancements in food manufacturing technology. This process can lead to intense freezing stress on yeast cells, which reduces the bread quality. In the previous study, Watanabe et al. found that the expression level of proteasome-related genes was reduced, and ubiquitinated proteins were accumulated by freeze-thaw stress in strains with reduced fermentation ability after cryopreservation. These results suggested that proteasome function is involved in fermentation ability after freezing stress. Expanding upon their finding, our present results indicate that the Rox1-Anb1 pathway represses freezing stress tolerance by altering proteasome activity. Moreover, ROX1 deletion or ANB1 overexpression improved fermentation performance after the stress exposure. These findings provide insight into tolerance mechanisms to freezing stress and their potential applications in the development of baker’s yeast strains.” (P. 17, L. 606-617).

 

#9 The discussion mentions potential contradictions in the observed effects of overexpression of certain proteins on proteasome activity. Addressing these contradictions and proposing explanations or hypotheses for the observed discrepancies would strengthen the discussion and highlight areas for future research.

We have changed the original statements as follows: “Previous reports have suggested that Fub1 inhibits proteasome activity in vitro but assists in stabilizing proteasome conformation and degrading substrate proteins in vivo. Freezing and thawing are complexed stresses, including ice crystal formation, ROS generation, and dehydration. The proteasome was reportedly vulnerable to oxidative damage, which occurs under freezing stress conditions, and ROS can disrupt proteasome structure. Therefore, Fub1 may stabilize the proteasome structure under freezing stress conditions by binding to the proteasome. The Fub1-bound proteasome may also facilitate the recruitment of substrate proteins and efficiently degrade denatured proteins generated during freezing stress. Under the low-oxygen conditions where ANB1 expression is induced due to the release of Rox1 inhibition, mitochondrial ROS production is increased. Rox1 is hypothesized to induce ANB1 expression in response to low oxygen, thereby protecting the proteasome from oxidative damage. Additionally, the Fub1-bound proteasome may facilitate the recruitment of substrate proteins, potentially enabling efficient degradation of denatured proteins arising from freezing stress. The mechanism underlying freezing stress tolerance in the Rox1-Anb1 pathway may be clarified through observations of the proteasome structure and the identification of denatured proteins that accumulate under freezing stress conditions.” (P. 18, L. 646-662).

 

#10 The discussion briefly touches on potential future directions, such as investigating proteasome structure and identifying denatured proteins under freezing stress conditions. Expanding on these potential avenues for future research and discussing their significance would provide a clearer roadmap for further investigation in this area.

Since proteasome inhibitors such as MG132 cannot penetrate yeast cell membranes, it is quite hard to show sensitivity or resistance to proteasome inhibitors. In this study, MG132-sensible yeast strains were constructed using genetic recombination technology. The use of genetic engineering techniques for yeast breeding is challenging due to consumer and public concerns regarding its safety. Here, we have the possibility that treatment with low concentrations of amphotericin, which binds to ergosterol, may increase the permeability of proteasome inhibitors into yeast cells, leading to increased sensitivity to them. We plan to investigate this screening method in practical baker’s yeast in the future. Additionally, this breeding method could be applied to creating yeast strains with enhanced tolerance to stresses other than freezing, because the proteasome is crucial for many stresses that induce protein denaturation. In fermentation environments other than bread making, there are various factors that induce stress conditions, such as high ethanol concentrations, high osmotic pressure, low or high temperatures, and oxidation, all of which can denature intracellular proteins. Hence, the above breeding strategy has the potential to be utilized in the construction of various practical yeast strains beyond baker’s yeast. We have included these statements in the revised manuscript (P. 18, L. 663-678).

 

#11 For me it is very strange that an extensive work like this has no conclusion. In my opinion, this work deserves a good conclusion section. Conclusions are an essential component of any scientific manuscript as they serve to succinctly summarize the key findings and implications of the study. While some journals may consider the conclusions section optional (including the Fermentation journal), it is crucial to emphasize the importance of including conclusions for the overall understanding and impact of the manuscript. First and foremost, conclusions provide a cohesive summary of the study's results, allowing readers to quickly grasp the main findings without having to sift through the entire manuscript. This is especially valuable in fields where research articles can be dense and technical, as concise conclusions help readers extract the most relevant information efficiently. Furthermore, conclusions offer the authors an opportunity to contextualize their findings within the broader scientific landscape. By discussing the significance of their results and potential implications for the field, authors can demonstrate the novelty and importance of their research. This not only enhances the credibility of the study but also facilitates knowledge dissemination and contributes to the advancement of scientific knowledge. Additionally, conclusions can serve as a springboard for future research directions (considering the lack of these information in your summary.). By highlighting areas for further investigation or unresolved questions raised by the study, authors can inspire future studies and collaborations. This helps foster a continuous cycle of scientific inquiry and innovation.

