Nitrogen availability and TOR signalling are important for preventing catastrophic mitosis in fission yeast

  1. Viacheslav Zemlianski
  2. Anna Marešová
  3. Jarmila Princová
  4. Roman Holič
  5. Robert Häsler
  6. Manuel José Ramos del Río
  7. Laurane Lhoste
  8. Maryia Zarechyntsava
  9. Martin Převorovský

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Mitosis is a critical stage in the cell cycle, controlled by a vast network of regulators responding to multiple internal and external factors. The fission yeast Schizosaccharomyces pombe may demonstrate catastrophic mitotic phenotypes due to mutations or drug treatments. One of the factors provoking catastrophic mitosis is a disturbed lipid metabolism, resulting from e.g. mutations in acetyl-CoA/biotin carboxylase ( cut6 ), in fatty acid synthase ( fas2/lsd1 ), or in the transcriptional regulator of lipid metabolism ( cbf11 ) genes, as well as treatment with inhibitors of fatty acid synthesis. It was previously shown that mitotic fidelity in lipid metabolism mutants can be partially rescued by ammonium chloride. In this study we demonstrate that mitotic fidelity can be improved by multiple good nitrogen sources. Moreover, this rescue is not limited to lipid metabolism disturbances but also applies to a number of unrelated mitotic mutants. Interestingly, the rescue is not achieved by restoring the lipid metabolism state, but rather indirectly. We found that the TOR regulatory network plays a major role in mediating such rescue, highlighting a novel role for TOR in mitotic fidelity.

Article activity feed

  1. Review Commons

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

    Learn more at Review Commons

    Reply to the reviewers

    'The authors do not wish to provide a response at this time.'

  2. 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 #4

    Evidence, reproducibility and clarity

    Summary:

    In this work, Zemlianski and colleagues exploit S. pombe mutations responsible for catastrophic mitoses, in particular those leading to a cut / cut-like phenotypes, whereby cytokinesis takes place without proper DNA segregation, trapping DNA molecules by septum formation in between the two separating cells. The work builds on the team's previous observation that these defects can be alleviated when cells are grown in a nitrogen-rich medium, and motivate their efforts to understand this better. The manuscript is written in a concise, neat and informative manner, and the results are presented clearly, with consistence in the format and the style all along. The analyses appear to have been, in general, conducted under the best standards. The findings are important and the data are of good quality. I have, however, important concerns that will be detailed below, and which, as I hope will be made clear, question the pertinence of including "TOR signaling" in the title, and making a distinction between "good" and "poor" nitrogen sources in the abstract.

    Major comments:

    Results

    The conclusion that the phenotype is suppressed by "good" but not "poor" nitrogen sources is not sufficiently supported. First, this interpretation is based on comparing only two or three sources of each type; Second, the "good" source glutamate needed to be raised for it to have a significant effect; 3) there is a strange datum, as Glu 100 mM in Graph 1D looks exactly the same as Glu 50 mM in Graph 1E, I guess there is a mistake in the plotting; 4) and, more important, the fact that the authors had the nice initiative of reproducing their YES medium experiments for every graph led to the inevitable fact that slightly different values were obtained every time, which is normal. While the values yield very similar data for panels 1B, 1C and even 1D, the frequency of catastrophic mitoses for the cbf11 mutant in YES in panel 1E is much lower than in panel Figure 1B, for example. This has the consequence of making the suppression obtained when adding 'poor' sources, such as proline or uracil, non-significantly different. Thus, the authors conclude that 'poor' nitrogen sources are not good at suppressing the phenotype. I suggest that the authors pool all their YES data (they will have 12 repeats of their experiment) and plot, in a single graph, all the other treatments. By performing the analyses again, using the appropriate statistical test for that, perhaps they will have a surprise. After which, the question is, is it so important to put the emphasis on whether the source is good or poor? The incontestable observation is that, in general, there is clear trend of suppression of the phenotype.

    In Figure 2, images should be shown as an example of what was seen, what was quantified, how the "decrease in nuclear cross-section area" looked like indeed.

