Spatial activation of Kinesin-1 by Ensconsin shapes microtubule networks via ncMTOCs recruitment

Article activity feed

1. 

   Review Commons

   Nov 25, 2025

   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

   Reviewer #1

   Evidence, reproducibility and clarity

   This paper addresses a very interesting problem of non-centrosomal microtubule organization in developing Drosophila oocytes. Using genetics and imaging experiments, the authors reveal an interplay between the activity of kinesin-1, together with its essential cofactor Ensconsin, and microtubule organization at the cell cortex by the spectraplakin Shot, minus-end binding protein Patronin and Ninein, a protein implicated in microtubule minus end anchoring. The authors demonstrate that the loss of Ensconsin affects the cortical accumulation non-centrosomal microtubule organizing center (ncMTOC) …

   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

   Reviewer #1

   Evidence, reproducibility and clarity

   This paper addresses a very interesting problem of non-centrosomal microtubule organization in developing Drosophila oocytes. Using genetics and imaging experiments, the authors reveal an interplay between the activity of kinesin-1, together with its essential cofactor Ensconsin, and microtubule organization at the cell cortex by the spectraplakin Shot, minus-end binding protein Patronin and Ninein, a protein implicated in microtubule minus end anchoring. The authors demonstrate that the loss of Ensconsin affects the cortical accumulation non-centrosomal microtubule organizing center (ncMTOC) proteins, microtubule length and vesicle motility in the oocyte, and show that this phenotype can be rescued by constitutively active kinesin-1 mutant, but not by Ensconsin mutants deficient in microtubule or kinesin binding. The functional connection between Ensconsin, kinesin-1 and ncMTOCs is further supported by a rescue experiment with Shot overexpression. Genetics and imaging experiments further implicate Ninein in the same pathway. These data are a clear strength of the paper; they represent a very interesting and useful addition to the field.

   The weaknesses of the study are two-fold. First, the paper seems to lack a clear molecular model, uniting the observed phenomenology with the molecular functions of the studied proteins. Most importantly, it is not clear how kinesin-based plus-end directed transport contributes to cortical localization of ncMTOCs and regulation of microtubule length.

   Second, not all conclusions and interpretations in the paper are supported by the presented data.

   We thank the reviewer for recognizing the impact of this work. In response to the insightful suggestions, we performed extensive new experiments that establish a well-supported cellular and molecular model (Figure 7). The discussion has been restructured to directly link each conclusion to its corresponding experimental evidence, significantly strengthening the manuscript.

   Below is a list of specific comments, outlining the concerns, in the order of appearance in the paper/figures.

   Figure 1. The statement: "Ens loading on MTs in NCs and their subsequent transport by Dynein toward ring canals promotes the spatial enrichment of the Khc activator Ens in the oocyte" is not supported by data. The authors do not demonstrate that Ens is actually transported from the nurse cells to the oocyte while being attached to microtubules. They do show that the intensity of Ensconsin correlates with the intensity of microtubules, that the distribution of Ensconsin depends on its affinity to microtubules and that an Ensconsin pool locally photoactivated in a nurse cell can redistribute to the oocyte (and throughout the nurse cell) by what seems to be diffusion. The provided images suggest that Ensconsin passively diffuses into the oocyte and accumulates there because of higher microtubule density, which depends on dynein. To prove that Ensconsin is indeed transported by dynein in the microtubule-bound form, one would need to measure the residence time of Ensconsin on microtubules and demonstrate that it is longer than the time needed to transport microtubules by dynein into the oocyte; ideally, one would like to see movement of individual microtubules labelled with photoconverted Ensconsin from a nurse cell into the oocyte. Since microtubules are not enriched in the oocyte of the dynein mutant, analysis of Ensconsin intensity in this mutant is not informative and does not reveal the mechanism of Ensconsin accumulation.

   As noted by Reviewer 3, the directional movement of microtubules traveling at ~140 nm/s from nurse cells toward the oocyte through Ring Canals was previously reported using a tagged Ens-MT binding domain reporter line by Lu et al. (2022). We have therefore added the citation of this crucial work in the novel version of the manuscript (lane 155-157) and removed the photo-conversion panel.

