Map7D2 and Map7D1 facilitate microtubule stabilization through distinct mechanisms in neuronal cells

  1. Koji Kikuchi
  2. Yasuhisa Sakamoto
  3. Akiyoshi Uezu
  4. Hideyuki Yamamoto
  5. Kei-ichiro Ishiguro
  6. Kenji Shimamura
  7. Taro Saito
  8. Shin-ichi Hisanaga
  9. Hiroyuki Nakanishi

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Microtubule (MT) dynamics are modulated through the coordinated action of various MT-associated proteins (MAPs). However, the regulatory mechanisms underlying MT dynamics remain unclear. We show that the MAP7 family protein Map7D2 stabilizes MTs to control cell motility and neurite outgrowth. Map7D2 directly bound to MTs through its N-terminal half and stabilized MTs in vitro. Map7D2 localized prominently to the centrosome and partially on MTs in mouse N1-E115 neuronal cells, which expresses two of the four MAP7 family members, Map7D2 and Map7D1. Map7D2 loss decreased the resistance to the MT-destabilizing agent nocodazole without affecting acetylated/detyrosinated stable MTs, suggesting that Map7D2 stabilizes MTs via direct binding. In addition, Map7D2 loss increased the rate of random cell migration and neurite outgrowth, presumably by disturbing the balance between MT stabilization and destabilization. Map7D1 exhibited similar subcellular localization and gene knockdown phenotypes to Map7D2. However, in contrast to Map7D2, Map7D1 was required for the maintenance of acetylated stable MTs. Taken together, our data suggest that Map7D2 and Map7D1 facilitate MT stabilization through distinct mechanisms in cell motility and neurite outgrowth.

Article activity feed

  1. Version published to 10.26508/lsa.202201390
  2. Version published to 10.1101/2021.10.27.466197v3 on bioRxiv
  3. 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

    Manuscript number: RC-2021-01129

    Corresponding author(s): Koji Kikuchi

    Reviewer #1

    Evidence, reproducibility and clarity (Required):

    In this manuscript, Kikuchi et al describe the characterization of MAP7D2 and MAP7D1, two MAP7 family members in mouse with specific expression patterns. Focusing mostly on MAP7D2, they assess its expression pattern across the body and find that it is mostly expressed in certain neuronal subsets. They then characterize the MT-related properties of MAP7D2 based on previous knowledge of other MAP7 family members. They show that MAP7D2 binds MTs (via the N-terminus), determine the binding affinity, and show that it can stimulate MT polymerization (or stabilization) both in vitro and in vivo. Using a specific antibody, they localize MAP7D2 to centrosomes, midbody and neurites in N1-E115 cells. Functionally, they show that loss of MAP7D1/2 mildly affects microtubule stability as judged by acetyl-tubulin staining, and properties of these cells that rely on cytoskeletal elements such as cell migration and neurite growth. Interestingly, there might be a feedback loop regulating MAP7D1/2 expression, as knockdown of MAP7D1 upregulates MAP7D2.

    Overall, the experiments and conclusions are very solid and convincing, such that I would not ask for further experiments. This is in part because the experiments are largely based on previous characterizations of other MAP7 family members, which are largely confirmed. The presentation of the data is also very clear.

    Significance (Required):

    I see the value of the study in the fact that it provides solid and specific research tools for MAP7D1/2 which could be __very useful __for the microtubule/neuronal cytoskeleton community.

    Response: We thank the reviewer very much for appreciating the content of our manuscript.

    **Referees cross-commenting**

    Reviewers 2 and 3 criticize that the evidence for an effect of MAP7D1/2 on MT dynamics is weak. I would agree in that ac-tub stainings and in vitro experiments are rather indirect. The experiments suggested by reviewer 2 should clarify this (esp. nocodazole should be easy). I also agree that an experiment addressing the potential involvement of kinesin-1 would help, the involvement of which seems to have been omitted by the authors. A kinesin-binding deficient mutant would add another MAP7D1/2 tool and increase the value for the community.

    __Response: __As for the reviewer’s suggestions listed above, please refer to our responses to the comments of Reviewer #2.

    __Reviewer #2 __

    Evidence, reproducibility and clarity (Required):

    In this study, the authors investigate 2 members from the MAP7 family Map7D2 and Map7D1. They first address the tissue distribution of Map7D2, by northern blotting using a variety of rat tissues. To complement their analysis, they also raised an antibody to look at the protein distribution. From their studies, they concluded that Map7D2 is abundantly expressed in the brain and testis. The authors went on to perform a series of functional assays. First, they biochemically demonstrated that rat Map7D2 directly binds to MTs by MT co-sedimentation assay. The MT binding domain was mapped to the N-terminal half. They performed MT turbidity assay to demonstrate enhanced MT polymerisation in the presence of Map7D2, suggesting that this Map stabilises MTs. The authors went on to characterise in detail the subcellular localisation of Map7D2 which was predominantly present in the centrosome and partially localised to MTs including within neurites from N1-E115 cells. Kikuchi et al. further revealed the overlap in expression between Map7D2 and another family member, Map7D1. The authors continued these studies by a series of functional studies in N1-E115 cells where they performed single or combined knock-downs of Map7D2 and Map7D1 and studied the levels of acetylated and detyrosinated tubulins and the effect of the knock-downs on migration and neurite extension. The main conclusion from this work was that Map7D2 and Map7D1 facilitate MT stabilization through distinct mechanisms which are important in controlling cell motility and neurite outgrowth. Map7D2 is proposed to stabilise MTs by direct binding whereas Map7D1 does it indirectly by affecting acetylation.