    Also, important for Figure 2, the authors used the nuclear cross-section area as a readout for nuclear envelope expansion versus shrinkage. For that, they did not use a fluorescent marker for the nuclear envelope that is continuous, but a nucleoporin (Cut11-GFP). In my experience, nucleoporins being discontinuously distributed throughout the nuclear envelope, the area encompassed by the signal may be underestimated in the event of a strong nuclear envelope deformation, as I have tried to illustrate in the scheme below: I WILL SEND THE SCHEME BY MAIL TO THE EDITOR, AS I CANNOT COPY-PASTE IT IN THE SYSTEM BOX Given that the photos from which the data were retrieved have not been shown, I cannot at present judge whether the use of a nuclear envelope marker providing continuous signals is absolutely necessary or not, and whether this consideration will affect (or not at all) the conclusions.

    The authors do not seem to comment or pay any attention to a very crucial result they obtain: the addition of ammonium to the WT strain has the effect of also restricting the nuclear cross-section area. They indeed say in their text "we did not observe any differences between cultures grown with or without ammonium supplementation (Fig.2)". I guess they refer here to the cbf11 mutant, in which case the sentence is true (although unfair to the WT). But by neglecting that the supplementation with ammonium had the power of reducing the cross-section area of WT nuclei, they are misled (or misleading) in their interpretation. The same, although milder, is true for Figure 5C, where the addition of ammonium to the WT culture does not alter the median value of prophase + metaphase duration, however has the virtue of very much rendering sharp (less scattered) the population of values, suggesting that the accuracy / control of the process is enhanced. What does this mean? I think it should be carefully thought about and considered as a whole.

    In the same line as above, the authors omit the RNA-seq analysis concerning the treatment of the WT with ammonium (Figure 3). This is very important to understand the standpoint of what this treatment elicits. It would also help unravel the observations I mentioned above that the authors did not assess in their descriptions. Also regarding Figure 3, it is completely obscure why the authors decided to show the genes on the right axis, and not others. Knowing how vast the lipid pathways are, there are likely many other hits that could be relevant. A particular thought goes for the proteins in charge of filling lipid droplets, such as sterol- and fatty acid-esterifying enzymes. Unless a very justified reason is provided, the choice at present seems arbitrary and it would be better to show a more unbiased data representation.

    In the same vein, related to the effect of ammonium onto the WT, in Figure S1 (I want to congratulate the authors for showing their 3 experimental replicates), the results very neatly show that ammonium supplementation to the WT leads to a neat and reproducible increase in TAG, a fact on which the authors do not comment. In the mutant, irrespective of ammonium presence or absence, a huge increase in squalene and steryl esters (SE) are seen. I think the work would benefit from actually quantifying the intensity of these bands and thus materializing this in the form of values. TAG, squalene and SE are all neutral lipids, and are all stored within LD to prevent lipotoxicity if accumulated in the endoplasmic reticulum. While ammonium elicits strong TAG accumulation in the WT, this is not the case in the mutant, likely because the massive occupation of LD storage capacity is overwhelmed with squalene and SE. Could this have something to do with the suppression they are studying?

    In the section of results where the authors comment the TLC analysis, they write "suggesting failed coordination between sterol and TAG lipid metabolism pathway". As it stands, the sentence is rather devoid of real meaning and may be even misleading, when considering what I wrote before.

    My biggest concern has to do with the very last part, when they explore the implications of TOR:

    • First, all the data presented in the two concerned panels of Figure 7 (B and C) and of Figure S3 lack the values obtained for the single mutants with which cbf11 was combined. This is not acceptable from a genetic point of view, and may prevent us from having important information. For example: if the authors were right that Tor2/TORC1 is ensuring successful progression through closed mitosis (last sentence of results), then one would predict that the tor2-S allele leads to an increase, already per se, of the frequency of catastrophic mitoses. However, at present, I cannot check that.
    • the authors turn to use a ∆ssp2 mutant to "increase Tor2 activity". However, this is a pleiotropic strategy, as AMP-kinase is the major sensor and responder to energy depletion, frequently triggered by glucose shortage, thus I am not sure the effects associated to its absence can be unequivocally be ascribed to a Tor2 raise.
    • there is a counterintuitive observation: rapamycin, which mimics nitrogen shortage, has the same effect than ammonium supplementation. This is strangely bypassed in the discussion, where the authors wrote "we showed increased mitotic fidelity in cbf11 cells when the stress-response branch of the TOR network was suppressed, either by ablation of Tor1/TORC2 or by boosting the activity of the pro-growth Tor2/TORC1 branch. These data are in agreement with previous findings that Tor2/TORC1 inhibition mimics nitrogen starvation".
    • last, and irrespective of what was said above, the authors conclude that the phenotype suppression is due to "a role for Tor2/TORC1 in ensuring successful progression through mitosis". If, as stated by the authors, Tor1/TORC2 absence not only abrogates Tor1/TORC2 activity, but it simultaneously raises Tor2/TORC1 activity, and if reciprocally Tor2/TORC1 increased activity concurs with Tor1/TORC2 attenuation, it cannot therefore be discerned if the suppression is due to Tor2/TORC1 raise or to Tor1/TORC2 dampening.