   Critically, however, our study provides mechanistic insight that was missing from this earlier work: this mechanism is also crucial to enrich MAPs in the oocyte. The fact that Dynein mutants fail to enrich Ensconsin is a crucial piece of evidence: it supports a model of Ensconsin-loaded MT transport (Figure 1D-1F).

   Figure 2. According to the abstract, this figure shows that Ensconsin is "maintained at the oocyte cortex by Ninein". However, the figure doesn't seem to prove it - it shows that oocyte enrichment of Ensonsin is partially dependent on Ninein, but this applies to the whole cell and not just to the cell cortex. Furthermore, it is not clear whether Ninein mutation affects microtubule density, which in turn would affect Ensconsin enrichment, and therefore, it is not clear whether the effect of Ninein loss on Ensconsin distribution is direct or indirect.

   Ninein plays a critical role in Ensconsin enrichment and microtubule organization in the oocyte (new Figure 2, Figure 3, Figure S3). Quantification of total Tubulin signal shows no difference between control and Nin mutant oocytes (new Figure S3 panels A, B). We found decreased Ens enrichment in the oocyte, and Ens localization on MTs and to the cell cortex (Figure 2E, 2F, and Figure S3C and S3D).

   Novel quantitative analyses of microtubule orientation at the anterior cortex, where MTs are normally preferentially oriented toward the posterior pole (Parton et al. 2011), demonstrate that Nin mutants exhibit randomized MT orientation compared to wild-type oocytes (new Figure 3C-3E).These findings establish that Ninein (although not essential) favors Ensconsin localization on MTs, Ens enrichment in the oocyte, ncMTOC cortical localization, and more robust MT orientation toward the posterior cortex. It also suggests that Ens levels in the oocyte acts as a rheostat to control Khc activation.

   The observation that the aggregates formed by overexpressed Ninein accumulate other proteins, including Ensconsin, supports, though does not prove their interactions. Furthermore, there is absolutely no proof that Ninein aggregates are "ncMTOCs". Unless the authors demonstrate that these aggregates nucleate or anchor microtubules (for example, by detailed imaging of microtubules and EB1 comets), the text and labels in the figure would need to be altered.

   We have modified the manuscript, we now refer to an accumulation of these components in large puncta, rather than aggregates, consistent with previous observations (Rosen et al., 2000). We acknowledge in the revised version that these puncta recruit Shot, Patronin and Ens without mentioning direct interaction (lane 218).

   Importantly, we conducted a more detailed characterization of these Ninein/Shot/Patronin/Ens-containing puncta in a novel Figure S4. To rigorously assess their nucleation capacity, we analyzed Eb1-GFP-labeled MT comets, a robust readout of MT nucleation (Parton et al., 2011, Nashchekin et al., 2016). While few Eb1-positive comets occasionally emanate from these structures, confirming their identity as putative ncMTOCs, these puncta function as surprisingly weak nucleation centers (new Figure S4 E, Video S1) and, their presence does not alter overall MT architecture (new Figure S4 F). Moreover, these puncta disappear over time, are barely visible at stage 10B, they do not impair oocyte development or fertility (Figure S4 G and Table 1).

   Minor comment: Note that a "ratio" (Figure 2C) is just a ratio, and should not be expressed in arbitrary units.

   We have amended this point in all the figures.

   Figure 3B: immunoprecipitation results cannot be interpreted because the immunoprecipitated proteins (GFP, Ens-GFP, Shot-YFP) are not shown. It is also not clear that this biochemical experiment is useful. If the authors would like to suggest that Ensconsin directly binds to Patronin, the interaction would need to be properly mapped at the protein domain level.

   This is a good point: the GFP and Ens-GFP immunoprecipitated proteins are now much clearly identified on the blots and in the figure legend (new Figure 4G). Shot-YFP IP, was used as a positive control but is difficult to be detected by Western blot due to its large size (>106 Da) using conventional acrylamide gels (Nashchekin et al., 2016).

