Rvb1/Rvb2 proteins couple transcription and translation during glucose starvation

  1. Yang S Chen
  2. Wanfu Hou
  3. Sharon Tracy
  4. Alex T Harvey
  5. Vince Harjono
  6. Fan Xu
  7. James J Moresco
  8. John R Yates
  9. Brian M Zid

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

During times of unpredictable stress, organisms must adapt their gene expression to maximize survival. Along with changes in transcription, one conserved means of gene regulation during conditions that quickly repress translation is the formation of cytoplasmic phase-separated mRNP granules such as P-bodies and stress granules. Previously, we identified that distinct steps in gene expression can be coupled during glucose starvation as promoter sequences in the nucleus are able to direct the subcellular localization and translatability of mRNAs in the cytosol. Here, we report that Rvb1 and Rvb2, conserved ATPase proteins implicated as protein assembly chaperones and chromatin remodelers, were enriched at the promoters and mRNAs of genes involved in alternative glucose metabolism pathways that we previously found to be transcriptionally upregulated but translationally downregulated during glucose starvation in yeast. Engineered Rvb1/Rvb2-binding on mRNAs was sufficient to sequester mRNAs into mRNP granules and repress their translation. Additionally, this Rvb tethering to the mRNA drove further transcriptional upregulation of the target genes. Further, we found that depletion of Rvb2 caused decreased alternative glucose metabolism gene mRNA induction, but upregulation of protein synthesis during glucose starvation. Overall, our results point to Rvb1/Rvb2 coupling transcription, mRNA granular localization, and translatability of mRNAs during glucose starvation. This Rvb-mediated rapid gene regulation could potentially serve as an efficient recovery plan for cells after stress removal.

Article activity feed

  1. Version published to 10.7554/elife.76965 on eLife
  2. Review Commons

    Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    This is a very interesting paper with novel observations. The authors find that, in yeast, Rvb1/2 AAA+ ATPases couple transcription, mRNA granular localization, and mRNAs translatability during glucose starvation. Rvb1 and Rvb2 were found to be enriched at the promoters and mRNAs of genes involved in alternative glucose metabolism pathways that are transcriptionally upregulated but translationally downregulated during glucose starvation.

    The following are some comments

    Introduction

    1. "Structural studies have shown that they form a dodecamer comprised of a stacked Rvb1 hexametric ring and a Rvb2 hexametric ring." o Rvb1 and Rvb2 form a heterohexameric ring with alternating arrangement (not homohexamers that stack on top of each other as suggested by this sentence) o In yeast, they oligomerize mostly as single hexametric rings, with dodecamers reported being less than 10% in frequency in vivo (eg Jeganathan et al. 2015 https://doi.org/10.1016/j.jmb.2015.01.010)

    Results Section: Rvb1/Rvb2 are identified as potential co-transcriptionally loaded protein factors on the alternative glucose metabolism genes

    1. "These two proteins are generally thought to act on DNA but have been found to be core components of mammalian and yeast cytoplasmic stress granules" • These two papers extensively show Rvb1/Rvb2 localization to granules/condensates under stress/nutrient starvation conditions and should be cited. The Rvb1/2 foci were named Rbits: i. Rizzolo et al. 2017 https://doi.org/10.1016/j.celrep.2017.08.074 ii. Kakihara et al. 2014 https://doi.org/10.1186/s13059-014-0404-4
    2. "a portion of them becomes localized to cytoplasmic granules that are not P-bodies in both 15-minute and 30-minute glucose starvation conditions (Figure 1-figure supplement 2)" • Supplement figure 2 only includes results under 30-min glucose starvation, no 15-min data was shown
    3. Figure 1C, unclear whether p-value here is for FC of GLC3 over HSP or FC of GLC3 over CRAPome. In addition, both FC datasets should have p-values.