              We have understood the importance of the Conclusion section and summarized our study as follows: In this study, we investigated the relationship between the ubiquitin-proteasome system and freezing stress tolerance in S. cerevisiae. Screening with MG132 isolated some mutants with enhanced freezing stress tolerance, characterized by increased intracellular proteasome activity, particularly trypsin-like proteasome activity, and by decreased sensitivity to MG132 inhibition of chymotrypsin-like activity. Whole-genome sequencing showed that loss-of-function mutations in ROX1, a heme-dependent re-pressor of hypoxic genes, result in enhanced freezing stress tolerance. Therefore, transcription analysis revealed ANB1 as a potential downstream target gene of Rox1. This study proposes a new mechanism for enhancing stress tolerance during freezing mediated by the ubiquitin-proteasome system. By applying the knowledge gained from this study to the development of baker’s yeast strains, we can improve the baking process of frozen dough to produce higher-quality bread. Furthermore, elucidation of the regulatory mechanism underlying the ubiquitin-proteasome system is expected to provide valuable insights into protein quality control mechanisms. (P. 18, L. 679-P. 19, L. 693).

 

#12 The language used in the manuscript is clear, and the scientific concepts are effectively communicated. However, there are a few areas where minor improvements could be made for clarity and conciseness, such as streamlining certain sentences and organizing the content for better readability. Please review the English of your text carefully and pay attention as a whole.

Thank you so much for your careful review of our manuscript. We have now checked the revised manuscript appropriately. We would appreciate it if you would kindly accept our responses.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The authors screened mutants with enhanced freezing stress tolerance, and genomic analysis of the best mutants revealed mutations in the ROX1 gene, which is a transcription repressor of hypoxic genes. The authors then found that the ROX1 deletion strain displays improved freezing stress tolerance. Furthermore, transcription analysis identified ANB1as a potential downstream target of Rox1. The authors also proved that overexpression of ANB1 enhanced freezing stress tolerance with increased proteasome activity.  These data provide new insights into mechanisms of freezing stress tolerance. The work is nicely designed and the results are convincing. I have suggestions for minor revision and discussion:

1. Abstract: please revise the grammar mistake, for example, ...These mutants showed elevated the activity of intracellular proteasome,..here 'the' should be removed. Please also check other possible mistakes in English writing. 

2. The fermentation changes are for me not so convincing, although the values are of significance statistically. Please discuss. 

3. The authors used the erg6 and pdr5 double deletion mutants for the screening, is it possible that these mutations also affect the effect of ROX1? 

4. In the discussion, the authors mentioned about Pre2 and Fub1, but there are no data on the functions of these two genes in the study. I would suggest that the authors carefully revise the discussion part. 

5. The authors stated that Rox1 contributes to changes in proteasome activity and freezing stress tolerance through the function of Anb1, but Anb1 may not be the only reason, please discuss. 

 

Author Response

Reviewer: 2

The authors screened mutants with enhanced freezing stress tolerance, and genomic analysis of the best mutants revealed mutations in the ROX1 gene, which is a transcription repressor of hypoxic genes. The authors then found that the ROX1 deletion strain displays improved freezing stress tolerance. Furthermore, transcription analysis identified ANB1as a potential downstream target of Rox1. The authors also proved that overexpression of ANB1 enhanced freezing stress tolerance with increased proteasome activity. These data provide new insights into mechanisms of freezing stress tolerance. The work is nicely designed and the results are convincing. I have suggestions for minor revision and discussion:

Thank you so much for your careful review of our manuscript. According to these insightful comments, we have revised our manuscript appropriately by changing the overstated text and adding new statements, as you suggested.

 

1. Abstract: please revise the grammar mistake, for example, ...These mutants showed elevated the activity of intracellular proteasome,..here 'the' should be removed. Please also check other possible mistakes in English writing.

Thank you so much for your careful review of our manuscript. We have now checked the revised manuscript appropriately. We would appreciate it if you would kindly accept our responses.

 

2. The fermentation changes are for me not so convincing, although the values are of significance statistically. Please discuss.