   We now explicitly state that immunoprecipitations were performed at 4°C, where microtubules are fully depolymerized, thereby excluding undirect microtubule-mediated interactions. We agree with this reviewer: we cannot formally rule out interactions through bridging by other protein components. This is stated in the revised manuscript (lane 238-239).

   One of the major phenotypes observed by the authors in Ens mutant is the loss of long microtubules. The authors make strong conclusions about the independence of this phenotype from the parameters of microtubule plus-end growth, but in fact, the quality of their data does not allow to make such a conclusion, because they only measured the number of EB1 comets and their growth rate but not the catastrophe, rescue or pausing frequency."Note that kinesin-1 has been implicated in promoting microtubule damage and rescue (doi: 10.1016/j.devcel.2021).In the absence of such measurements, one cannot conclude whether short microtubules arise through defects in the minus-end, plus-end or microtubule shaft regulation pathways.

   We thank the reviewer for raising this important point. Our data demonstrate that microtubule (MT) nucleation and polymerization rates remain unaffected under Khc RNAi and ens mutant conditions, indicating that MT dynamics alterations must arise through alternative mechanisms.

   As the reviewer suggested, recent studies on Kinesin activity and MT network regulation are indeed highly relevant. Two key studies from the Verhey and Aumeier laboratories examined Kinesin-1 gain-of-function conditions and revealed that constitutively active Kinesin-1 induces MT lattice damage (Budaitis et al., 2022). While damaged MTs can undergo self-repair, Aumeier and colleagues demonstrated that GTP-tubulin incorporation generates "rescue shafts" that promote MT rescue events (Andreu-Carbo et al., 2022). Extrapolating from these findings, loss of Kinesin-1 activity could plausibly reduce rescue shaft formation, thereby decreasing MT rescue frequency and stability. Although this hypothesis is challenging to test directly in our system, it provides a mechanistic framework for the observed reduction in MT number and stability.

   Additionally, the reviewer highlighted the role of Khc in transporting the dynactin complex, an anti-catastrophe factor, to MT plus ends (Nieuwburg et al., 2017), which could further contribute to MT stabilization. This crucial reference is now incorporated into the revised Discussion.

   Importantly, our work also demonstrates the contribution of Ens/Khc to ncMTOC targeting to the cell cortex. Our new quantitative analyses of MT organization (new Figure 5 B) reveal a defective anteroposterior orientation of cortical MTs in mutant conditions, pointing to a critical role for cortical ncMTOCs in organizing the MT network.

   Taken together, we propose that the observed MT reduction and disorganization result from multiple interconnected mechanisms: (1) reduced rescue shaft formation affecting MT stability; (2) impaired transport of anti-catastrophe factors to MT plus ends; and (3) loss of cortical ncMTOCs, which are essential for minus-end MT stabilization and network organization. The Discussion has been revised to reflect this integrated model in a dedicated paragraph (“A possible regulation of MT dynamics in the oocyte at both plus end minus MT ends by Ens and Khc” lane 415-432).

   It is important to note in that a spectraplakin, like Shot, can potentially affect different pathways, particularly when overexpressed.

   We agree that Shot harbors multiple functional domains and acts as a key organizer of both actin and microtubule cytoskeletons. Overexpression of such a cytoskeletal cross-linker could indeed perturb both networks, making interpretation of Ens phenotype rescue challenging due to potential indirect effects.

   To address this concern, we selected an appropriate Shot isoform for our rescue experiments that displayed similar localization to “endogenous” Shot-YFP (a genomic construct harboring shot regulatory sequences) and importantly that was not overexpressed.

   Elevated expression of the Shot.L(A) isoform (see Western Blot Figure S8 A), considered as the wild-type form with two CH1 and CH2 actin-binding motifs (Lee and Kolodziej, 2002), showed abnormal localization such as strong binding to the microtubules in nurse cells and oocyte confirming the risk of gain-of-function artifacts and inappropriate conclusions (Figure S8 B, arrows).