    Section: Rvb1/Rvb2 are enriched at the promoters of endogenous alternative glucose metabolism genes

    1. "Here, we performed ChIP-seq on Rvb1, Rvb2, and the negative control Pgk1 in 10 minutes of glucose starvation (Figure 2-figure supplement 3, left)" • Unclear what figure is being referred to, panel A or panel B?
    2. "Structural studies have shown that Rvb1/Rvb2 can form a dodecamer complex. Their overlapped enrichment also indicates that Rvb1 and Rvb2 may function together." • They function together regardless of forming a dodecamer or not, as they assemble as heterohexamers

    Section: Engineered Rvb1/Rvb2 tethering to mRNAs directs the cytoplasmic localization and repressed translation

    1. Does binding of any protein to PP7 loop in this construct alter cytoplasmic fate? A control such as GFP-CP or any other protein attached to CP should be used.
    2. No statistical analysis was done for Figure 4E quantification
    3. "Results showed that after replenishing the glucose to the starved cells, the translation of those genes is quickly induced, with an ~8-fold increase in ribosome occupancy 5 minutes after glucose readdition for Class II mRNAs (Figure 4-figure supplement 9)" o Would be important to see this recovery (increase in translation after glucose replenishment) in one of the reporter constructs used in the paper, such as GL3 promoter driven CFP.

    Section: Engineered Rvb1/Rvb2 binding to mRNAs increases the transcription of corresponding genes

    1. How many biological replicates is in Figure 5B? There does not seem to be any error bars/gray sections indicating sample variation. P-value was also not calculated.

    Reviewer #1 (Significance (Required)):

    This is a very interesting manuscript that ascribes yet another function of the highly conserved RVB1/2 AAA+ ATPases.

    **Referee Cross-commenting**

    All reviewers agree that this an interesting paper. However, the reviewers do suggest specific experiments to verify some of the results. Carrying out these experiments will definitely improve the paper.

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    In their manuscript entitled "Rvb1/Rvb2 proteins couple transcription and translation during glucose starvation", Chen and co-authors use genetics and microscopy to demonstrate how budding yeast regulate cytoplasmic translation by their promoter sequences by two conserved ATPases Rvb1 and Rvb2 during nutrient stress. The authors show that these two ATPases repress translation of target mRNAs and then propose that these two proteins also recruit mRNAs to P bodies. The authors show that Rvb1/2 preferentially binds in the presence of Class II promoters using CoTrIP, that Rvb1/2 binds specifically at Class II promoters using ChIP-Seq, that Rvb1/2 are bound to transcripts with Class II promoters using RIP-Seq, that tethering of Rvb1/2 to a transcript decreases its translatability and that Rvb1/2 binding to a transcript increases its transcript levels by increasing transcription and not slowing mRNA decay.

    The CoTrIP experiment is clever and for the most part well executed. The key conclusions are largely convincing but some clarifications are nevertheless needed (see below). Overall, this paper is well written with well executed experiments that largely support the authors' model. No major additional experiments are needed to support the claims of the paper. There are a few minor concerns that should be addressed before this manuscript gets published. These are: Minor comments:

    1. Are Rvb1/2 components (enriched in) of P bodies? The model proposed by the authors suggests this but no data is show.
    2. Fig. 1A: The model proposed by the authors indicates that Rvb1/2 and other proteins are recruited to the mRNAs in a promoter-dependent manner and not mRNA sequence dependent manner. This is largely supported by the data presented in the paper. However the authors should also discuss the possibility that RNA sequences could nevertheless contribute as only a uniform ORF has been tested. Could the promoter recruit Rvb1/2 similarly regardless of the ORF sequence tested? Please provide a sequence of the uniform ORF, discuss what this "uniformity" means and how a change in RNA sequence could affect the outcome of the experiment outlined in Fig. 1A.
    3. Fig. 2: The authors use Pgk 1 in their ChIP control but this is not the appropriate control for the experiment as Pgk 1 is not nuclear and thus cannot demonstrate non-specific interaction with genetic regions of tested genes. Regardless, the data is convincing enough to support the model that Rvb1/2 are specifically recruited to the promoters of Class II stress-induced genes and not Class I stress-induced genes. GFP-NLS would be a better control. The authors should discuss in their materials and methods section why they chose a cytoplasmic protein for their normalization control but preferably perform ChIP with GFP-NLS or other nuclear protein that could bind to chromatin non-specifically to further demonstrate the specificity of Rvb1/2 enrichment at Class II promoters.
    4. The authors claim that Rvb1/Rvb2 binding to transcripts leads to formation of granules that are non-colocalized with P-bodies and instead co-localized to SGs, but no SG fluorescent marker is used to demonstrate this claim. The authors should provide this data or remove this claim from their manuscript.
    5. Fluorescent images are fuzzy, very small and difficult to interpret. mRNA puncta are difficult to observe and it is hard to conclude which green puncta colocalize with P bodies and which do not (and how frequently). It is difficult to differentiate between the cytoplasm and nucleus. Consider adding DAPI overlay.
    6. The relevance of Figure 2B is not clear - please discuss.
    7. Fig 5A modeling adds little supporting evidence to the entire figure. The experimental results are more convincing. Consider moving to the Supplement.
    8. Fig. 4 and 3B. The authors suggest that Rvb1/2 loaded by the promoters onto the mRNA determine accumulation of mRNAs to P bodies. To test this model, the authors tether Rvb1/2 onto the mRNA using MS2-MCP system and then look for co-localization of the mRNA with P bodies. However, if the authors' model is correct, this experiment could have been achieved already using the constructs in Fig. 3B. The authors should look at the P body localization pattern using chimeras used in Fig. 3B.
    9. Fig. 6: The authors present a model where mRNAs transcribed from Class II promoters are decorated with Rvb1/2 co-transcriptionally, exported into the cytoplasm, recruited to P bodies and translationally repressed. However, this model is not fully supported by the data shown. Specifically, the authors have not shown that localization of mRNAs to P bodies induces translational repression or whether the recruitment is a consequence of this repression. The authors should revise their model to reflect this uncertainty. Also, the numbering of steps 1,2 3 is confusing. Does it imply a temporal sequences? Some of these steps could be occurring simultaneously (like 1 and 3). How does step 3 lead from step 2? Please clarify this model.
    10. Consider showing data-points in Fig 1 figure supplement 1. The box/whisker plot doesn't give a good sense of the enrichment alone
    11. Figure 1 Fig supplement 2 shows that the fluorophore seems to influence the % of cells with foci. Why is this the case?
    12. List gene names in Fig 2 fig supp 5.
    13. Throughout the paper the graph axis labels are very small and difficult to read.
    14. Figure 4 fig supplement 7C and 8E: on the y-axis the legend says proportion of cells (%), so the value on the y-axis might be 25, 50, 75 and not 0.25, 0.50 and 0.75.
    15. The last paragraph of the Introduction (page 2) detailed how Rvb1/Rvb2 are core components of the stress granule. Yet most experiments were conducted to relate Rvb1/Rvb2 with P-bodies. Maybe some information about the known roles Rvb1/Rvb2 play in the P-bodies in the Introduction section could help.

    Reviewer #2 (Significance (Required)):

    Ruvb helicase has been shown to regulate the formation of stress granules in human U2OS cells during oxidative stress (Parker lab, Cell, 2016). Thus, the authors suggest that Rvb proteins could have a broad and conserved role in the formation of RNA granules, which advances our understanding of how biomolecular condensates could form. In addition, translationally-repressed mRNAs have been shown to preferentially recruit to diverse RNA granules, from stress granules P bodies in human cells as well as germ granules in C. elegans and Drosophila. These observations have gained considerable attention in the past 5 years and exact molecular principles behind this phenomenon are not entirely clear. Long and exposed RNA sequences are thought to be sufficient for this enrichment. The authors however suggest that specific proteins (Rvb1/2) could also trigger enrichment either directly by interacting with P bodies or indirectly by repressing translation and exposing RNA sequences. This finding will be particularly relevant to the field of biomolecular condensates. My expertise is in the area of RNA biology, mRNA decay, RNA granules and mRNA localization.