Fermentation ability is significantly influenced by cell viability. Our data showed a decrease in cell viability after freezing stress. Therefore, the fermentation changes are convincing. You might assume that the data on fermentation ability are unnecessary in this study. However, researchers in the field of food engineering often prioritize fermentation ability rather than cell viability. Hence, we have included both data sets in this manuscript.

 

3. The authors used the erg6 and pdr5 double deletion mutants for the screening, is it possible that these mutations also affect the effect of ROX1?

Both ERG6 and ROX1 are important for adaptation under anaerobic conditions. Therefore, a genetic link between ERG6 and ROX1 may exist. However, the changes in proteasome and freezing stress tolerance induced by ROX1 deletion were observed in the presence or absence of ERG6. Thus, we believe that the double deletion of ERG6 and PDR5 might not affect Rox1 functions, at least in freezing tolerance.

 

4. In the discussion, the authors mentioned about Pre2 and Fub1, but there are no data on the functions of these two genes in the study. I would suggest that the authors carefully revise the discussion part.

We think that the discussion part needs new perspectives suggested by the present data. Therefore, we have kept the sentences related to Pre2 and Fub1 functions, even although we do not have any actual data for Fub1 or Pre2. We have toned down the text because some parts are overstated, as you pointed out. We would appreciate it if you would kindly accept our responses.

 

5. The authors stated that Rox1 contributes to changes in proteasome activity and freezing stress tolerance through the function of Anb1, but Anb1 may not be the only reason, please discuss.

              The changes in proteasome activity and freezing stress tolerance caused by ROX1 deletion were almost similar to those caused by ANB1 overexpression. Furthermore, ANB1 deletion canceled the increased survival observed in the ROX1 deletion strain was completely canceled. Our genetic data clearly showed that Anb1 is the only downstream of Rox1, although further studies, including biochemical and physiological analyses, are needed.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript describes a screen to identify genes that alter proteasome activity following freeze stress in bakers yeast. The authors found that mutations in rox1 led to improved freeze tolerance and increased proteasome activities. Additionally, the authors found that transcription of ANB1 is increased in a rox1∆ and that ANB1 activity is required for rox1∆-mediated tolerance to freeze stress. Overall this manuscript makes a nice contribution, is very well written and a pleasure to read. There are some statistical concerns, and more information is needed regarding the bioinformatic analyses, though I think these are easily addressed and will not change the final story. Comments are meant to improve the manuscript and its impact.

 

Specific comments: (major concerns/comments shown in bold)

1.     Line 244: Proteasome activity wasn’t actually measured here and so I think it’s a reach to say that proteasome activity correlates with freezing stress tolerance. Could this be reworded?

 

2.     Figure 1B and C: How do the MT mutants grow as compared to the wt strain? If a wild type strain was not included on these plates, could the number of incubation days that the plates shown in Fig1B and 1C were done be included to provide a rough comparison?

 

3.     Methods: 2.4 Cell viability. The authors should state how 100% viability was estimated (presumably this was an aliquot of the cells plated to YPD prior to freezing). Approximate numbers of colonies counted would be an appropriate detail to add, as there are statistical insight as to whether 10% is derived from (100/1000 vs. 1/10 CFUs), and may explain the differences between the cell viability of the erg6∆pdr5∆ strain between Fig 1D (~5%) and Fig2A (~30% +/-~5%).

 

4.     The authors should provide the bioinformatic pipelines for how the Illumina reads were processed to reveal the rox1 mutations. Overall sequence coverages and if additional (as yet unexplored) mutations that were detected would be appropriate to include.

 

5.     Fig2 B-D, Fig 3B-D; Fig 4 A-C, Fig 5B-D. Fig 6A-C. Instead of t-tests, these data should be analyzed using two-way ANOVAs (genotype x environment) followed by posthoc Tukey tests. This would allow for all comparisons to be made (mutants vs ∆∆ treated, mutants vs ∆∆ untreated, and each strain treated vs. untreated, etc.). It’s not clear (to me at least) what the percentages shown in the figures are referring to. They don’t appear to be consistent with the change in activity, nor the starting or ending activity levels. It’s possible that the MG132 treatment lowers enzyme activities to a similarly low level, regardless of the initial enzyme activity. If there are no differences between the enzyme activities between the ∆∆ and mutant treated cells (ie..MG132 destroys enzyme activity to a baseline regardless of starting activity levels), then it may not make (biological) sense to compare the changes in enzyme activities (if that is what the % numbers on the plots are indeed referring to..?).   perhaps related:  What is the importance of “the numerical value represents the residual activity in the presence of MG132”?