    Discussion

    The authors invoke that TOR controls lipin, despite what they go on to dismiss the link between TOR and lipids by saying "we did not observe any major changes in phospholipid composition when cells were grown in ammonium-supplemented YES medium compared to plain YES (Figure S2)", with this reinforcing their conclusion that ammonium does not suppress lipid-related cut mutants through directly correcting lipid metabolism defects. While I agree with that reasoning, I invoke again that they nevertheless neglected the clear change observed in their three replicates (Figure S2) that ammonium addition to WT cells strongly increases the amount of TAG (esterified fatty acids). Since lipin activity promotes DAG formation, which then leads to TAG accumulation, this aspect should not be neglected.

    The emphasis on TOR, which expands several paragraphs of the Discussion, should be revisited if the evidence provided for this part of the data is not reinforced.

    To finish, if I may provide some personal thoughts that may be useful for the authors, I would first remind that TAG storage prevents the channeling of phosphatidic acid towards novel phospholipid synthesis thus antagonizes NE expansion, which agrees with their neglected observation for the WT in Figure 2A. The antagonization of NE expansion can be achieved through autophagy (DOI 10.1038/s41467-023-39172-3; DOI 10.1177/25152564231157706), and indeed rapamycin addition (a very potent inducer of autophagy) also suppressed the cut phenotype (Figure 7A). What is more, in S. cerevisiae, autophagy has been shown as important to transition through mitosis conveniently and to prevent mitotic aberrations (DOI 10.1371/journal.pgen.1003245), and to impose a "genome instability" intolerance threshold by restricting NE expansion (DOI 10.1177/25152564231157706). In the first mentioned work, the authors proposed that autophagy may help raising aminoacid levels, which could assist cell cycle progression. This would have the virtue of reconciling the otherwise counterintuitive observation of the authors that rapamycin, which mimics nitrogen shortage, has the same effect than ammonium supplementation. It could be that ammonium supplementation mimics the downstream signal of a complex cascade initiated by actual aminoacid shortage, known to elicit autophagy-like processes (thus explaining why TAG raise, why the NE does not expand), and may culminate with launching a program for more accurate mitosis and genome segregation. In further support, TORC1 inhibition (as elicited by +rapamycin) is a central node that integrates multiple cues, not only nitrogen availability, but also carbon shortage (DOI 10.1016/j.molcel.2017.05.027), and even genetic instability cues (DOI 10.1016/j.celrep.2014.08.053), perhaps helping unravel why ammonium (via TOR) suppresses very diverse cut mutants, irrespective of whether they stem from lipid or chromatid cohesion deficiencies. These previous works should be considered by the authors.

    Minor

    There was no speculation about why the suppressions are partial.

    Reference 15, cited in the text, is absent from the references list.

    An explanation of which statistical tests were chosen and why they were chosen would be necessary.

    In particular, for the analyses performed for Figure 5, one-way ANOVA should be applied instead of several t-tests.

    A small section in M&M about how data in general was acquired, quantified, plotted and analyzed would be appropriate.

    In the discussion, the sentence "this could mean that the signaling of availability of a good nitrogen source is by itself more important for mitotic fidelity than the actual physical presence of the nutrients" is a rather void sentence. Because, from the point of view of how a cell "works", the signal is important for the basic reason that it is supposed to represent the actual real cue eliciting it.

    In the second part of Results, when the phenotype of cbf11 mutants concerning LD is mentioned, the authors said "aberrant LD content". It would be good if they can mention already at this stage which type of aberration this was: more LD? less LD? bigger? smaller?

    What is the difference between the two SE bands in Figure S2? What exactly does SE-1 and SE-2 mean?

    In Figure 2, the two graphs, presented side by side, would be more easily comparable if they could be plotted with the same y-axis scale.