   By contrast, our rescue experiments using the Shot.L(C) isoform (that only harbors the CH2 motif) provide strong evidence against such artifacts for three reasons. First, Shot-L(C) is expressed at slightly lower levels than a Shot-YFP genomic construct (not overexpressed), and at much lower levels than Shot-L(A), despite using the same driver (Figure S8 A). Second, Shot-L(C) localization in the oocyte is similar to that of endogenous Shot-YFP, concentrating at the cell cortex (Figure S8 B, compare lower and top panels). Taken together, these controls rather suggest our rescue with the Shot-L(C) is specific.

   Note that this Shot-L(C) isoform is sufficient to complement the absence of the shot gene in other cell contexts (Lee and Kolodziej, 2002).

   Unjustified conclusions should be removed: the authors do not provide sufficient data to conclude that "ens and Khc oocytes MT organizational defects are caused by decreased ncMTOC cortical anchoring", because the actual cortical microtubule anchoring was not measured.

   This is a valid point. We acknowledge that we did not directly measure microtubule anchoring in this study. In response, we have revised the discussion to more accurately reflect our observations. Throughout the manuscript, we now refer to "cortical microtubule organization" rather than "cortical microtubule anchoring," which better aligns with the data presented.

   Minor comment: Microtubule growth velocity must be expressed in units of length per time, to enable evaluating the quality of the data, and not as a normalized value.

   This is now amended in the revised version (modified Figure S7).

   A significant part of the Discussion is dedicated to the potential role of Ensconsin in cortical microtubule anchoring and potential transport of ncMTOCs by kinesin. It is obviously fine that the authors discuss different theories, but it would be very helpful if the authors would first state what has been directly measured and established by their data, and what are the putative, currently speculative explanations of these data.

   We have carefully considered the reviewer's constructive comments and are confident that this revised version fully addresses their concerns.

   First, we have substantially strengthened the connection between the Results and Discussion sections, ensuring that our interpretations are more directly anchored in the experimental data. This restructuring significantly improves the overall clarity and logical flow of the manuscript.

   Second, we have added a new comprehensive figure presenting a molecular-scale model of Kinesin-1 activation upon release of autoinhibition by Ensconsin (new Figure 7D). Critically, this figure also illustrates our proposed positive feedback loop mechanism: Khc-dependent cytoplasmic advection promotes cortical recruitment of additional ncMTOCs, which generates new cortical microtubules and further accelerates cytoplasmic transport (Figure 7 A-C). This self-amplifying cycle provides a mechanistic framework consistent with emerging evidence that cytoplasmic flows are essential for efficient intracellular transport in both insect and mammalian oocytes.

   Minor comment: The writing and particularly the grammar need to be significantly improved throughout, which should be very easy with current language tools. Examples: "ncMTOCs recruitment" should be "ncMTOC recruitment"; "Vesicles speed" should be "Vesicle speed", "Nin oocytes harbored a WT growth,"- unclear what this means, etc. Many paragraphs are very long and difficult to read. Making shorter paragraphs would make the authors' line of thought more accessible to the reader.

   We have amended and shortened the manuscript according to this reviewer feed-back. We have specifically built more focused paragraphs to facilitates the reading.

   Significance

   This paper represents significant advance in understanding non-centrosomal microtubule organization in general and in developing Drosophila oocytes in particular by connecting the microtubule minus-end regulation pathway to the Kinesin-1 and Ensconsin/MAP7-dependent transport. The genetics and imaging data are of good quality, are appropriately presented and quantified. These are clear strengths of the study which will make it interesting to researchers studying the cytoskeleton, microtubule-associated proteins and motors, and fly development.

   The weaknesses of this study are due to the lack of clarity of the overall molecular model, which would limit the impact of the study on the field. Some interpretations are not sufficiently supported by data, but this can be solved by more precise and careful writing, without extensive additional experimentation.