 

6.     Line 357: Remove the word “carefully”. (unless you mean to suggest that the other analyses were not performed with similar care!).

 

7.     Fig 4EF and 6FG, Fig S1B: It’s not clear what the 2 way ANOVAs used here are comparing. A repeated measures ANOVA that takes into account that multiple measures were taken from the same sample (acknowledging that variances between time points are not independent) would be more appropriate. This could also include an interaction term between sample and timepoint. If the shape of the time course is not of interest, then a simple one-way anova comparison of the end time point to show that more CO₂ is produced in the rox1∆ strain would be appropriate. Also, the key states that the data are presented with  ±SD though no error bars are presented on the graphs.

 

8.     RNAseq: The bioinformatic pipeline for how the RNAseq reads were analyzed should be included.

 

9.     The title of Table S3 is misleading as it doesn’t indicate the actual criteria for the listed genes. The authors should provide a table with all the genes presented in Fig S3, with appropriate statistics (log2fold change, p values ,etc).

 

10.  I’m surprised that expression of COX5b is not in the list of genes that are differentially expressed between all 3 conditions. Does COX5b (or any of the hypoxic genes known to be repressed by Rox1) appear in any component of the venn diagram shown in Fig S3? GO ontology analyses could help to provide connections to genes and pathways involved, and I’m somewhat surprised these aren’t reported. I’m also wondering what expression differences between wt and rox1∆ cells at steady state (where cells are most likely undergoing diauxic shift) vs cells wakening from a frozen (quiescent?) state actually mean. Could the authors comment on this?

 

11.  Line 456-458: states that overexpression of ANB1 reduces MG132 sensitivity to a level that was similar to the rox1∆. This isn’t shown, as far as I can tell. The authors should direct the reader to this data, or if not shown, state that is it not shown (if that is permissible by the journal) and remove the “more importantly” qualifier.

 

12.  If ANB1 overexpression leads to a similar phenotype as a rox1 mutant, then I am thinking back to the original screen. All 3 of the sequenced strains had rox1 mutations (and with identical mutations), suggesting to me that the screen might be saturated. It would be interesting to know if the other un-sequenced MT mutants also had rox1 mutations or if there are potentially additional as yet uncharacterized mutations in this screen (such as in the promoter of ANB1). I guess a good old fashioned complementation test could determine this, or Sanger sequencing of candidate genes. I don’t suggest that this needs to be done for this publication, but maybe the authors could reflect on the experiments presented here, including the screen, why tryspin/caspase/chymotrypsin-like activities differ in the various genotypes and treatments , big picture with expression differences, etc, in the Discussion before delving into potential next experiments involving ANB1/HYP2, Fub1.

Author Response

Reviewer 3:

This manuscript describes a screen to identify genes that alter proteasome activity following freeze stress in baker’s yeast. The authors found that mutations in rox1 led to improved freeze tolerance and increased proteasome activities. Additionally, the authors found that transcription of ANB1 is increased in a rox1∆ and that ANB1 activity is required for rox1∆-mediated tolerance to freeze stress. Overall this manuscript makes a nice contribution, is very well written and a pleasure to read. There are some statistical concerns, and more information is needed regarding the bioinformatic analyses, though I think these are easily addressed and will not change the final story. Comments are meant to improve the manuscript and its impact.

Thank you so much for your positive comments. We have revised our manuscript appropriately by changing the overstated text and adding new statements, as you suggested.

 

Specific comments: (major concerns/comments shown in bold)

Line 244: Proteasome activity wasn’t actually measured here and so I think it’s a reach to say that proteasome activity correlates with freezing stress tolerance. Could this be reworded?

              According to your suggestion, we have changed the original sentence to “The ubiquitin-proteasome system has a relationship with freezing stress tolerance.” (P. 6, L. 298-299).

 

2. Figure 1B and C: How do the MT mutants grow as compared to the wt strain? If a wild type strain was not included on these plates, could the number of incubation days that the plates shown in Fig1B and 1C were done be included to provide a rough comparison?

              Since the WT strain was not included in Fig. 1B, we cannot directly compare its growth with the MT mutants. However, all plates were incubated for the same period (3 days). Rough comparisons suggest that the MT strains grew slightly slower than the WT strain. We have added this statement in the revised manuscript (P. 6, L. 310-P. 7, L. 312).