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    Dr. Brian Zid has previously published in Nature that, in response to glucose starvation, promoters of some genes ("class II") can control synthesis of mRNAs that are sequestered in cytoplasmic P bodies or Stress granules, away from the translation apparatus. In this paper, his group reports about the underlying mechanism. They have found proteins that bind preferentially class II promoters as well as their transcripts and are capable of repressing their translation and stimulating their assembly with P bodies. They found a correlation between the capacity of Rvb1/2 binding to promoters and binding to mRNAs. Using a tethering technique, they found that Rb1/Rvb2 recruitment to reporter mRNA (not class II) led to the association of the transcript with PBs and its translation repression. Interestingly, Binding of Rvb1/Rvb2 to the studied transcript increased transcription of its own gene, probably by remodeling the nearby chromatin. The paper uncovers a mechanism to sequester mRNAs as translationally repressed in RNA granules during starvation and warrants a publication in a good journal, after responding to various comments below.

    1. CoTrIP is a method to identify proteins that differentially bind plasmids carrying different promoters/genes. However, the claim that it identifies proteins bound to nascent mRNAs is an overreach, as the proteins bind both DNA and RNA and the purified plasmid contains both types of nucleic acids. Therefore, the title of section 1 ("Rvb1/Rvb2 are identified as potential co-transcriptionally loaded protein factors on the alternative glucose metabolism genes") should be changed to something like: Rvb1/Rvb2 are identified as proteins that are co-purified with a plasmid expressing alternative glucose metabolism genes. Description of CoTrIP and its results should be discussed throughout the manuscript accordingly.

    2. The engineered Rvb1/Rvb2 tethering to mRNAs of choice is a potentially convincing way to show the causative effect of Rvb1/Rvb2 on RNA performance. Using this method, the authors show that attachment of Rvb1/Rvb2 to an engineered mRNA mediate its association with granules and inhibits its translation. However, this experiment takes Rvb1/2 out of its natural context such that its behavior in this case may not be exemplative of its endogenous function. The authors are encouraged to support their results by depleting Rvbs with AID and examine the outcome of this depletion on PBs formation and translation of class II genes (and class I as controls).

    3. The tethering experiments, shown in Fig. 4, would be more convincing by including an additional control. To rule out the possibility that any bulky protein that is recruited to the 3'-UTR by the PP7 element affects translation (not an unlikely possibility), they want to consider fusing irrelevant protein (e.g., Pgk1p) to CP, in place of Rvb1/2.

    1. The proposal that Rvb1 binds class II transcripts during transcription is a plausible possibility (which I personally believe to represent the reality), but by no means demonstrated. This should be clearly addressed in the paper.

    2. An optional suggestion: The paper can be upgraded by performing ribosome profiling, as shown in Supplemental Fig. 9, after a short depletion of Rvb1/2 by AID (see comment 2). This, in combination with the results already shown in Supp Fig. 9, can demonstrate the role of Rvb1/2 in mRNA storage in granules and in translation shortly after glucose refeeding. The large data sets thus produced (in particular the ratio between depleted and non-depleted signal per each gene) can be used to try correlate the extent of ribosome occupancy (or the above mentioned ratio) with cis-element(s) or known trans-acting elements within the promoters. This may identify elements within the promoters that recruit (directly or indirectly) Rvb1/2. If successful, it can pave the way to demonstrate co-transcriptional RNA binding. I also suggest moving Supp Fig. 9 as an additional panel of the main Fig. 4. Minor point:

    3. The original reference about "mRNA imprinting" was published by Choder in Cellular logistics 2011.

    4. The graph in 5B does not have error bars and the number of replicates is unclear.

    Reviewer #3 (Significance (Required)):

    The paper uncovers a mechanism to sequester mRNAs as translationally repressed in RNA granules during starvation. This significantly advances our understanding of how gene expression in yeast responds to the environment and warrants a publication in a good journal, after responding to the various comments, indicated above. My expertise is regulation of gene expression.

