Tubulin autoregulation controls the biosynthesis of γ-tubulin to ensure mitotic fidelity

  1. Maxim Assaf
  2. Chahrazed Lacheheub
  3. Ivana Gasic

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Microtubule organization relies on the precise control of tubulin abundance to ensure accurate cytoskeletal function and faithful cell division. While α- and β-tubulin levels are controlled by a well-characterized autoregulatory pathway that triggers co-translational mRNA decay in response to excess soluble tubulin, how cells regulate the abundance of the core microtubule nucleator γ-tubulin has remained unclear. Here, we show that γ-tubulin is regulated by the canonical tubulin autoregulatory machinery. We find that γ-tubulin-encoding mRNAs are downregulated in response to elevated soluble αβ-tubulin levels. This regulation requires TTC5 to co-translationally recognize a conserved amino-terminal MPREI motif in nascent γ-tubulin proteins, and further recruit SCAPER and CCR4-NOT complex, targeting γ-tubulin mRNAs for decay. Disruption of this regulatory mechanism elevates γ-tubulin protein levels, increases centrosomal microtubule nucleation output, and compromises mitotic fidelity. Together, our findings establish γ-tubulin as a previously unrecognized substrate of tubulin autoregulation and reveal coordinated control of tubulin biosynthesis as a key mechanism for tuning microtubule nucleation and ensuring accurate chromosome segregation.

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 #3

    Evidence, reproducibility and clarity

    Microtubule dynamics depend on the concentration of soluble αβ-tubulins. When cells detect an increase in soluble αβ-tubulin, they trigger degradation of tubulin mRNAs via a process termed tubulin autoregulation. In this pathway, the ribosome-associated factor TTC5 recognizes nascent amino-terminal autoregulatory MREC and MREI motifs in αβ-tubulins. Upon recognition of the nascent tubulin chain, TTC5 recruits the adaptor protein SCAPER, which in turn engages the CCR4-NOT complex to promote mRNA decay. While this mechanism has been well characterized for α- and β-tubulin transcripts, how cells regulate the abundance of the core microtubule nucleator γ-tubulin remains poorly understood. Here, Assaf et al. show that γ-tubulin-encoding mRNAs are also downregulated through the same tubulin autoregulation pathway (the TTC5-SCAPER-CCR4-NOT axis) in response to elevated soluble αβ-tubulin. They demonstrated that disruption of this pathway, through knockout or mutation of TTC5, SCAPER, or CNOT11, leads to increased γ-tubulin mRNA levels following treatment with the microtubule destabilizer combretastatin A-4 (CA4). Furthermore, mutation of the autoregulatory MPREI motifs in TUBG1 and TUBG2 (TUBGR3H) results in a modest increase in γ-tubulin protein levels. This elevation enhances centrosomal γ-tubulin during mitosis, increases microtubule nucleation capacity (as measured by microtubule regrowth after cold treatment), and ultimately reduces mitotic fidelity.

    Major comments:

    1. The authors concluded that tubulin autoregulation-associated mitotic defects are largely driven by elevated γ-tubulin protein levels. However, it is somewhat surprising that such a modest increase in γ-tubulin protein level (1.10-, 1.18-, 1.23-fold in TTC5 KO, TUBGR3H, TTC5 KO+TUBGR3H cells, respectively) leads to chromosome alignment and segregation defects. Given that γ-tubulin is a relatively abundant protein, with only a small fraction localized at centrosomes [PMID: 27539480], it remains unclear whether this magnitude of increase is sufficient to account for the observed mitotic defects. To more directly test whether a modest increase in γ-tubulin is sufficient to impair mitotic fidelity, it would be informative to perform live-cell imaging of γ-tubulin-GFP and chromosomes in cells moderately overexpressing wild-type or R3H γ-tubulin (as in Figure 3A), in the presence and absence of siTUBG1. This approach would help determine whether a comparable increase in γ-tubulin levels is sufficient to induce chromosome missegregation. Alternatively, the authors should consider tempering or revising their conclusion.