 

3. Methods: 2.4 Cell viability. The authors should state how 100% viability was estimated (presumably this was an aliquot of the cells plated to YPD prior to freezing). Approximate numbers of colonies counted would be an appropriate detail to add, as there are statistical insight as to whether 10% is derived from (100/1000 vs. 1/10 CFUs), and may explain the differences between the cell viability of the erg6∆pdr5∆ strain between Fig 1D (~5%) and Fig2A (~30% +/-~5%).

As you pointed out, cell viability was calculated as follows: (numbers of colonies after freezing stress treatment/numbers of colonies before freezing stress treatment) × 100. We have now added the sentence in the Material and Methods section (P. 4, L. 187-189).

The freezing stress tests currently being conducted by many researchers (including us), are affected by various parameters, such as door opening/closing and room temperature. There is a great deal of variability in experimental data. Therefore, we cannot simply compare the results between each figure. A better freezing stress test should be developed, but we now think that this is the best method we can do. We believe that our experimental system is acceptable because the tendency of each mutant to be tolerant remains exactly the same. We would appreciate it if you would kindly accept our responses.

 

4. The authors should provide the bioinformatic pipelines for how the Illumina reads were processed to reveal the rox1 mutations. Overall sequence coverages and if additional (as yet unexplored) mutations that were detected would be appropriate to include.

A quality control process was applied to the raw paired-end sequence reads that passed the FastQC. Low-quality (< 20) bases and adapter sequences were trimmed by Trimmomatic software (Version 0.38) and obtained a final coverage of 96-98%. The reads were mapped to the S. cerevisiae S288C reference genome (version R64-1-1) with Bwa-mem (v0.7.17-r1188), using default parameters. The number of duplicates was calculated using Picard MarkDuplicates, and single nucleotide variants (SNVs) and short indels were called using Bcftools (Version 1.9). To identify the putative effects on protein translation and high-impact mutations, identified variants were annotated using SnpEff (Version 4.3t) (P. 5, L. 223-231).

We have not yet determined other mutations, but we have deposited all data into the DDBJ Sequenced Read Archive. We expect other researchers’ responses.

 

5. Fig2 B-D, Fig 3B-D; Fig 4 A-C, Fig 5B-D. Fig 6A-C. Instead of t-tests, these data should be analyzed using two-way ANOVAs (genotype x environment) followed by posthoc Tukey tests. This would allow for all comparisons to be made (mutants vs ∆∆ treated, mutants vs ∆∆ untreated, and each strain treated vs. untreated, etc.). It’s not clear (to me at least) what the percentages shown in the figures are referring to. They don’t appear to be consistent with the change in activity, nor the starting or ending activity levels. It’s possible that the MG132 treatment lowers enzyme activities to a similarly low level, regardless of the initial enzyme activity. If there are no differences between the enzyme activities between the ∆∆ and mutant treated cells (ie..MG132 destroys enzyme activity to a baseline regardless of starting activity levels), then it may not make (biological) sense to compare the changes in enzyme activities (if that is what the % numbers on the plots are indeed referring to..?). perhaps related: What is the importance of “the numerical value represents the residual activity in the presence of MG132”?

We apologize for the confusing explanation. The numerical value represents the percentage of residual activity in the presence of MG132. It was calculated by dividing the activity with MG132 by the activity without MG132 (P. 5, L. 248-250). This number indicates the inhibitory effects against MG132. The activity in the absence of MG132 varies for each strain. Therefore, we believe that it is more informative to show the readers the effect of inhibition by MG132 based on relative values rather than absolute values. Also, for the above reasons, the discussion of absolute differences in the presence of MG132 does not make sense and has not been included in the text. Hence, we dared to use Student’s t-test instead of one/two-way ANOVA for statistical processing. We appreciate your understanding.

 

6. Line 357: Remove the word “carefully”. (unless you mean to suggest that the other analyses were not performed with similar care!).

              We have simply deleted it.

 

7. Fig 4EF and 6FG, Fig S1B: It’s not clear what the 2 way ANOVAs used here are comparing. A repeated measures ANOVA that takes into account that multiple measures were taken from the same sample (acknowledging that variances between time points are not independent) would be more appropriate. This could also include an interaction term between sample and timepoint. If the shape of the time course is not of interest, then a simple one-way anova comparison of the end time point to show that more CO2 is produced in the rox1∆ strain would be appropriate. Also, the key states that the data are presented with ±SD though no error bars are presented on the graphs.