    Major comments:

    The main conclusion from this work that Map7D2 and Map7D1 facilitate MT stabilization and that this is necessary for correct migration and neurite extension has not been convincingly demonstrated. In my opinion, a more detailed study of MT properties to demonstrate a role in MT stabilisation would greatly benefit the work, eg. experiments using MT destabilising agents such as nocodazole. In addition, a series of experiments aiming to study MT dynamics would help to understand the function of these MT regulators. The authors proposed an elevation in microtubule dynamics to explain the increase in migration and neurite extension but no experimental proof was provided.

    Response: According to the reviewer’s suggestion, we plan to assess the role of MT stabilization in greater detail by analyzing the sensitivity to the MT-destabilizing agent, nocodazole.

    To study MT dynamics, methods such as analyzing the velocity and direction of an EB1-GFP comet are commonly used. We have previously analyzed the roles of Map7 and Map7D1 in MT dynamics using HeLa cells stably expressing EB1-GFP (Kikuchi et al., EMBO Rep., 2018). However, no such tools have been developed for analyzing MT dynamics in N1-E115 cells, which were used in this study. In addition, it is difficult to analyze MT dynamics by transient expression of EB1-GFP because of the low plasmid transfection efficiency. Therefore, we instead plan to assess the effect on MT dynamics by measuring the EB1 comet length by immunofluorescence, referring to Fig. 7D in EMBO J. 32:1293–1306, 2013.

    Moreover, considering the possibility that the Map7D2 dynamics are altered when MT stability is changed, e.g., before and after differentiation induction, we analyzed the Map7D2 dynamics at the centrosome by fluorescence recovery after photobleaching (FRAP) using N1-E115 cells stably expressing EGFP-rMap7D2. We found that the dynamics were altered between the proliferative and differentiated states (see the figure below). Compared to the proliferative state, the recovery rate of EGFP-Map7D2 was reduced (lower left panel), and the immobile fraction of Map7D2 was increased in the differentiated state (lower right panel). As these data suggest that the increase in immobile Map7D2 may enhance MT stabilization, we will present them in a new figure in our manuscript along with the results of the above two experiments.

    It has been previously demonstrated that loss of MAP7D2 leads to a decrease in axonal cargo entry to axons resulting in defects in axon development and neuronal migration. The C-terminus is necessary for this function as it mediates interaction with Kinesin-1 (Pan et al., 2019). Such mechanisms could also explain the defects in migration and neurite growth that the authors observed. This possibility has not been considered but instead, the subtle changes in total α-tubulin led to suggest MT stabilisation as a key function without proof of causation. Could the authors provide some further experimental evidence to demonstrate that stability is the main contributor to the phenotypes observed? Eg. by rescuing migration and neurite phenotypes with a variant of MAP7D2 which cannot bind kinesin1.

    Response: The reviewer states “Such mechanisms could also explain the defects in migration and neurite growth that the authors observed;” however, our results showed that loss of Map7D2 elevated the rates of both cell motility and neurite outgrowth (original Fig. 5). In contrast, it has been reported in several papers that when Kinesin-1 function is impaired, both cell motility and neurite outgrowth are reduced (Curr. Biol., 23: 1018–1023, 2013; Mol. Cell. Biol., 39: e00109–19, 2019; etc.). Therefore, it is likely that the phenotypes we observed are independent of the functions associated with Kinesin-1 in N1-E115 cells. It is indeed possible that the experiment suggested by the reviewer may reveal relationships between Map7D2 and kinesin-1 in terms of cell motility and neurite outgrowth, however, it is difficult to conduct such an experiment because transient expression of Map7D2 induces MT bundling, as shown in original Fig. 2F. Based on the above, we plan to add a discussion of the relationship between Map7D2 and Kinesin-1.

    A key conclusion proposed by the authors is that Map7D2 and Map7D1 facilitate MT stabilization through distinct mechanisms. Such different roles in MT stabilisation are important in controlling cell motility and neurite outgrowth. In my opinion, their data does not fully support this statement and the findings using MT readouts do not match the defects in migration and neurite growth. Loss of Map7D2 leads to a very subtle phenotype on α-tubulin, while Map7D1 decreases both α-tubulin and acetylated tubulin, but Map7D1 seems to have a milder or similar effect on migration and neurite growth than Map7D2. Furthermore, it would be expected that the combined loss of function would lead to a stronger phenotype in cell migration when compared to the single loss of functions due to their distinct roles on MT stability, however, this seems not to be the case.

    __Response: __The fact that no stronger phenotype was observed may be because, besides Map7D2 and Map7D1, other molecules are involved in MT stabilization. Another possible explanation is that the increases in both cell motility and neurite outgrowth caused by decreased MT stabilization are offset by Kinesin-1 dysfunction. We plan to add a discussion of the above two possibilities.