    In Figure 1A, it would be useful for non-specialists of this phenotype and non-pombe readers to show a control of how it looks to be "normal".

    Referees cross-commenting

    Overall, there is a striking consensus on the need to either address experimentally or remove the emphasis put on the TOR/mitotic fidelity connection, and of clarifying the counter-intuitive notions associated to the results obtained with rapamycin. Also, the need for revisiting / improving / justifying the means by which nuclear envelope deformation is assessed has been raised at least twice. I therefore guess that the common guidelines for improving this manuscript are clearly established.

    Significance

    In view of all of the above, my feeling is that the authors have put the accent on the TOR message, which is weak, while they have less put the accent on very strong and elegant findings they do: The authors discover that the suppression of cut(-like) mutant phenotype by addition of NH4 is not due to a correction in lipid metabolism defects, suggesting that the effect is indirect. In support, cut-like mutants whose molecular defect stems from lipid-unrelated defects are also suppressed by ammonium addition. What is more, the authors refine the type of cut-like mutants susceptible of being "corrected" by ammonium addition, finding a "novel definition of cuts" that invoke a temporal rule. This important observation has relevant implications:

    • the long-standing interpretation (commented by the authors) that lipid-related cut mutants are defective because of insufficient synthesis of lipids to be able to grow their nuclear envelope membranes seems now inappropriate in light of their data;
    • this has the immediate implication that perhaps the importance of nitrogen supplementation for accurate mitosis is no longer a fact that may apply only to (yeast) organisms performing closed mitosis, which may broaden the implications of their finding substantially;
    • the nature of the temporal ruler they discover that makes defects appearing early susceptible of being suppressed by nitrogen supplementation deserves analysis in further works, thus opening an immediate perspective.
  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

    In this manuscript, Zemlianski et al conducted careful analysis of a group of lipid metabolism mutants exhibiting mitotic defects. They demonstrate that supplementing a good nitrogen source in the medium can rescue the mitotic defects in these mutants. Notably, this rescue occurs independently of addressing lipid composition defects or altering the expression of lipid metabolism genes. Furthermore, the study implicates TORC1 activity as a key player in integrating nitrogen availability for the effective execution of mitosis. Despite well-controlled and meticulously executed experiments, the overall study lacks comprehensiveness and appears to add to the list of existing reports without offering mechanistic insights into the unexplained impact that the TOR pathway (or nitrogen source) has on mitosis.

    Major Comments:

    1. The discussed link between Tor and mitosis is not a novel finding. An yet unexplained link between mitosis and the tor pathway has been previously reported by Yanagida lab and several years later by the Hauf Lab. Recent reports from the Hauf lab suggest that the relation of Tor to mitotic fidelity could be associated with the translational sensitivity of mitotic proteins to tor pathway or more directly to translational response to nitrogen availability for growth. Therefore, based on these leads it would be informative to see if the authors could expand on this idea and explore more on the mechanistic aspect of how nitrogen availability which feeds into tor functionality can influence mitotic progression.

    Based on the results presented here, it is reasonable to assume that in the lipid metabolism mutants which are rescued on nitrogen supplementation, TORC1 would be rendered inactivate as these cells are apparently nitrogen starved. TORC1 inactivation is known to downregulate translation and could impact the levels of critical mitotic genes. Therefore, it warrants the testing of this possibility.

    1. TORC1 is known to restrain mitotic progression by opposing securin-separase and TORC2 to aid G2 to M transition by regulating the timing of Cdc2 de-phosphorylation. Earlier studies have seen rescue of mitotic defects in securin and separase by tor2 mutants (TORC1). However, here the rescue is executed by increasing the activity of TORC1 or impairing the TORC2 pathway by mutations in tor1. It might be good to present this result in context of previous reports and discuss how mitotic defects exhibited by lipid metabolism defects differ from those of mutants in core mitotic pathway such as separase and securin. The current discussion section does not explicitly explain this difference.

    Significance

    The inquiry central to the present study, namely the investigation into the impact of the TOR pathway on the proficient execution of mitosis, holds significant scientific relevance. Unraveling the mechanisms through which TOR enhances mitotic fidelity has the potential to enhance current drug interventions and pave the way for the development of informed and efficient therapeutic strategies, particularly in cancer.However, in the current form, the study lacks mechanistic insights and does not add much to the already known literature as I have detailed above in my comment.