   We thank the reviewer for raising these important concerns regarding clarity and data interpretation. We have thoroughly revised the manuscript to address these issues on multiple fronts. First, we have substantially rewritten key sections to ensure that our conclusions are clearly articulated and directly supported by the data. Second, we have performed several new experiments that now allow us to propose a robust mechanistic model, presented in new figures. These additions significantly strengthen the manuscript and directly address the reviewer's concerns.

   My expertise is cell biology and biochemistry of the microtubule cytoskeleton, including both microtubule-associated proteins and microtubule motors.

   Reviewer #2

   Evidence, reproducibility and clarity

   In this manuscript, Berisha et al. investigate how microtubule (MT) organization is spatially regulated during Drosophila oogenesis. The authors identify a mechanism in which the Kinesin-1 activator Ensconsin/MAP7 is transported by dynein and anchored at the oocyte cortex via Ninein, enabling localized activation of Kinesin-1. Disruption of this pathway impairs ncMTOC recruitment and MT anchoring at the cortex. The authors combine genetic manipulation with high-resolution microscopy and use three key readouts to assess MT organization during mid-to-late oogenesis: cortical MT formation, localization of posterior determinants, and ooplasmic streaming. Notably, Kinesin-1, in concert with its activator Ens/MAP7, contributes to organizing the microtubule network it travels along. Overall, the study presents interesting findings, though we have several concerns we would like the authors to address. Ensconsin enrichment in the oocyte

   1. Enrichment in the oocyte • Ensconsin is a MAP that binds MTs. Given that microtubule density in the oocyte significantly exceeds that in the nurse cells, its enrichment may passively reflect this difference. To assess whether the enrichment is specific, could the authors express a non-Drosophila MAP (e.g., mammalian MAP1B) to determine whether it also preferentially localizes to the oocyte?

   To address this point, we performed a new series of experiments analyzing the enrichment of other Drosophila and non-Drosophila MAPs, including Jupiter-GFP, Eb1-GFP, and bovine Tau-GFP, all widely used markers of the microtubule cytoskeleton in flies (see new Figure S2). Our results reveal that Jupiter-GFP, Eb1-GFP, and bovine Tau-GFP all exhibit significantly weaker enrichment in the oocyte compared to Ens-GFP. Khc-GFP also shows lower enrichment. These findings indicate that MAP enrichment in the oocyte is MAP-dependent, rather than solely reflecting microtubule density or organization. Of note, we cannot exclude that microtubule post-translational modifications contribute to differential MAP binding between nurse cells and the oocyte, but this remains a question for future investigation.

   The ability of ens-wt and ens-LowMT to induce tubulin polymerization according to the light scattering data (Fig. S1J) is minimal and does not reflect dramatic differences in localization. The authors should verify that, in all cases, the polymerization product in their in vitro assays is microtubules rather than other light-scattering aggregates. What is the control in these experiments? If it is just purified tubulin, it should not form polymers at physiological concentrations.

   The critical concentration Cr for microtubule self-assembly in classical BRB80 buffer found by us and others is around 20 µM (see Fig. 2c in Weiss et al., 2010). Here, microtubules were assembled at 40 µM tubulin concentration, i.e., largely above the Cr. As stated in the materials and methods section, we systematically induced cooling at 4°C after assembly to assess the presence of aggregates, since those do not fall apart upon cooling. The decrease in optical density upon cooling is a direct control that the initial increase in DO is due to the formation of microtubules. Finally, aggregation and polymerization curves are widely different, the former displaying an exponential shape and the latter a sigmoid assembly phase (see Fig. 3A and 3B in Weiss et al., 2010).

   Photoconversion caveatsMAPs are known to dynamically associate and dissociate from microtubules. Therefore, interpretation of the Ens photoconversion data should be made with caution. The expanding red signal from the nurse cells to the oocyte may reflect a any combination of dynein-mediated MT transport and passive diffusion of unbound Ensconsin. Notably, photoconversion of a soluble protein in the nurse cells would also result in a gradual increase in red signal in the oocyte, independent of active transport. We encourage the authors to more thoroughly discuss these caveats. It may also help to present the green and red channels side by side rather than as merged images, to allow readers to assess signal movement and spatial patterns better.