    Minor comments:

    1) In the first result section, the author refers to Fig. S3 to suggest the expression of MAP7D2 in the cerebral cortex, however, there are no transcripts in the cerebral cortex according to the figure. Similarly, the immunofluorescence analysis done by the authors shows marginal expression of MAP7D2 in the cerebral cortex.

    __Response: __According to the reviewer’s comment, we have changed the order of the data shown in Fig. 1C, top panels. The data from the olfactory bulb, cerebellum, and hippocampus, in which Map7D2 expression was detected in the database, were arranged in the top three rows, and the data from the cerebral cortex, in which Map7D2 expression was not detected in the database, were moved to the bottom row as a negative control. In addition, we have revised the relevant part of the Results section as follows: “Based on RNA-seq CAGE, RNA-Seq, and SILAC database analysis (Expression Atlas, https://www.ebi.ac.uk/gxa/home/), Map7D2 expression was detected in the cerebellum, hippocampus, and olfactory bulb, and not in the cerebral cortex (Fig. S3). We further confirmed Map7D2 expression in the above four brain tissue regions of postnatal day 0 mice by immunofluorescence. Among these regions, Map7D2 was the most highly expressed in the Map2-negative area of the olfactory bulb, i.e., the glomerular layer (Fig. 1C). Weak signals were detected in the cerebellum, and marginal signals were observed in the hippocampus and cerebral cortex (Fig. 1C).” (page 5, lines 4–11)

    2) The authors use γ-Tubulin as a housekeeping gene in Fig. 3D, since Map7D2 is enriched in centrosomes this may not be the most appropriate choice.

    __Response: __γ-Tubulin is abundant in both the cytosol and the nuclear compartments of cells (Sig. Transduct. Target Ther. 3: 24, 2018). As it has been used for similar purposes in several other studies (Cancer Res., 61: 7713–7718, 2001; J. Biol. Chem., 291: 23112–23125, 2016; etc.), we considered it acceptable for use as a loading control for immunoblotting.

    3) According to the authors, knockdown of Map7D2 leads to a decrease in the intensity of α-tubulin and Map7D1 (Fig. 4C and D). This data doesn't agree with the previous statement made by the authors where they show that Map7D2 knockdown or knockout did not affect Map7D1 expression by Western Blot Analysis (Fig. S2C and S5B)

    Response: The immunoblotting results indicate that the total amount of Map7D1 in the cells is not affected by loss of Map7D2. In contrast, the immunofluorescence results indicate that the amount (distribution) of Map7D1 localized around the centrosome is decreased by loss of Map7D2, presumably due to a reduction in the number of MT structures that can serve as scaffolds for Map7D1. We plan to add this interpretation in the Results section.

    4) Line 6 page 7 "Endogenous Map7D2 expression is suppressed in N1-E115 cells stably expressing EGFP-rMap7D2 and was restored by specific knock-down of EGFP-rMap7D2 using gfp siRNA (Fig. 3D)". No quantifications and stats are shown. Also, endogenous Map7D2 after knock-down of EGFP-rMap7D2 is not comparable to the control.

    __Response: __According to the reviewer’s suggestion, we have quantified the amount of endogenous Map7D2 or EGFP-rMap7D2, normalized it to the amount of γ-tubulin, and calculated relative values to endogenous Map7D2 in the parental control. The amount of endogenous Map7D2 was decreased to 53% in N1-E115 cells stably expressing EGFP-rMap7D2, suggesting that EGFP-rMap7D2 expression suppressed endogenous Map7D2 expression. In this cell line, the total amount of Map7D2 (EGFP-rMap7D2 + endogenous Map7D2) was increased, however, when EGFP-rMap7D2 was depleted using sigfp in this cell line, endogenous Map7D2 was expressed to the same level as EGFP-rMap7D2 before knock-down. Together with the finding that Map7d1 knock-down increased the amount of Map7D2, these findings indicate that the amount of Map7D2 in the cells is regulated in response to the amount of Map7D1 and exogenous Map7D2. We have added this interpretation in the Results section. (page 7, lines 8–15)

    In addition, we have changed the legend of the original Fig. 3D to clarify the quantification method, as follows: “(D) Generation of N1-E115 cells stably expressing EGFP-rMap7D2. To check the expression level of EGFP-rMap7D2, lysates derived from the indicated cells were probed with anti-GFP (top panel) and anti-Map7D2 (middle panel) antibodies. The blot was reprobed for γ-tubulin as a loading control (bottom panel). The amount of endogenous Map7D2 or EGFP-rMap7D2 was normalized to the amount of γ-tubulin, and the value relative to endogenous Map7D2 in the parental control was calculated.” (page 22, lines 18–20)

    5) Line 8 page 7 "These results suggest that the expression of Map7D2 was influenced by changes in that of Map7D1" This statement seems in the wrong place, after the Map7D2 and EGFP-rMap7D2 experiment. Instead for clarity, it would be better placed after line 5 where the authors explain the effect of Map7D1 knock-down on the levels of Map7D2.