    Minor comments:

    1. On page 4, the authors state that measuring TUBG pre-mRNA and mRNA levels allows for distinguishing transcriptional (pre-mRNA) from post-transcriptional (mRNA) regulation, citing reference [33: PMID: 15367667]. While this approach is appropriate, it is unclear why reference [33: PMID: 15367667] is cited here, as it does not appear to directly describe this methodology. Please provide a more relevant reference or clarify the rationale for this citation.
    2. For the immunoprecipitation shown in Fig. 3B, an appropriate negative control is needed. For example, a parental cell line lacking γ-tubulin-FLAG expression should be included to assess background binding.
    3. Comparing γ-tubulin localization at centrosomes between Fig. 3C, D and Fig. 4B, the difference between parental and mutant cell lines (TTC5 KO, TBUGR3H and TTC5KO + TBUGR3H) appears more pronounced in Fig. 4B. It would be helpful if the authors could quantify centrosomal γ-tubulin localization in Fig. 4B to facilitate a direct comparison. In addition, could the authors comment on whether the cold treatment used in Fig. 4B might influence the soluble αβ-tubulin levels? If so, this could potentially enhance tubulin autoregulation in parental cells, leading to reduced γ-tubulin mRNA levels, while this response would be impaired in the mutant cell lines. Such an effect might contribute to the increased difference in centrosomal γ-tubulin observed under these conditions.

    Significance

    The manuscript provides significant and new mechanistic insights into microtubule regulation by identifying γ-tubulin as a target of the microtubule autoregulation pathway. It further suggests a new model that cells coordinately adjust both microtubule building blocks and nucleation capacity in response to changes in soluble tubulin pools through a common molecular machinery. The data are clearly presented, the experiments are rigorous, including the well-controlled cell lines, and the manuscript is well written and easy to follow. Based on its quality, novelty, and significance, I strongly support publication.

  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 #2

    Evidence, reproducibility and clarity

    In this manuscript, Assaf et al. address an important question by investigating the post-transcriptional-regulation of gtubulin mediated by the tubulin autoregulation mechanism and its functional role in microtubule nucleation and chromosome segregation. Autoregulation of a/b-tubulin has been shown previously, including by the current authors, but whether other tubulin genes are regulated in a similar manner was unknown. Using cell culture models and treatment with microtubule-destabilising/stabilising drugs, they demonstrate that cells regulate gtubulin levels in response to changes in soluble a-b tubulin levels. This regulation depends on the tubulin autoregulation mechanism that the authors previously identified for a/b-tubulin and the authors identify a similar motif in the N-term of gtubulin that is recognized by TTC5. By using a R3H g-tubulin mutant, which specifically disrupts the binding of TTC5, they examine the consequences of deregulated gtubulin biosynthesis. They showed that loss of g-tubulin mRNA regulation moderately increases overall gtubulin levels, which is nevertheless sufficient to enhance microtubule nucleation and induce mitotic defect.

    Specific points:

    • As pointed out by the authors in the discussion, there needs to be an explanation for how regulating just the mRNA levels of g-tubulin (and not other complex components) can influence the overall protein levels of g-TuRC in order to achieve a functional output for the regulation. The authors provide a nice explanation for when g-tubulin levels drop - this would potentially expose Ubi sites on GCPs that would lead to their ubiquitination and degradation. First, can the authors show that the GCP protein levels are also decreased, like g-tubulin, when they increase the pool of a/b-tubulin dimers? Second, what do the authors think happens when g-tubulin levels are increased? How does this lead to an increase in GCP levels? I don't think this is essential to answer, and certainly not experimentally, but if there is no simple answer then the authors should at least acknowledge this in the discussion.
    • Statistics - In several cases (eg. Fig 3D), the authors compare multiple conditions to one control and use Mann-Whitney or t-tests. The need to use one-way ANOVA with correction for multiple comparisons. This is also true when they compare different conditions to each other, while also comparing to controls. The statistical analysis should be done in a single ANOVA analysis, not with multiple different individual tests. E.g. Fig 4C.
    • N numbers: In many of the experiments, the authors perform 3-4 biological replicates, plotting the value from each replicate e.g. only 3 or 4 values. In Fig 3D and 4C, however, where they examine g-tubulin levels at centrosomes and microtubule nucleation after cold treatment, they plot the individual centrosome values from each of the replicates. Given that there may be variability between the replicates, and the number of centrosomes are not equal between replicates, it would be better to plot the average value from each replicate, which would better match how data is plotted in other experiments. For example, in Figure 4C, perhaps there is not really a significant difference in microtubule nucleation between TTC5 KO and TTC5 KO + TUBGR3H. This result is a bit odd considering the levels of centrosomal g-tubulin are not different (Fig 3D). The authors try to address this in the discussion (without mentioning the result in the results), but I am not fully convinced by their arguments.
    • In the IP shown in Figure 3B, a negative control, such as a construct expressing the FLAG tag alone, is necessary to confirm the specificity of the interactions.
    • In Figure 4B, the authors should include an image that is representative of their quantification. Based on the current image, we would conclude that nucleation is reduced in the TTC5 + TubGR3H condition compared to TubGR3H alone, which is not consistent with the quantification shown in Figure 4C.
    • In Figure S7A, the authors show that all cell lines show similar timing; however, this is not evident from the examples shown in Fig 5B. This is likely because there is a lot of variation in cell division timing and the authors chose to show example images from cells that happened to have different timings. However, this may appear confusing to readers. I t may be better to include the timing graph in the main figure, along with single images to highlight phenotypes (rather than time series for each condition). The time series images could be moved to supplementary.
    • The authors conclude that « partial depletion of g-tubulin restored mitotic fidelity to levels comparable to those of control cells » however in fig 5E, and 5F they only statistically compare the siRNAi control to the siTUBG1 for each genotype (which shows a reduction). But to say that they go back down to control levels, they should also statistically test the difference with the parental siRNAi control.