   This is a valid point that mirrors the comment of Reviewers 1 and 3. The directional movement of microtubules traveling at ~140 nm/s from nurse cells toward the oocyte via the ring canals was previously reported by Lu et al. (2022) with excellent spatial resolution. Notably, this MT transport was measured using a fusion protein containing the Ens MT-binding domain. We now cite this relevant study in our revised manuscript and have removed this redundant panel in Figure 1.

   Reduction of Shot at the anterior cortex• Shot is known to bind strongly to F-actin, and in the Drosophila ovary, its localization typically correlates more closely with F-actin structures than with microtubules, despite being an MT-actin crosslinker. Therefore, the observed reduction of cortical Shot in ens, nin mutants, and Khc-RNAi oocytes is unexpected. It would be important to determine whether cortical F-actin is also disrupted in these conditions, which should be straightforward to assess via phalloidin staining.

   As requested by the reviewer, we performed actin staining experiments, which are now presented in a new Figure S5. These data demonstrate that the cortical actin network remains intact in all mutant backgrounds analyzed, ruling out any indirect effect of actin cytoskeleton disruption on the observed phenotypes.

   MTs are barely visible in Fig. 3A, which is meant to demonstrate Ens-GFP colocalization with tubulin. Higher-quality images are needed.

   The revised version now provides significantly improved images to show the different components examined. Our data show that Ens and Ninein localize at the cell cortex where they co-localize with Shot and Patronin (Figure 2 A-C). In addition, novel images show that Ens extends along microtubules (new Figure 4 A).

   MT gradient in stage 9 oocytesIn ens-/-, nin-/-, and Khc-RNAi oocytes, is there any global defect in the stage 9 microtubule gradient? This information would help clarify the extent to which cortical localization defects reflect broader disruptions in microtubule polarity.

   We now provide quantitative analysis of microtubule (MT) array organization in novel figures (Figure 3D and Figure 5B). Our data reveal that both Khc RNAi and *ens *mutant oocytes exhibit severe disruption of MT orientation toward the posterior (new Figure 5B). Importantly, this defect is significantly less pronounced in Nin-/- oocytes, which retain residual ncMTOCs at the cortex (new Figure 3D). This differential phenotype supports our model that cortical ncMTOCs are critical for maintaining proper MT orientation toward the posterior side of the oocyte.

   Role of Ninein in cortical anchoringThe requirement for Ninein in cortical anchorage is the least convincing aspect of the manuscript and somewhat disrupts the narrative flow. First, it is unclear whether Ninein exhibits the same oocyte-enriched localization pattern as Ensconsin. Is Ninein detectable in nurse cells? Second, the Ninein antibody signal appears concentrated in a small area of the anterior-lateral oocyte cortex (Fig. 2A), yet Ninein loss leads to reduced Shot signal along a much larger portion of the anterior cortex (Fig. 2F)-a spatial mismatch that weakens the proposed functional relationship. Third, Ninein overexpression results in cortical aggregates that co-localize with Shot, Patronin, and Ensconsin. Are these aggregates functional ncMTOCs? Do microtubules emanate from these foci?

   We now provide a more comprehensive analysis of Ninein localization. Similar to Ensconsin (Ens), endogenous Ninein is enriched in the oocyte during the early stages of oocyte development but is also detected in NCs (see modified Figure 2 A and Lasko et al., 2016). Improved imaging of Ninein further shows that the protein partially co-localizes with Ens, and ncMTOCs at the anterior cortex and with Ens-bound MTs (Figure 2B, 2C).

   Importantly, loss of Ninein (Nin) only partially reduces the enrichment of Ens in the oocyte (Figure 2E). Both Ens and Kinesin heavy chain (Khc) remain partially functional and continue to target non-centrosomal microtubule-organizing centers (ncMTOCs) to the cortex (Figure 3A). In Nin-/- mutants, a subset of long cortical microtubules (MTs) is present, thereby generating cytoplasmic streaming, although less efficiently than under wild-type (WT) conditions (Figure 3F and 3G). As a non-essential gene, we envisage Ninein as a facilitator of MT organization during oocyte development.