              We are not experts in statistics, but we understand the difference between one-way ANOVA and two-way ANOVA. Two-way ANOVA is used to analyze the effects of two categorical independent variables (factors) on a continuous dependent variable. Since these fermentation data are time-course experiments, we think that two-way ANOVA is appropriate for statistical analysis.

              The sentence “Data are presented as means ± SD (n = 3)” in the legend was our simple mistake. We performed three independent experiments, but due to the large variation in absolute values, only representative results were presented in the figures. We have now included the correct sentence in the revised manuscript (P. 13, L. 480-481; P. 13, L. 484-486; P. 17, L.596-598; P. 17, L. 602-603; the legend of Fig. S1B).

 

8. RNAseq: The bioinformatic pipeline for how the RNAseq reads were analyzed should be included.

Quality control process was applied to the raw paired-end sequence reads that passed the FastQC. Low quality (< 20) bases and adapter sequences were trimmed by Trimmomatic software (Version 0.38). Reads were aligned to the S. cerevisiae reference genome (version R64-1-1) using RNA-seq aligner HISAT2 (Version 2.1.0). The coverage was 96-98%. The expression count matrix was generated using featureCounts (Version 1.6.3). The raw read counts were normalized with transcripts per million (TPM). We now added bioinformatics information in the Materials and Methods section (P. 6, L. 273-279).

 

9. The title of Table S3 is misleading as it doesn’t indicate the actual criteria for the listed genes. The authors should provide a table with all the genes presented in Fig S3, with appropriate statistics (log²fold change, p values ,etc).

              According to your suggestion, we have provided a list of all genes under each condition with relative expression levels (new Table S3-6).

 

10. I’m surprised that expression of COX5b is not in the list of genes that are differentially expressed between all 3 conditions. Does COX5b (or any of the hypoxic genes known to be repressed by Rox1) appear in any component of the venn diagram shown in Fig S3? GO ontology analyses could help to provide connections to genes and pathways involved, and I’m somewhat surprised these aren’t reported. I’m also wondering what expression differences between wt and rox1∆ cells at steady state (where cells are most likely undergoing diauxic shift) vs cells wakening from a frozen (quiescent?) state actually mean. Could the authors comment on this?

As you pointed out, the expression level of COX5b (shown as YIL111W in new Table S3-4) in rox1Δ was more than 2-fold compared to WT before freezing and at 20 min after thawing. However, the expression level was 1.8-fold at 40 min after thawing. Since we extracted genes whose expression was increased more than 2-fold in rox1Δ compared to the WT in all conditions, COX5b was not selected as one of candidate genes.

GO analysis is useful for obtaining a picture of the cellular response, but it is not suitable for analyzing individual genes. We performed the RNAseq analysis to search for individual genes associated with the proteasome and freezing stress. Our strategy in analyzing the RNAseq data was to identify candidate genes based on expression patterns and chymotrypsin activity in the presence of MG132 rather than GO analysis. Therefore, GO analysis is not necessary for this study. To allow readers interested in the cellular response to freezing stress or Rox1 itself to perform GO analysis by themselves, we have already deposited the RNAseq data and provided a list of gene names (new Table S3-6).

 

11. Line 456-458: states that overexpression of ANB1 reduces MG132 sensitivity to a level that was similar to the rox1∆. This isn’t shown, as far as I can tell. The authors should direct the reader to this data, or if not shown, state that is it not shown (if that is permissible by the journal) and remove the “more importantly” qualifier.

              We apologize for some confusion. The previous sentence was unclear. The term “MG132 sensitivity” does not refer to cells, but to sensitivity to chymotrypsin activity (Fig. 5A). The original sentence has been rectified as follows: “More importantly, only the strain overexpressing ANB1 exhibited decreased sensitivity of chymotrypsin-like activity to MG132.” You can now confirm the revised manuscript (P. 13, L. 505-507).

 

12. If ANB1 overexpression leads to a similar phenotype as a rox1 mutant, then I am thinking back to the original screen. All 3 of the sequenced strains had rox1 mutations (and with identical mutations), suggesting to me that the screen might be saturated. It would be interesting to know if the other un-sequenced MT mutants also had rox1 mutations or if there are potentially additional as yet uncharacterized mutations in this screen (such as in the promoter of ANB1). I guess a good old fashioned complementation test could determine this, or Sanger sequencing of candidate genes. I don’t suggest that this needs to be done for this publication, but maybe the authors could reflect on the experiments presented here, including the screen, why tryspin/caspase/chymotrypsin-like activities differ in the various genotypes and treatments, big picture with expression differences, etc, in the Discussion before delving into potential next experiments involving ANB1/HYP2, Fub1.