    Minor comments:

    • Figure 1C : the line above "SCAPER KO" + should be only on the last 2 columns
    • Figure 1E : It would improve clarity if the authors indicated in the figure that the immunoprecipitation was performed using TTC5
    • Figure 5E-F : It is not clear from the graph whether the parental cells were also transfected with the siRNA, although this appears to be the case based on the figure legend.
    • In the legend of Figure S4E, the cell lines appear to be TTC5 KO and TUBGR3H, rather than mutant TTC5 as indicated in the legend

    Significance

    The manuscript is well written, the data is well presented and overall the data supports the conclusions being drawn. The results and conclusions are significant and will be of interest to a broad readership.

  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 #1

    Evidence, reproducibility and clarity

    Microtubule nucleation and dynamics are essential for proper microtubule organization and for diverse cellular functions, including cell division. Nucleation is templated by a ring of y-tubulins within the γ-tubulin ring complex (yTuRC) and therefore depends on the cellular availability of γ-tubulin. Nucleation rates are regulated not only by yTuRC activators but also by the pool of soluble αβ-tubulin available for microtubule polymerization. Cellular αβ-tubulin abundance is controlled by a previously identified autoregulatory pathway that fine-tunes αβ-tubulin mRNA stability in response to changes in soluble tubulin levels. Unexpectedly, y-tubulin transcripts were found to be downregulated in parallel with αβ-tubulin under conditions of elevated soluble tubulin, prompting Assaf et al. to investigate whether y-tubulin abundance is regulated by the canonical tubulin autoregulatory mechanism. By combining transcriptomic reanalysis, targeted genetic perturbation, biochemical interaction assays, and functional cell biological approaches, the authors show that γ-tubulin expression is regulated similarly to αβ-tubulin through a post-transcriptional mechanism in response to soluble tubulin levels, identifying the TTC5-SCAPER-CCR4-NOT axis as essential for the decay of γ-tubulin mRNA, as previously shown for αβ-tubulins. The authors also claim tha loss of γ-tubulin mRNA regulation leads to increased γ-tubulin protein levels and enhanced microtubule nucleation, which ultimately affects mitotic fidelity. Interestingly, they show that just subtle changes in γ-tubulin levels are sufficient to compromise mitotic fidelity, suggesting that γ-tubulin-mediated nucleation is a particularly sensitive control point for mitosis.

    Major comments:

    • Interpretation of TUBG2 transcript data (Fig. 1A, Fig. S1): the authors state that "a similar trend was observed for TUBG2, although the changes in transcript levels were more variable across cell lines." Given that TUBG2 mRNA is expressed at very low levels in non-neuronal cells, conclusions drawn from these datasets are inherently less reliable. This is also reflected by the fact that the authors do not pursue TUBG2 regulation further, in contrast to the detailed analysis of TUBG1. In addition, the trends observed for TUBG2 do not appear as consistent as for TUBG1. I therefore suggest toning down claims regarding TUBG2, or clearly stating that these observations are preliminary. If the authors wish to strengthen this point, experiments in neuronal cells - using microtubule-stabilizing or -destabilizing drugs, or cold-induced changes in soluble αβ-tubulin - would be more appropriate to assess TUBG2 regulation.
    • To improve clarity and focus, I suggest reorganizing Figure 1 and Supplementary Figure 1- move TUBG1 transcriptomic data from Fig. S1A to Fig. 1A and TUBG2 data from Fig. 1A to the Supplementary Figures. This would emphasize the main γ-tubulin analyzed throughout the manuscript under conditions of altered soluble αβ-tubulin, while relegating the less robust data to supplementary material. In this context, Fig. 1E could be moved to Fig. S1, while Fig. S1B could be promoted to Fig. 1, as the fold change observed for TUBG1 in rat heart myocardium is relatively large and appears biologically meaningful.
    • The final summary of results section 1concludes that "TUBG mRNA is regulated post-transcriptionally in response to changes in soluble αβ-tubulin levels." While the effects of the drugs used are well characterized, microtubule-targeting agents can differ in magnitude and kinetics across cell types. To fully support this statement, it would be important to directly show that soluble αβ-tubulin levels change under the conditions used, for example by biochemical fractionation (polymerized vs soluble tubulin). If this is not feasible, the conclusion should be softened to state that TUBG mRNA responds to microtubule-stabilizing and -destabilizing treatments, rather than inferred changes in soluble αβ-tubulin.
    • In Figure 2D, the fold change observed for CNOT11 is less pronounced than for other components of the autoregulation pathway, which is somewhat unexpected given its proposed role as the downstream effector. It would be helpful to clarify whether this regulation is specific to CNOT11, or whether other CCR4-NOT subunits might compensate or contribute. Relatedly, in Fig. S2C, the increase in TUBG1 pre-mRNA levels in CNOT11-KO cells - although not statistically significant - suggests that baseline transcription or mRNA stability may already be altred. This raises some uncertainty regarding the specificity of CNOT11 in regulating γ-tubulin mRNA decay. Additional discussion or clarification would strengthen the interpretation.
    • Quantification of centrosomal fluorescence in fgs. 3D, 4C, S5B-C: Centrosomal γ-tubulin and α-tubulin levels are quantified using integrated density measured within a fixed-size circular ROI. Because ROI area is constant, this approach effectively reflects mean fluorescence intensity, rather than total centrosomal content, unless the ROI fully encompasses all centrosomal signal. Given that centrosome size and γ-tubulin spatial distribution may vary between conditions, a brief justification of how the ROI size was chosen, or a control analysis demonstrating robustness to ROI size (beyond fig S6B), would strengthen the conclusions.
    • Fig. 3E: the change in γ-tubulin levels is modest; complementary measurements of global γ-tubulin levels would strengthen this conclusion.
    • In Fig. 4B, the representative TUBG-R3H cell appears not only to nucleate more microtubules, but also to display faster microtubule polymerization. This is unexpected, as this mutation is proposed to specifically affect γ-tubulin regulation rather than αβ-tubulin availability. In contrast, TTC5-KO cells, where both γ-tubulin and αβ-tubulin regulation may be affected, would more intuitively show such a phenotype. This discrepancy makes the quantification in Fig. 4C difficult to reconcile with the representative images, where microtubule regrowth (based on α-tubulin signal) appears highest in TUBG-R3H cells. Additional clarification or discussion would be helpful.