  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

    Zemlianski et al. present an analysis of the interaction between various mitotic phenotypes and Tor1-dependent nitrogen signaling in fission yeast. They make two interesting observations. First, the mitotic disruption caused by defects in lipid metabolism are not due to direct effects of such defects on mitotic mechanisms because the mitotic phenotypes can be suppressed by nitrogen supplementation without resolving the lipid metabolism defects. Second, the effects of nitrogen supplementation are due to nitrogen's effect on TOR signaling, not to the direct effect of the nutrient, because TOR mutants have similar effects. The work is straight forward, appears to be well done, and the conclusions are well supported by the data.

    The one part of the manuscript that I do not understand is the effect of rapamycin on mitotic fidelity. The presented genetics suggest that nitrogen increases mitotic fidelity by activating TORC1. However, rapamycin inhibits TORC1, yet also increases mitotic fidelity. The authors need to state and address this apparent contradiction much more directly than they currently do (unless I badly misunderstand something, and then they need only explain it to me).

    Referees cross-commenting

    I agree with the comments of the other reviews. They all seem reasonable and addressable by the authors.

    Significance

    The significance of the work is limited by the lack of mechanistic insight or even a plausible hypothesis as to how TOR-dependent nitrogen signaling is affecting mitotic fidelity. In something of an understatement, the authors note that "the exact mechanism of the nitrogen-mediated rescue of mitotic fidelity remains to be characterised in detail". Until that mechanism can be at least suggested, these observation do not provide much biological insight into the question of mitotic regulation.

    The work will be of interest to workers specifically involved in the regulation of mitotic fidelity in yeast, but, until more mechanistic insight can be generated, not much beyond that group.

  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

    The authors show that mitotic fidelity can be improved by a good nitrogen source. Such 'rescue' applies to a range of unrelated mitotic mutants in fission yeast (S.pombe). Rescue appears not to be achieved by restoring lipid metabolism. Instead they argue for an indirect mechanism of suppression, and find that the TOR signalling network is involved. The paper is well written and the data clearly presented.

    Major comments:

    • most claims and conclusions are supported by the data. The catastrophic mitosis (Fig 1A) should be better described in this manuscript, rather than referring the reader to Ref. 18.
    • 'rescue' is often used in figure legends and text (eg. Fig 5-7 titles). Is this the most appropriate word? In most cases this rescue is partial. Perhaps 'suppresses' is more appropriate?
    • The authors write in the results: "taken together the ammonium-mediated rescue of mitotic defects.........seems to operate early in the cell cycle, prior to anaphase." Can they be more precise here? If cell cycle checkpoints were activated, to lengthen G2 or early M, this may not always help reduce chromosome mis-segregation. The authors have previously shown that combining a sac mutation with cbf11 did not rescue mitotic defects (ref 18). Have the authors tested these double mutants to see if the prolonged mitosis observed in cbf11 is shortened? Have other checkpoints been tested, apart from the sac?
    • "Figure 7. The TOR network is critical for the ammonium-mediated rescue of Δcbf11 mitotic defects." The data shows that inhibiting the TOR network (rapamycin) has a similar impact to ammonia. These are not additive, and it is argued that both rapamycin and ammonium must affect the same pathway. However, they do not test the impact of nitrogen sources in the genetic tor mutant backgrounds. Where is the mechanistic evidence that tor signalling is required for the ammonium-mediated rescue?
    • Optional: can the authors support their interpretation by providing some biochemical evidence that the tor signalling pathway is active in relevant conditions. For example, is tor signalling reduced when a good nitrogen source is added? Is tor signalling enhanced in a cbf11 mutant?

    Minor comments:

    • Methods: how was the area of nuclear cross-section measured (for Figure 2)?
    • I question whether the statistical t-tests used are always appropriate. In some experiments (eg. Fig2 and 4C) should ANOVA be performed? I am no expert in this, but the authors should get advice.

    Significance

    The data presented will be of interest to those studying cell division, lipid homeostasis and TOR signalling networks. However, in my opinion the mechanistic link with TOR signalling (Fig 7) should be strengthened.

    I am an expert on mitotic regulation and chromosome segregation in yeast.

  6. Version published to 10.1101/2023.12.25.573293v1 on bioRxiv