    **Referee Cross-commenting**

    In general all reviewers feel that the paper deals with a significant issue, each from his/her point of view, and is basically of high quality.

    I concur with all the comments of Reviewer 1 and 2. In particular, two comments drove my attention. Reviewer 1: Would be important to see increase in translation after glucose replenishment in one of the reporter constructs used in the paper, such as GL3 promoter driven CFP. Reviewer 2: The authors should look at the P body localization pattern using chimeras used in Fig. 3B.

    There are comments common to more than one reviewer.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Dr. Brian Zid has previously published in Nature that, in response to glucose starvation, promoters of some genes ("class II") can control synthesis of mRNAs that are sequestered in cytoplasmic P bodies or Stress granules, away from the translation apparatus. In this paper, his group reports about the underlying mechanism. They have found proteins that bind preferentially class II promoters as well as their transcripts and are capable of repressing their translation and stimulating their assembly with P bodies. They found a correlation between the capacity of Rvb1/2 binding to promoters and binding to mRNAs. Using a tethering technique, they found that Rb1/Rvb2 recruitment to reporter mRNA (not class II) led to the association of the transcript with PBs and its translation repression. Interestingly, Binding of Rvb1/Rvb2 to the studied transcript increased transcription of its own gene, probably by remodeling the nearby chromatin.
    The paper uncovers a mechanism to sequester mRNAs as translationally repressed in RNA granules during starvation and warrants a publication in a good journal, after responding to various comments below.

    1. CoTrIP is a method to identify proteins that differentially bind plasmids carrying different promoters/genes. However, the claim that it identifies proteins bound to nascent mRNAs is an overreach, as the proteins bind both DNA and RNA and the purified plasmid contains both types of nucleic acids. Therefore, the title of section 1 ("Rvb1/Rvb2 are identified as potential co-transcriptionally loaded protein factors on the alternative glucose metabolism genes") should be changed to something like: Rvb1/Rvb2 are identified as proteins that are co-purified with a plasmid expressing alternative glucose metabolism genes. Description of CoTrIP and its results should be discussed throughout the manuscript accordingly.
    2. The engineered Rvb1/Rvb2 tethering to mRNAs of choice is a potentially convincing way to show the causative effect of Rvb1/Rvb2 on RNA performance. Using this method, the authors show that attachment of Rvb1/Rvb2 to an engineered mRNA mediate its association with granules and inhibits its translation. However, this experiment takes Rvb1/2 out of its natural context such that its behavior in this case may not be exemplative of its endogenous function. The authors are encouraged to support their results by depleting Rvbs with AID and examine the outcome of this depletion on PBs formation and translation of class II genes (and class I as controls).
    3. The tethering experiments, shown in Fig. 4, would be more convincing by including an additional control. To rule out the possibility that any bulky protein that is recruited to the 3'-UTR by the PP7 element affects translation (not an unlikely possibility), they want to consider fusing irrelevant protein (e.g., Pgk1p) to CP, in place of Rvb1/2.
    4. The proposal that Rvb1 binds class II transcripts during transcription is a plausible possibility (which I personally believe to represent the reality), but by no means demonstrated. This should be clearly addressed in the paper.
    5. An optional suggestion: The paper can be upgraded by performing ribosome profiling, as shown in Supplemental Fig. 9, after a short depletion of Rvb1/2 by AID (see comment 2). This, in combination with the results already shown in Supp Fig. 9, can demonstrate the role of Rvb1/2 in mRNA storage in granules and in translation shortly after glucose refeeding. The large data sets thus produced (in particular the ratio between depleted and non-depleted signal per each gene) can be used to try correlate the extent of ribosome occupancy (or the above mentioned ratio) with cis-element(s) or known trans-acting elements within the promoters. This may identify elements within the promoters that recruit (directly or indirectly) Rvb1/2. If successful, it can pave the way to demonstrate co-transcriptional RNA binding. I also suggest moving Supp Fig. 9 as an additional panel of the main Fig. 4.