    Response: According to the reviewer’s suggestion, we have rephrased the relevant sentence as “Interestingly, Map7d1 knock-down upregulated Map7D2 expression, as confirmed with three different siRNAs (Fig. S2C), suggesting that Map7D2 expression is affected by changes in Map7D1 expression, not by off-target effects of a particular siRNA.” (page 7, lines 7, 8)

    6) Line 8 page 8 "Although the physiological role of the C-terminal region of Map7D2 is currently unknown..." This statement seems not adequate as there are several studies reporting the role of the C-terminal region of Map7D2 in Kinesin1- mediated transport. The authors mention such studies in the discussion.

    __Response: __According to the reviewer’s suggestion, we plan to add a discussion of the relationship between Map7D2 and kinesin-1.

    7) Line 6 page 9 " Further, the knock-down of either resulted in a comparable reduction of MT intensity (Fig. 4C and D) ..." This is not visible and/or justified by the images provided and would benefit from some sort of quantification at other regions such as neurites.

    Response: Considering the cell motility, quantification of α-tubulin/Ace-tubulin/Map7D1/Map7D2 intensities in neurites is not appropriate. Instead, we have added arrowheads indicating α-tubulin/Ace-tubulin/Map7D1/Map7D2 in Fig. 4C, for better understanding.

    8) In Fig. 2B, a band corresponding to his6-rMAP7D2 of molecular weight >97 kDa co-sedimented with the microtubules. However, the cloned rMAP7D2 had a molecular weight of 84.82 kDa and the addition of 6XHis-Tag would add another 2-3 kDa, therefore, the final protein band observed should be less than 90 kDa. It would be beneficial if the authors could specify the molecular weight of the purified protein after the addition of the V5-his tag and/or if there was addition of amino acids due to cloning strategy.

    __Response: __In Fig. 2B, we used full-length GST-tagged rMap7D2, like in Fig. 2E and D; therefore, we have corrected His6-rMap7D2 as GST-rMap7D2. We apologize for the mistake.

    9) In Fig. 2C, there is misalignment of the western blot with the panel or text underneath.

    Response: We thank the reviewer for pointing this out; we have corrected the misalignment of the CBB staining in Fig. 2C.

    10) In Fig. 3C the inset from the first panel seems to correspond to a different focal plane than the main image.

    __Response: __We have revised the relevant part of the figure legend as follows: “In C, images of differentiated cells were captured by z-sectioning, because the focal planes of the centrosome and neurites are different. Each inset shows an enlarged image of the region indicated with a white box at each focal plane. Arrowheads indicate the centrosomal localization of Map7D2.”

    11) In Fig. 4A, the cell type is not specified and is referred as "indicated cells", also the material and methods section seems to omit the specific cells used.

    __Response: __We have added “in N1-E115 cells treated with each siRNA” in the legend of Fig. 4A.

    12) Fig. S6 is not mentioned in the results.

    __Response: __We apologize for having referred to Fig. S6 only in the Discussion section in the original manuscript. We plan to describe the findings shown in the original Fig. S6 to the Results section and renumber the figures accordingly.

    Significance (Required):

    MTs play essential roles in practically every cellular process. Their precise regulation is therefore crucial for cellular function and viability. MAPs are specialised proteins that interact with MTs and regulate their behaviour in different manners. Understanding their precise function in different cellular contexts is of utmost importance for many biological and biomedical fields.

    MAPs are well known for their ability to promote MT polymerization, bundling and stabilisation in vitro (Bodakuntla et al., 2019). Several members of the Map7 family have been shown to regulate microtubule stability. For instance, MAP7 can prevent nocodazole-induced MT depolymerization and maintain stable microtubules at branch points in DRG neurons (Tymanskyj & Ma, 2019). Ensconsin, the Drosophila Map, is required for MT growth in mitotic neuroblasts by regulating the mean rate of MT polymerization (Gallaud et al., 2014). However, this family of Maps seems to have diverse functions encompassing a variety of mechanisms, as exemplified by a series of studies demonstrating the involvement of MAP7 family proteins in the recruitment and activation of kinesin1 (Hooikaas et al., 2019; Pan et al., 2019) and in microtubule remodelling and Wnt5a signalling (Kikuchi et al., 2018). Further understanding of this family of Maps and how its members differ in their function is important and will help to advance the field.

    __Response: __We appreciate the reviewer’s comments. We believe that our revision plan will greatly improve the quality of our manuscript.

    __Reviewer #3 __

    Evidence, reproducibility and clarity (Required):

    Summary:

    Microtubule Associated Proteins (MAPs) are important regulators of microtubule dynamics, microtubule organization and vesicular transport by modulating motor protein recruitment and processivity. In the current manuscript the authors have characterized 2 members of the MAP7 protein family, MAP7D1 and MAP7D2. The authors characterized MAP7D2 expression pattern in the brain and its microtubule binding properties in vitro and in cells. In cells both proteins localize to the centrosome and to microtubules and upon depletion centrosome localized microtubules seem reduced, and cell migration and neurite outgrowth are increased. Surprisingly, they find that microtube acetylation (a common marker for stable microtubules) is reduced upon MAP7D1 depletion but not MAP7D2 depletion. Based on this finding the authors conclude that these proteins have a distinct mechanism in stabilizing MTs to affect cell migration and neurite outgrowth; MAP7D2 stabilizes by binding to MTs, whereas MAP7D1 stabilizes MTs by acetylation.