              Thank you for your insightful suggestion. We are interested in determining if the other MT mutants also had rox1 mutations. Currently, we have attempted Sanger sequencing of the ROX1 ORF in the other MT mutants. As you mentioned, we are also interested in targeting the promoter of the ANB1 gene. Although we do not yet have solid data, we believe that these experiments will help us understand the overall picture of the proteasome system and freezing stress tolerance. Since we do not want to confuse the readers with further overstatements, we hope to provide the big picture in a future publication after we obtain more reliable data.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

In the reviewed article "Enhancing freezing stress tolerance through regulation of the ubiquitin-proteasome system in Saccharomyces cerevisiae" submitted by Tanahashi et al., the authors aim to minimize the negative effects of freezing on the performance of baker’s yeast strains during subsequent baking. As a general remark, the article is professionally written, and the overall appearance is entirely satisfactory. Moreover, the work fits the journal's scope well and might be of interest to a broad readership. Nonetheless, some questions and uncertainties arose while reviewing the manuscript, which should be addressed and/or answered by the authors before the article might be accepted (minor revision). I hope my comments below are helpful and will contribute to improving the quality of the manuscript.

General comment:

1. The authors use extensive metabolic engineering and strain characterization techniques to obtain candidates with improved freezing stress capabilities. However, might other approaches, such as the use of cryoprotectants or optimization of yeast culture conditions, not be more straightforward? What advantage would the use of altered strains offer in an industrial process? Please add a short paragraph to the manuscript discussing this point.

Specific comments:

1. Lines 34-35: Add a reference.

2. Lines 44-45: It is written that "Therefore, yeast strains used in frozen dough baking must have advanced freezing stress tolerance." This sounds like it is an indispensable requirement. However, it would only be beneficial. Hence, please rephrase the statement.

3. Lines 55-85: While the authors do a proper job explaining the ubiquitin–proteasome system in the introduction, it would be much easier accessible to prospective readers when a figure for the main mechanisms and pathways would be added to the manuscript. Hence, it would be much appreciated if the authors added one additional figure, illustrating the major steps of the mechanisms outlined in lines 55-85.

4. Section 2.1: How were the different media sterilized? Please add this information to the manuscript.

5. Lines 159-160: How were the cells cultivated (system, volume, temperature, pH, rpm, time)? Please add this information to this section.

6. Figure S1: Why was the fermentation ability of the 10-day-old cells not tested and compared (also for the other generated strains, a freezing incubation time of 5 days was used, why)? Please comment. If further data from 10 days freezing are available, please add them to the manuscript

7. Lines 238-245 and Figure S2: How can it be concluded that the observed effects can solely be attributed to the targeted gene deletions? Did the generated strains show any further differences from the WT strain that also might have contributed to the lower or higher freezing tolerance? Please comment and add additional information when available.

 

Author Response

Reviewer: 4

In the reviewed article "Enhancing freezing stress tolerance through regulation of the ubiquitin-proteasome system in Saccharomyces cerevisiae" submitted by Tanahashi et al., the authors aim to minimize the negative effects of freezing on the performance of baker’s yeast strains during subsequent baking. As a general remark, the article is professionally written, and the overall appearance is entirely satisfactory. Moreover, the work fits the journal's scope well and might be of interest to a broad readership. Nonetheless, some questions and uncertainties arose while reviewing the manuscript, which should be addressed and/or answered by the authors before the article might be accepted (minor revision). I hope my comments below are helpful and will contribute to improving the quality of the manuscript.

We are really grateful for your very positive comments, which indeed were helpful in improving our manuscript.

 

General comment:

1. The authors use extensive metabolic engineering and strain characterization techniques to obtain candidates with improved freezing stress capabilities. However, might other approaches, such as the use of cryoprotectants or optimization of yeast culture conditions, not be more straightforward? What advantage would the use of altered strains offer in an industrial process? Please add a short paragraph to the manuscript discussing this point.