    Minor point:

    1. The original reference about "mRNA imprinting" was published by Choder in Cellular logistics 2011.
    2. The graph in 5B does not have error bars and the number of replicates is unclear.

    Significance

    The paper uncovers a mechanism to sequester mRNAs as translationally repressed in RNA granules during starvation. This significantly advances our understanding of how gene expression in yeast responds to the environment and warrants a publication in a good journal, after responding to the various comments, indicated above.

    My expertise is regulation of gene expression.

    Referee Cross-commenting

    In general all reviewers feel that the paper deals with a significant issue, each from his/her point of view, and is basically of high quality.

    I concur with all the comments of Reviewer 1 and 2. In particular, two comments drove my attention. Reviewer 1: Would be important to see increase in translation after glucose replenishment in one of the reporter constructs used in the paper, such as GL3 promoter driven CFP. Reviewer 2: The authors should look at the P body localization pattern using chimeras used in Fig. 3B.

    There are comments common to more than one reviewer.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    In their manuscript entitled "Rvb1/Rvb2 proteins couple transcription and translation during glucose starvation", Chen and co-authors use genetics and microscopy to demonstrate how budding yeast regulate cytoplasmic translation by their promoter sequences by two conserved ATPases Rvb1 and Rvb2 during nutrient stress. The authors show that these two ATPases repress translation of target mRNAs and then propose that these two proteins also recruit mRNAs to P bodies. The authors show that Rvb1/2 preferentially binds in the presence of Class II promoters using CoTrIP, that Rvb1/2 binds specifically at Class II promoters using ChIP-Seq, that Rvb1/2 are bound to transcripts with Class II promoters using RIP-Seq, that tethering of Rvb1/2 to a transcript decreases its translatability and that Rvb1/2 binding to a transcript increases its transcript levels by increasing transcription and not slowing mRNA decay.

    The CoTrIP experiment is clever and for the most part well executed. The key conclusions are largely convincing but some clarifications are nevertheless needed (see below). Overall, this paper is well written with well executed experiments that largely support the authors' model. No major additional experiments are needed to support the claims of the paper. There are a few minor concerns that should be addressed before this manuscript gets published. These are:

    Minor comments:

    1. Are Rvb1/2 components (enriched in) of P bodies? The model proposed by the authors suggests this but no data is show.

    2. Fig. 1A: The model proposed by the authors indicates that Rvb1/2 and other proteins are recruited to the mRNAs in a promoter-dependent manner and not mRNA sequence dependent manner. This is largely supported by the data presented in the paper. However the authors should also discuss the possibility that RNA sequences could nevertheless contribute as only a uniform ORF has been tested. Could the promoter recruit Rvb1/2 similarly regardless of the ORF sequence tested? Please provide a sequence of the uniform ORF, discuss what this "uniformity" means and how a change in RNA sequence could affect the outcome of the experiment outlined in Fig. 1A.

    3. Fig. 2: The authors use Pgk 1 in their ChIP control but this is not the appropriate control for the experiment as Pgk 1 is not nuclear and thus cannot demonstrate non-specific interaction with genetic regions of tested genes. Regardless, the data is convincing enough to support the model that Rvb1/2 are specifically recruited to the promoters of Class II stress-induced genes and not Class I stress-induced genes. GFP-NLS would be a better control. The authors should discuss in their materials and methods section why they chose a cytoplasmic protein for their normalization control but preferably perform ChIP with GFP-NLS or other nuclear protein that could bind to chromatin non-specifically to further demonstrate the specificity of Rvb1/2 enrichment at Class II promoters.