    Main comments:

    - Both MAP7 proteins show strong localization to the centrosome and to a lesser degree to MTs. Knockdown of either protein leads to reduced MTs around the centrosome, which lead the authors to conclude the MAP7s are stabilizing the MTs. However, the effect could just as well be an indirect effect due to a function of these MAPs at the centrosome. To address this authors could e.g. quantify microtubule properties in postmitotic cells. In addition, antibody specificity should be tested using knockdown of knockout cells, as this centrosome localization was not observed in Hela cells (Hooikaas, 2019; Kikuchi, 2018). Maybe this localization is specific to rat MAP7s or to the cell line used.

    Response: We think that this comment partly overlaps with the comments raised by Reviewer #2. We plan to assess the role of MT stabilization in greater detail by analyzing the sensitivity to the MT-destabilizing agent, nocodazole, and the effect on MT dynamics by measuring the EB1 comet length by immunofluorescence.

    Regarding the reviewer’s concern about antibody specificity, we had carefully confirmed the antibody specificity, as shown in Fig. S2 of the original manuscript. Subsequently, Map7D2 localization was confirmed in N1-E115 cells stably expressing EGFP-rMap7D2, as shown in Fig. 3D, E of the original manuscript. In addition, we are currently conducting analyses using Map7d1-egfp knock-in mice, which confirmed that Map7D1 localizes around the centrosome in cortical neurons, as shown below (we would like to disclose these unpublished data to the reviewers only). Therefore, it is thought that the localization pattern of Map7D2 and Map7D1 differs depending on the cell type and cell line. We plan to add this interpretation to the Results section.

    - Centrosome nucleated microtubules are typically highly dynamic and little modified. Therefore is the Ac-tub staining at the centrosome really MTs? I cannot identify MTs in the fluorescent images in 4C. Maybe authors could consider ac-tub/alpha-tub ratio in non centrosomal region (e.g. neurites). Moreover, as both Acetylation and detyrosination are associated with long-lived/stable MTs, it is surprising that only acetylated tubulin goes down on WB. Does this suggest that long-lived MTs are still present to normal level? If so, can one still argue that the loss of acetylation is the cause of the lower MT levels? This should at least be discussed.

    Response: As for the reviewer’s statement “Centrosome nucleated microtubules are typically highly dynamic and little modified. Therefore is the Ac-tub staining at the centrosome really MTs?”, it has been previously reported that tubulin acetylation is observed around the centrosome in some cell lines (J. Neurosci., 30: 7215–7226, 2010; PLoS One, 13: e0190717, 2018; etc.). N1-E115 is one of the cell lines in which tubulin acetylation is observed around the centrosome.

    It is not surprising that “only acetylated tubulin goes down on WB,” as it has been previously reported that acetylated and detyrosinated tubulins are sometimes not synchronous (J. Neurosci., 23: 10662–10671, 2003; J. Neurosci., 30: 7215–7226, 2010; J. Cell Sci., 132: jcs225805, 2019., etc.). For instance, Montagnac et al. (Nature, 502: 567–570, 2013) showed that defects in the α-tubulin acetyltransferase αTAT1-clathrin-dependent endocytosis axis reduce only tubulin acetylation, resulting in a shift from directional to random cell migration. Although the details of the molecular function of Map7D1 are beyond the main purpose of this study, we plan to add a discussion of the reduced tubulin acetylation by Map7d1 knock-down based on the above.

    - MAP7D1 and MAP7D2 depletion leads to subtle defect in cell migration and neurite outgrowth, which the author suggest is caused by reduced MT stability. However, MAP7 proteins have well characterized functions in kinesin-1 transport, and thus the phenotypes may well be caused by defects in kinesin-1 transport. Ideally the authors would do rescue experiments with FL or just the MT binding N-termini to separate these functions. Moreover this is needed to substantiate the claim of the authors that MAP7D1 effect on MT stability is not mediated by direct binding.

    __Response: __As this comment largely overlaps with the comments raised by Reviewer #2, please refer to our responses to the comments of Reviewer #2.

    - The authors do not refer well to published work. Several papers have published very similar work (especially to Fig1+2) and it would help the reader much if this would be discussed/compared along the results section and not briefly mention these in the results section. In addition, authors overstate the novelty of their results e.g. page 3: these proteins are not "functionally uncharacterized" nor are their expression patter and biochemical properties analyzed for the first time in this manuscript; page 8 "Although the physiological role of the C-terminal region of Map7D2 is currently unknow, ..." There is a clear function for the C-terminus for the recruitment/activation of kinesin-1.

    Response: According to the reviewer’s suggestion, we plan to add a comparison with data on the Map7 family members presented in previous papers in the Results section and rephrase the relevant part regarding the physiological role of the C-terminal region of Map7D2.