Yeast cells use compatible solutes, such as trehalose, glycerol, and proline, to tolerate freezing stress by preventing intracellular ice crystal formation. The addition of these compounds to the culture medium and the optimization of culture conditions can increase the amount of compatible solutes inside the cells, making them more tolerant to freezing stress. However, the addition of such solutes can raise production costs and may alter the taste, flavor, and texture of the bread, which may not be acceptable to both producers and consumers. An alternative approach to obtaining strains with numerous qualities is breeding. Breeding strategies that use metabolic engineering to target specific cellular systems have less impact on other systems that affect bread taste and aroma. Furthermore, the generated strains can be distributed worldwide, allowing them to make high-quality bread without changing their traditional bread-making methods. Therefore, Breeding strategies with metabolic engineering have advantages in obtaining practical candidates with improved freezing stress capabilities. We have now included the above statements in the revised manuscript (P. 3, L. 101-112).

 

Specific comments:

1. Lines 34-35: Add a reference.

We obtained the related information “The baking industry has been expanding, with market size reaching about US$220 billion in fiscal year 2022” from our coauthors affiliated with TableMark Co., Ltd. They conducted an independent analysis of the bakery market. As this information is not available to the public, we have included “TableMark Co., Ltd, personal communication” instead of an official reference (P. 1, L. 36).

 

2. Lines 44-45: It is written that "Therefore, yeast strains used in frozen dough baking must have advanced freezing stress tolerance." This sounds like it is an indispensable requirement. However, it would only be beneficial. Hence, please rephrase the statement.

According to your suggestion, we have changed the sentence to ‘Therefore, yeast strains with advanced freezing stress tolerance would be promising for use in frozen dough baking.” (P. 2, L. 46-47).

 

3. Lines 55-85: While the authors do a proper job explaining the ubiquitin–proteasome system in the introduction, it would be much easier accessible to prospective readers when a figure for the main mechanisms and pathways would be added to the manuscript. Hence, it would be much appreciated if the authors added one additional figure, illustrating the major steps of the mechanisms outlined in lines 55-85.

              A schematic diagram might be helpful. However, I'm not an expert on the ubiquitin-proteasome system, and I don't think an overly indulgent diagram would benefit readers. Therefore, I have cited a useful review article (Marshall et al., Front Mol Biosci, 2019) on the ubiquitin-proteasome system (P. 2, L. 71, 77, 86; P. 19, L. 743-744). Your understanding in this response would be greatly appreciated.

 

4. Section 2.1: How were the different media sterilized? Please add this information to the manuscript.

We have added the sentence “All media were sterilized at 121ºC for 20 min.” in the Materials and Methods section (P3, L144).

 

5. Lines 159-160: How were the cells cultivated (system, volume, temperature, pH, rpm, time)? Please add this information to this section.

We have now added additional information as follows: “Yeast strains were cultured in 50 mL of SC or SC-Ura medium until reaching the stationary phase (30°C, 200 rpm), and cells with an OD₆₀₀ units of 20 were collected. The cells were then washed with distilled water, suspended in distilled water, and incubated at -30°C for 5 days. After thawing at 25°C for 20 min, cells were suspended in 25 mL of the model dough medium. The fermentation process was continuously monitored by measuring the volume of evolved CO₂ gas using a Fermograph II apparatus (ATTO Technology Inc., Amherst, NY, USA) under static incubation conditions at 30°C.” We have included the above information in the revised manuscript (P. 4, L. 196-203).

 

6. Figure S1: Why was the fermentation ability of the 10-day-old cells not tested and compared (also for the other generated strains, a freezing incubation time of 5 days was used, why)? Please comment. If further data from 10 days freezing are available, please add them to the manuscript.

              We did not measure the fermentation ability of the 10-day-old cells because our data showed no significant difference in cell viability between the 5-day-old and 10-day-old cells. However, exploring how long cells can be preserved for baking is an interesting issue for future research. We aim to provide relevant data in future publications.

 

7. Lines 238-245 and Figure S2: How can it be concluded that the observed effects can solely be attributed to the targeted gene deletions? Did the generated strains show any further differences from the WT strain that also might have contributed to the lower or higher freezing tolerance? Please comment and add additional information when available.

We only have the growth data for WT, rpn4Δ, and ubr2Δ strains cultured under non-stressed conditions. There was no significant difference in growth between the strains. Therefore, the difference in cell viability of each strain may be attributed to the targeted gene deletions. As you pointed out, the evidence may be weak, but the data encourage our motivation to start the screening shown in this work. We have now included growth data for WT, rpn4Δ, and ubr2Δin the revised manuscript (P. 4, L. 191-194; P. 6, L. 296-297; Fig. S2A).

Author Response File: Author Response.docx