    4. The authors claim that Rvb1/Rvb2 binding to transcripts leads to formation of granules that are non-colocalized with P-bodies and instead co-localized to SGs, but no SG fluorescent marker is used to demonstrate this claim. The authors should provide this data or remove this claim from their manuscript.

    5. Fluorescent images are fuzzy, very small and difficult to interpret. mRNA puncta are difficult to observe and it is hard to conclude which green puncta colocalize with P bodies and which do not (and how frequently). It is difficult to differentiate between the cytoplasm and nucleus. Consider adding DAPI overlay.

    6. The relevance of Figure 2B is not clear - please discuss.

    7. Fig 5A modeling adds little supporting evidence to the entire figure. The experimental results are more convincing. Consider moving to the Supplement.

    8. Fig. 4 and 3B. The authors suggest that Rvb1/2 loaded by the promoters onto the mRNA determine accumulation of mRNAs to P bodies. To test this model, the authors tether Rvb1/2 onto the mRNA using MS2-MCP system and then look for co-localization of the mRNA with P bodies. However, if the authors' model is correct, this experiment could have been achieved already using the constructs in Fig. 3B. The authors should look at the P body localization pattern using chimeras used in Fig. 3B.

    9. Fig. 6: The authors present a model where mRNAs transcribed from Class II promoters are decorated with Rvb1/2 co-transcriptionally, exported into the cytoplasm, recruited to P bodies and translationally repressed. However, this model is not fully supported by the data shown. Specifically, the authors have not shown that localization of mRNAs to P bodies induces translational repression or whether the recruitment is a consequence of this repression. The authors should revise their model to reflect this uncertainty. Also, the numbering of steps 1,2 3 is confusing. Does it imply a temporal sequences? Some of these steps could be occurring simultaneously (like 1 and 3). How does step 3 lead from step 2? Please clarify this model.

    10. Consider showing data-points in Fig 1 figure supplement 1. The box/whisker plot doesn't give a good sense of the enrichment alone.

    11. Figure 1 Fig supplement 2 shows that the fluorophore seems to influence the % of cells with foci. Why is this the case?

    12. List gene names in Fig 2 fig supp 5.

    13. Throughout the paper the graph axis labels are very small and difficult to read.

    14. Figure 4 fig supplement 7C and 8E: on the y-axis the legend says proportion of cells (%), so the value on the y-axis might be 25, 50, 75 and not 0.25, 0.50 and 0.75.

    15. The last paragraph of the Introduction (page 2) detailed how Rvb1/Rvb2 are core components of the stress granule. Yet most experiments were conducted to relate Rvb1/Rvb2 with P-bodies. Maybe some information about the known roles Rvb1/Rvb2 play in the P-bodies in the Introduction section could help.

    Significance

    Ruvb helicase has been shown to regulate the formation of stress granules in human U2OS cells during oxidative stress (Parker lab, Cell, 2016). Thus, the authors suggest that Rvb proteins could have a broad and conserved role in the formation of RNA granules, which advances our understanding of how biomolecular condensates could form.

    In addition, translationally-repressed mRNAs have been shown to preferentially recruit to diverse RNA granules, from stress granules P bodies in human cells as well as germ granules in C. elegans and Drosophila. These observations have gained considerable attention in the past 5 years and exact molecular principles behind this phenomenon are not entirely clear. Long and exposed RNA sequences are thought to be sufficient for this enrichment. The authors however suggest that specific proteins (Rvb1/2) could also trigger enrichment either directly by interacting with P bodies or indirectly by repressing translation and exposing RNA sequences. This finding will be particularly relevant to the field of biomolecular condensates.

    My expertise is in the area of RNA biology, mRNA decay, RNA granules and mRNA localization.