   Finally, our new analyses demonstrate that large puncta containing Ninein, Shot, Patronin, and despite their size, appear to be relatively weak nucleation centers (revised Figure S4 E and Video 1). In addition, their presence does not bias overall MT architecture (Figure S4 F) nor impair oocyte development and fertility (Figure S4 G and Table 1).

   Inconsistency of Khc^MutEns rescueThe Khc^MutEns variant partially rescues cortical MT formation and restores a slow but measurable cytoplasmic flow yet it fails to rescue Staufen localization (Fig. 5). This raises questions about the consistency and completeness of the rescue. Could the authors clarify this discrepancy or propose a mechanistic rationale?

   This is a good point. The cytoplasmic flows (the consequence of cargo transport by Khc on MTs) generated by a constitutively active KhcMutEns in an ens mutant condition, are less efficient than those driven by Khc activated by Ens in a control condition (Figure 6C). The rescued flow is probably not efficient enough to completely rescue the Staufen localization at stage 10.

   Additionally, this KhcMutEns variant rescues the viability of embryos from Khc27 mutant germline clones oocytes but not from ens mutants (Table1). One hypothesis is that Ens harbors additional functions beyond Khc activation.

   This incomplete rescue of Ens by an active Khc variant could also be the consequence of the “paradox of co-dependence”: Kinesin-1 also transport the antagonizing motor Dynein that promotes cargo transport in opposite directions (Hancock et al., 2016). The phenotype of a gain of function variant is therefore complex to interpret. Consistent with this, both KhcMutEns-GFP and KhcDhinge2 two active Khc only rescues partially centrosome transport in ens mutant Neural Stem Cells (Figure S10).

   Minor points:

   1. The pUbi-attB-Khc-GFP vector was used to generate the Khc^MutEns transgenic line, presumably under control of the ubiquitous ubi promoter. Could the authors specify which attP landing site was used? Additionally, are the transgenic flies viable and fertile, given that Kinesin-1 is hyperactive in this construct?

   All transgenic constructs were integrated at defined genomic landing sites to ensure controlled expression levels. Specifically, both GFP-tagged KhcWT and KhcMutEns were inserted at the VK05 (attP9A) site using PhiC31-mediated integration. Full details of the landing sites are provided in the Materials and Methods section. Both transgenic flies are homozygous lethal and the transgenes are maintained over TM6B balancers.

   On page 11 (Discussion, section titled "A dual Ensconsin oocyte enrichment mechanism achieves spatial relief of Khc inhibition"), the statement "many mutations in Kif5A are causal of human diseases" would benefit from a brief clarification. Since not all readers may be familiar with kinesin gene nomenclature, please indicate that KIF5A is one of the three human homologs of Kinesin heavy chain.

   We clarified this point in the revised version (lane 465-466).

   On page 16 (Materials and Methods, "Immunofluorescence in fly ovaries"), the sentence "Ovaries were mounted on a slide with ProlonGold medium with DAPI (Invitrogen)" should be corrected to "ProLong Gold."

   This is corrected.

   Significance

   This study shows that enrichment of MAP7/ensconsin in the oocyte is the mechanism of kinesin-1 activation there and is important for cytoplasmic streaming and localization non-centrosomal microtubule-organizing centers to the oocyte cortex

   We thank the reviewers for the accurate review of our manuscript and their positive feed-back.