    Minor comments

    - P6 Map7D3 also binds with its N-terminus to MTs, like other MAP7s (Yadav et al)

    __Response: __According to the reviewer’s comment, we have revised this as “Map7D3 binds through a conserved region on not only the N-terminal side, but also the C-terminal side (Sun, 2011; Yadav et al., 2014).” (page 6, lines 4, 5)

    - P7 "As Map7D2 has the potential to functionally compensate for Map7D1 loss" where is this based on?

    Response: For clarity, we have rephrased this as “As Ma7D2 expression was upregulated upon suppression of Map7D1 expression, Map7D2 has the potential to functionally compensate for Map7D1 loss.” (page 7, line 17, 18)

    - Fig2F quality of black-white images is low potentially due to conversion issues

    __Response: __We thank the reviewer for pointing out these conversion issues, and we have made the necessary corrections.

    Significance (Required):

    At this stage the conceptual advance is limited. Part of the findings are not novel. The finding that MAP7s depletion have a different effect on MTs acetylation may be interesting to cytoskeleton researchers, although the potential mechanism has not been addressed experimentally or textually.

    However, their conclusion that this leads to reduced MTs and then to cellar migration and neurite formation defects is not sufficiently supported by experimental evidence.

    __Response: __We appreciate the reviewer’s comments. We believe that our revision plan will greatly improve the quality of our manuscript.

    **Referees cross-commenting**

    I completely agree with reviewer #2: At this stage the paper's conclusions are not sufficiently supported by the data. Important will be to further characterize the effect om the MTs (do they really have a different effect) and to look at the possible involvement of the motor recruitment. Maybe that a 3 to 6 months revision time would have been more accurate.

    __Response: __Please refer to our responses to the comments of Reviewer #2.

  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 #3

    Evidence, reproducibility and clarity

    Summary:

    Microtubule Associated Proteins (MAPs) are important regulators of microtubule dynamics, microtubule organization and vesicular transport by modulating motor protein recruitment and processivity. In the current manuscript the authors have characterized 2 members of the MAP7 protein family, MAP7D1 and MAP7D2. The authors characterized MAP7D2 expression pattern in the brain and its microtubule binding properties in vitro and in cells. In cells both proteins localize to the centrosome and to microtubules and upon depletion centrosome localized microtubules seem reduced, and cell migration and neurite outgrowth are increased. Surprisingly, they find that microtube acetylation (a common marker for stable microtubules) is reduced upon MAP7D1 depletion but not MAP7D2 depletion. Based on this finding the authors conclude that these proteins have a distinct mechanism in stabilizing MTs to affect cell migration and neurite outgrowth; MAP7D2 stabilizes by binding to MTs, whereas MAP7D1 stabilizes MTs by acetylation.

    Main comments:

    • Both MAP7 proteins show strong localization to the centrosome and to a lesser degree to MTs. Knockdown of either protein leads to reduced MTs around the centrosome, which lead the authors to conclude the MAP7s are stabilizing the MTs. However, the effect could just as well be an indirect effect due to a function of these MAPs at the centrosome. To address this authors could e.g. quantify microtubule properties in postmitotic cells. In addition, antibody specificity should be tested using knockdown of knockout cells, as this centrosome localization was not observed in Hela cells (Hooikaas, 2019; Kikuchi, 2018). Maybe this localization is specific to rat MAP7s or to the cell line used.
    • Centrosome nucleated microtubules are typically highly dynamic and little modified. Therefore is the Ac-tub staining at the centrosome really MTs? I cannot identify MTs in the fluorescent images in 4C. Maybe authors could consider ac-tub/alpha-tub ratio in non centrosomal region (e.g. neurites). Moreover, as both Acetylation and detyrosination are associated with long-lived/stable MTs, it is surprising that only acetylated tubulin goes down on WB. Does this suggest that long-lived MTs are still present to normal level? If so, can one still argue that the loss of acetylation is the cause of the lower MT levels? This should at least be discussed.
    • MAP7D1 and MAP7D2 depletion leads to subtle defect in cell migration and neurite outgrowth, which the author suggest is caused by reduced MT stability. However, MAP7 proteins have well characterized functions in kinesin-1 transport, and thus the phenotypes may well be caused by defects in kinesin-1 transport. Ideally the authors would do rescue experiments with FL or just the MT binding N-termini to separate these functions. Moreover this is needed to substantiate the claim of the authors that MAP7D1 effect on MT stability is not mediated by direct binding.
    • The authors do not refer well to published work. Several papers have published very similar work (especially to Fig1+2) and it would help the reader much if this would be discussed/compared along the results section and not briefly mention these in the results section. In addition, authors overstate the novelty of their results e.g. page 3: these proteins are not "functionally uncharacterized" nor are their expression patter and biochemical properties analyzed for the first time in this manuscript; page 8 "Although the physiological role of the C-terminal region of Map7D2 is currently unknow, ..." There is a clear function for the C-terminus for the recruitment/activation of kinesin-1.

    Minor comments:

    • P6 Map7D3 also binds with its N-terminus to MTs, like other MAP7s (Yadav et al)
    • P7 "As Map7D2 has the potential to functionally compensate for Map7D1 loss" where is this based on?
    • Fig2F quality of black-white images is low potentially due to conversion issues

    Significance

    At this stage the conceptual advance is limited. Part of the findings are not novel. The finding that MAP7s depletion have a different effect on MTs acetylation may be interesting to cytoskeleton researchers, although the potential mechanism has not been addressed experimentally or textually.