  5. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    This is a very interesting paper with novel observations. The authors find that, in yeast, Rvb1/2 AAA+ ATPases couple transcription, mRNA granular localization, and mRNAs translatability during glucose starvation. Rvb1 and Rvb2 were found to be enriched at the promoters and mRNAs of genes involved in alternative glucose metabolism pathways that are transcriptionally upregulated but translationally downregulated during glucose starvation.

    The following are some comments

    Introduction

    1. "Structural studies have shown that they form a dodecamer comprised of a stacked Rvb1 hexametric ring and a Rvb2 hexametric ring." o Rvb1 and Rvb2 form a heterohexameric ring with alternating arrangement (not homohexamers that stack on top of each other as suggested by this sentence) o In yeast, they oligomerize mostly as single hexametric rings, with dodecamers reported being less than 10% in frequency in vivo (eg Jeganathan et al. 2015 https://doi.org/10.1016/j.jmb.2015.01.010)

    Results Section: Rvb1/Rvb2 are identified as potential co-transcriptionally loaded protein factors on the alternative glucose metabolism genes

    1. "These two proteins are generally thought to act on DNA but have been found to be core components of mammalian and yeast cytoplasmic stress granules" • These two papers extensively show Rvb1/Rvb2 localization to granules/condensates under stress/nutrient starvation conditions and should be cited. The Rvb1/2 foci were named Rbits: i. Rizzolo et al. 2017 https://doi.org/10.1016/j.celrep.2017.08.074 ii. Kakihara et al. 2014 https://doi.org/10.1186/s13059-014-0404-4
    2. "a portion of them becomes localized to cytoplasmic granules that are not P-bodies in both 15-minute and 30-minute glucose starvation conditions (Figure 1-figure supplement 2)" • Supplement figure 2 only includes results under 30-min glucose starvation, no 15-min data was shown
    3. Figure 1C, unclear whether p-value here is for FC of GLC3 over HSP or FC of GLC3 over CRAPome. In addition, both FC datasets should have p-values.

    Section: Rvb1/Rvb2 are enriched at the promoters of endogenous alternative glucose metabolism genes

    1. "Here, we performed ChIP-seq on Rvb1, Rvb2, and the negative control Pgk1 in 10 minutes of glucose starvation (Figure 2-figure supplement 3, left)" • Unclear what figure is being referred to, panel A or panel B?
    2. "Structural studies have shown that Rvb1/Rvb2 can form a dodecamer complex. Their overlapped enrichment also indicates that Rvb1 and Rvb2 may function together." • They function together regardless of forming a dodecamer or not, as they assemble as heterohexamers

    Section: Engineered Rvb1/Rvb2 tethering to mRNAs directs the cytoplasmic localization and repressed translation

    1. Does binding of any protein to PP7 loop in this construct alter cytoplasmic fate? A control such as GFP-CP or any other protein attached to CP should be used.
    2. No statistical analysis was done for Figure 4E quantification
    3. "Results showed that after replenishing the glucose to the starved cells, the translation of those genes is quickly induced, with an ~8-fold increase in ribosome occupancy 5 minutes after glucose readdition for Class II mRNAs (Figure 4-figure supplement 9)" o Would be important to see this recovery (increase in translation after glucose replenishment) in one of the reporter constructs used in the paper, such as GL3 promoter driven CFP.

    Section: Engineered Rvb1/Rvb2 binding to mRNAs increases the transcription of corresponding genes

    1. How many biological replicates is in Figure 5B? There does not seem to be any error bars/gray sections indicating sample variation. P-value was also not calculated.

    Significance

    This is a very interesting manuscript that ascribes yet another function of the highly conserved RVB1/2 AAA+ ATPases.

    Referee Cross-commenting

    All reviewers agree that this an interesting paper. However, the reviewers do suggest specific experiments to verify some of the results. Carrying out these experiments will definitely improve the paper.

  6. Version published to 10.1101/2021.10.17.464753v1 on bioRxiv