    Minor comments

    1. Introduction: while the authors thoroughly describe how cells respond to excess soluble αβ-tubulin through autoregulation, the manuscript does not address how cells initially sense changes in soluble tubulin levels. Even if this mechanism remains unresolved, briefly acknowledging this conceptual gap or discussing current hypotheses in the field would strengthen the Introduction and better frame the study.
    2. Figures and legends
      • Fig. 1 and Fig. S1 legends: reference 34 should be reference 36; ***p < 0.001 is mentioned but not shown; Fig. S1B should include a reference along with the GEO accession number.
      • The sentence "In line with previous results for TUBB transcripts (Fig. S1D)" is ambiguous. Please clarify whether this refers to previously published data only, newly generated data, or a combination. Similar clarification may be needed for Fig. S1C and S1E.
      • Fig. 1D: please briefly comment on why microtubule stabilization with PTX leads to a slight but significant decrease in pre-mRNA levels.
      • Replace "decay in tubulin autoregulation" with degradation of TUBA and TUBB mRNA.
      • Text where the call for Fig. 2E appears should read: "TUBG1 and TUBB mRNAs following CA4 treatment in TTC5-KO cells..."
      • Fig. 2F-H: use γ-tubulin and β-tubulin instead of TUBB abd TUBG; the positioning of the γ-tubulin nascent chain within the TTC5 pocket is not clearly illustrated based on author's claim: "the γ-tubulin nascent chain appears to be positioned deeper into the TTC5 pocket (Fig. 2F-G)", and an electrostatic interaction between γ-tubulin R3 and TCC5 D225 should be represented in 2H as in 2G to support the claim that "...the key electrostatic interactions (...) are predicted to be maintained...". Overall, these figures may need adjustement or clarification.
      • Fig. S3B: tubulin should be replaced by α-tubulin.
      • Fig. S3B-C: immunoblots should be accompanied by quantification of the five biological replicates.
      • Fig. S4F-G: these results are compelling and could be moved to the main figure.
      • The metaphase plates shown in Fig. 3C appear relatively homogeneous across conditions, which contrasts with the mitotic defects quantified later (e.g. Fig. 5C). Including representative examples with clearer chromosome alignment or segregation errors, particularly for TTC5-KO cells, would better illustrate the reported phenotypes.
      • In Figs. 3D, 4C, 5B-C, 6C, please specify whether each dot represents a single centrosome or the mean of both centrosomes per cell. Ideally, each dot should correspond to the mean value per cell.
      • Fig. 3E: the change in γ-tubulin levels is modest; complementary measurements of global γ-tubulin levels would strengthen this conclusion.
      • Fig. 4B y-axis should specify α-tubulin fluorescence intensity.
    3. Discussion
      • The Discussion would benefit from explicitly acknowledging limitations, such as the relatively - The statement "among γ-TuRC components, only γ-tubulin mRNA is subject to autoregulation" should be rephrased, or additional γ-TuRC subunits should be tested to support this claim.
    4. Methods
      • Consider separating imaging procedures from analysis into distinct sections (e.g., Immunofluorescence and Microscopy data analysis).
      • Correct 2.4 µm (not µM).
      • Clarify ROI selection and background subtraction strategy, as discussed above.
    5. Optional / stylistic
      • Use consistent placement of "n.s." and asterisks in bar plots.
      • Fig. 2A: if CA4 is included in the schematic, it should be mentioned in the legend. Alternatively, the model could depict a generic increase in soluble αβ-tubulin dimers.
      • Fig. S3A may be unnecessary; instead, consider summarizing homology percentages between αβ-tubulin and γ-tubulin, highlighting higher N-terminal and lower C-terminal conservation to further motivate shared autoregulation.

    Significance

    This study extends the concept of tubulin autoregulation beyond αβ-tubulin by identifying γ-tubulin as an additional target of the same post-transcriptional regulatory pathway. By doing so, it highlights a coordinated mechanism that links control of microtubule building blocks with regulation of microtubule nucleation capacity, which is central for maintaining proper microtubule organization and mitotic fidelity. The work therefore contributes to a more integrated view of how cells balance microtubule mass, number, and organization during cell division. The tubulin autoregulation pathway involving TTC5, SCAPER, and the CCR4-NOT complex has been well characterized for αβ-tubulin, and previous studies had already established that γ-tubulin levels are tightly controlled, with both overexpression and depletion leading to mitotic defects. While the extension of this regulatory mechanism to γ-tubulin is important, it builds on existing concepts rather than introducing a fundamentally new regulatory pathway. In this sense, the study refines and extends current knowledge by providing mechanistic insight into how γ-tubulin abundance is regulated. The finding that among γ-TuRC components only γ-tubulin mRNA appears to be subject to autoregulation raises interesting questions regarding the specificity and functional consequences of this selective regulation. Although the work does not introduce a completely novel concept, its detailed analysis of γ-tubulin autoregulation and its functional impact on microtubule nucleation and mitotic fidelity will be of interest to the microtubule and cell division research communities.

    My expertise lies in microtubule nucleation and minus-end regulation, with a focus on γ-TuRC function and the in vitro reconstitution of its regulation and activity.

  5. Version published to 10.64898/2026.01.09.698560 on bioRxiv