    However, their conclusion that this leads to reduced MTs and then to cellar migration and neurite formation defects is not sufficiently supported by experimental evidence.

    Referees cross-commenting

    I completely agree with reviewer #2: At this stage the paper's conclusions are not sufficiently supported by the data. Important will be to further characterize the effect om the MTs (do they really have a different effect) and to look at the possible involvement of the motor recruitment. Maybe that a 3 to 6 months revision time would have been more accurate.

  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 #2

    Evidence, reproducibility and clarity

    In this study, the authors investigate 2 members from the MAP7 family Map7D2 and Map7D1. They first address the tissue distribution of Map7D2, by northern blotting using a variety of rat tissues. To complement their analysis, they also raised an antibody to look at the protein distribution. From their studies, they concluded that Map7D2 is abundantly expressed in the brain and testis. The authors went on to perform a series of functional assays . First, they biochemically demonstrated that rat Map7D2 directly binds to MTs by MT co-sedimentation assay. The MT binding domain was mapped to the N-terminal half. They performed MT turbidity assay to demonstrate enhanced MT polymerisation in the presence of Map7D2, suggesting that this Map stabilises MTs. The authors went on to characterise in detail the subcellular localisation of Map7D2 which was predominantly present in the centrosome and partially localised to MTs including within neurites from N1-E115 cells. Kikuchi et al. further revealed the overlap in expression between Map7D2 and another family member, Map7D1. The authors continued these studies by a series of functional studies in N1-E115 cells where they performed single or combined knock-downs of Map7D2 and Map7D1 and studied the levels of acetylated and detyrosinated tubulins and the effect of the knock-downs on migration and neurite extension. The main conclusion from this work was that Map7D2 and Map7D1 facilitate MT stabilization through distinct mechanisms which are important in controlling cell motility and neurite outgrowth. Map7D2 is proposed to stabilise MTs by direct binding whereas Map7D1 does it indirectly by affecting acetylation.

    Major comments:

    The main conclusion from this work that Map7D2 and Map7D1 facilitate MT stabilization and that this is necessary for correct migration and neurite extension has not been convincingly demonstrated. In my opinion, a more detailed study of MT properties to demonstrate a role in MT stabilisation would greatly benefit the work, eg. experiments using MT destabilising agents such as nocodazole. In addition, a series of experiments aiming to study MT dynamics would help to understand the function of these MT regulators. The authors proposed an elevation in microtubule dynamics to explain the increase in migration and neurite extension but no experimental proof was provided.

    It has been previously demonstrated that loss of MAP7D2 leads to a decrease in axonal cargo entry to axons resulting in defects in axon development and neuronal migration. The C-terminus is necessary for this function as it mediates interaction with Kinesin-1 (Pan et al., 2019). Such mechanisms could also explain the defects in migration and neurite growth that the authors observed. This possibility has not been considered but instead, the subtle changes in total -tubulin led to suggest MT stabilisation as a key function without proof of causation. Could the authors provide some further experimental evidence to demonstrate that stability is the main contributor to the phenotypes observed? Eg. by rescuing migration and neurite phenotypes with a variant of MAP7D2 which cannot bind kinesin1.

    A key conclusion proposed by the authors is that Map7D2 and Map7D1 facilitate MT stabilization through distinct mechanisms. Such different roles in MT stabilisation are important in controlling cell motility and neurite outgrowth. In my opinion, their data does not fully support this statement and the findings using MT readouts do not match the defects in migration and neurite growth. Loss of Map7D2 leads to a very subtle phenotype on -tubulin, while Map7D1 decreases both -tubulin and acetylated tubulin, but Map7D1 seems to have a milder or similar effect on migration and neurite growth than Map7D2. Furthermore, it would be expected that the combined loss of function would lead to a stronger phenotype in cell migration when compared to the single loss of functions due to their distinct roles on MT stability, however, this seems not to be the case.

    Minor comments:

    1. In the first result section, the author refers to Fig. S3 to suggest the expression of MAP7D2 in the cerebral cortex, however, there are no transcripts in the cerebral cortex according to the figure. Similarly, the immunofluorescence analysis done by the authors shows marginal expression of MAP7D2 in the cerebral cortex.
    2. The authors use -Tubulin as a housekeeping gene in Fig. 3D, since Map7D2 is enriched in centrosomes this may not be the most appropriate choice.
    3. According to the authors, knockdown of Map7D2 leads to a decrease in the intensity of -tubulin and Map7D1 (Fig. 4C and D). This data doesn't agree with the previous statement made by the authors where they show that Map7D2 knockdown or knockout did not affect Map7D1 expression by Western Blot Analysis (Fig. S2C and S5B)
    4. Line 6 page 7 "Endogenous Map7D2 expression is suppressed in N1-E115 cells stably expressing EGFP-rMap7D2 and was restored by specific knock-down of EGFP-rMap7D2 using gfp siRNA (Fig. 3D)". No quantifications and stats are shown. Also, endogenous Map7D2 after knock-down of EGFP-rMap7D2 is not comparable to the control.
    5. Line 8 page 7 "These results suggest that the expression of Map7D2 was influenced by changes in that of Map7D1" This statement seems in the wrong place, after the Map7D2 and EGFP-rMap7D2 experiment. Instead for clarity, it would be better placed after line 5 where the authors explain the effect of Map7D1 knock-down on the levels of Map7D2.
    6. Line 8 page 8 "Although the physiological role of the C-terminal region of Map7D2 is currently unknown..." This statement seems not adequate as there are several studies reporting the role of the C-terminal region of Map7D2 in Kinesin1- mediated transport. The authors mention such studies in the discussion.
    7. Line 6 page 9 " Further, the knock-down of either resulted in a comparable reduction of MT intensity (Fig. 4C and D) ..." This is not visible and/or justified by the images provided and would benefit from some sort of quantification at other regions such as neurites.
    8. In Fig. 2B, a band corresponding to his6-rMAP7D2 of molecular weight >97 kDa co-sedimented with the microtubules. However, the cloned rMAP7D2 had a molecular weight of 84.82 kDa and the addition of 6XHis-Tag would add another 2-3 kDa, therefore, the final protein band observed should be less than 90 kDa. It would be beneficial if the authors could specify the molecular weight of the purified protein after the addition of the V5-his tag and/or if there was addition of amino acids due to cloning strategy.
    9. In Fig. 2C, there is misalignment of the western blot with the panel or text underneath.
    10. In Fig. 3C the inset from the first panel seems to correspond to a different focal plane than the main image.
    11. In Fig. 4A, the cell type is not specified and is referred as "indicated cells", also the material and methods section seems to omit the specific cells used.
    12. Fig. S6 is not mentioned in the results.

    Significance

    MTs play essential roles in practically every cellular process. Their precise regulation is therefore crucial for cellular function and viability. MAPs are specialised proteins that interact with MTs and regulate their behaviour in different manners. Understanding their precise function in different cellular contexts is of utmost importance for many biological and biomedical fields.

    MAPs are well known for their ability to promote MT polymerization, bundling and stabilisation in vitro (Bodakuntla et al., 2019). Several members of the Map7 family have been shown to regulate microtubule stability. For instance, MAP7 can prevent nocodazole-induced MT depolymerization and maintain stable microtubules at branch points in DRG neurons (Tymanskyj & Ma, 2019). Ensconsin, the Drosophila Map, is required for MT growth in mitotic neuroblasts by regulating the mean rate of MT polymerization (Gallaud et al., 2014). However, this family of Maps seems to have diverse functions encompassing a variety of mechanisms, as exemplified by a series of studies demonstrating the involvement of MAP7 family proteins in the recruitment and activation of kinesin1 (Hooikaas et al., 2019; Pan et al., 2019) and in microtubule remodelling and Wnt5a signalling (Kikuchi et al., 2018). Further understanding of this family of Maps and how its members differ in their function is important and will help to advance the field.

  6. 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

    In this manuscript, Kikuchi et al describe the characterization of MAP7D2 and MAP7D1, two MAP7 family members in mouse with specific expression patterns. Focusing mostly on MAP7D2, they assess its expression pattern across the body and find that it is mostly expressd in certain neuronal subsets. They then characterize the MT-related properties of MAP7D2 based on previous knowledge of other MAP7 family members. They show that MAP7D2 binds MTs (via the N-terminus), determine the binding affinity, and show that it can stimulate MT polymerization (or stabilization) both in vitro and in vivo. Using a specific antibody, they localize MAP7D2 to centrosomes, midbody and neurites in N1-E115 cells. Functionally, they show that loss of MAP7D1/2 mildly affects microtubule stability as judged by acetyl-tubulin staining, and properties of these cells that rely on cytoskeletal elements such as cell migration and neurite growth. Interestingly, there might be a feedback loop regulating MAP7D1/2 expression , as knockdown of MAP7D1 upregulates MAP7D2.

    Overall, the experiments and conclusions are very solid and convincing, such that I would not ask for further experiments. This is in part because the experiments are largely based on previous characterizations of other MAP7 family members, which are largely confirmed. The presentation of the data is also very clear.

    Significance

    I see the value of the study in the fact that,it provides solid and specific research tools for MAP7D1/2 which could be very useful for the microtubule/neuronal cytoskeleton community.

    Referees cross-commenting

    Reviewers 2 and 3 criticize that the evidence for an effect of MAP7D1/2 on MT dynamics is weak. I would agree in that ac-tub stainings and in vitro experiments are rather indirect. The experiments suggested by reviewer 2 should clarify this (esp. nocodazole should be easy). I also agree that an experiment addressing the potential involvement of kinesin-1 would help, the involvement of which seems to have been omitted by the authors. A kinesin-binding deficient mutant would add another MAP7D1/2 tool and increase the value for the community.

  7. Version published to 10.1101/2021.10.27.466197v1 on bioRxiv
  8. Version published to 10.1101/2021.10.27.466197v2 on bioRxiv