   Reviewer #3

   Evidence, reproducibility and clarity

   The manuscript of Berisha et al., investigates the role of Ensconsin (Ens), Kinesin-1 and Ninein in organisation of microtubules (MT) in Drosophila oocyte. At stage 9 oocytes Kinesin-1 transports oskar mRNA, a posterior determinant, along MT that are organised by ncMTOCs. At stage 10b, Kinesin-1 induces cytoplasmic advection to mix the contents of the oocyte. Ensconsin/Map7 is a MT associated protein (MAP) that uses its MT-binding domain (MBD) and kinesin binding domain (KBD) to recruit Kinesin-1 to the microtubules and to stimulate the motility of MT-bound Kinesin-1. Using various new Ens transgenes, the authors demonstrate the requirement of Ens MBD and Ninein in Ens localisation to the oocyte where Ens activates Kinesin-1 using its KBD. The authors also claim that Ens, Kinesin-1 and Ninein are required for the accumulation of ncMTOCs at the oocyte cortex and argue that the detachment of the ncMTOCs from the cortex accounts for the reduced localisation of oskar mRNA at stage 9 and the lack of cytoplasmic streaming at stage 10b. Although the manuscript contains several interesting observations, the authors' conclusions are not sufficiently supported by their data. The structure function analysis of Ensconsin (Ens) is potentially publishable, but the conclusions on ncMTOC anchoring and cytoplasmic streaming not convincing.

   We are grateful that the regulation of Khc activity by MAP7 was well received by all reviewers. While our study focuses on Drosophila oogenesis, we believe this mechanism may have broader implications for understanding kinesin regulation across biological systems.

   For the novel function of the MAP7/Khc complex in organizing its own microtubule networks through ncMTOC recruitment, we have carefully considered the reviewers' constructive recommendations. We now provide additional experimental evidence supporting a model of flux self-amplification in which ncMTOC recruitment plays a key role. It is well established that cytoplasmic flows are essential for posterior localization of cell fate determinants at stage 10B. Slow flows have also been described at earlier oogenesis stages by the groups of Saxton and St Johnston. Building on these early publications and our new experiments, we propose that these flows are essential to promote a positive feedback loop that reinforces ncMTOC recruitment and MT organization (Figure 7).

   1. The main conclusion of the manuscript is that "MT advection failure in Khc and ens in late oogenesis stems from defective cortical ncMTOCs recruitment". This completely overlooks the abundant evidence that Kinesin-1 directly drives cytoplasmic streaming by transporting vesicles and microtubules along microtubules, which then move the cytoplasm by advection (Palacios et al., 2002; Serbus et al, 2005; Lu et al, 2016). Since Kinesin-1 generates the flows, one cannot conclude that the effect of khc and ens mutants on cortical ncMTOC positioning has any direct effect on these flows, which do not occur in these mutants.

   We regret the lack of clarity of the first version of the manuscript and some missing references. We propose a model in which the Kinesin-1- dependent slow flows (described by Serbus/Saxton and Palacios/StJohnston) play a central role in amplifying ncMTOC anchoring and cortical MT network formation (see model in the new Figure 7).

   1. The authors claim that streaming phenotypes of ens and khs mutants are due to a decrease in microtubule length caused by the defective localisation of ncMTOCs. In addition to the problem raised above, However, I am not convinced that they can make accurate measurements of microtubule length from confocal images like those shown in Figure 4. Firstly, they are measuring the length of bundles of microtubules and cannot resolve individual microtubules. This problem is compounded by the fact that the microtubules do not align into parallel bundles in the mutants. This will make the "microtubules" appear shorter in the mutants. In addition, the alignment of the microtubules in wild-type allows one to choose images in which the microtubule lie in the imaging plane, whereas the more disorganized arrangement of the microtubules in the mutants means that most microtubules will cross the imaging plane, which precludes accurate measurements of their length.

   As mentioned by Reviewer 4, we have been transparent with the methodology, and the limitations that were fully described in the material and methods section.

   Cortical microtubules in oocytes are highly dynamic and move rapidly, making it technically impossible to capture their entire length using standard Z-stack acquisitions. We therefore adopted a compromise approach: measuring microtubules within a single focal plane positioned just below the oocyte cortex. This strategy is consistent with established methods in the field, such as those used by Parton et al. (2011) to track microtubule plus-end directionality. To avoid overinterpretation, we explicitly refer to these measurements as "minimum detectable MT length," acknowledging that microtubules may extend beyond the focal plane, particularly at stage 10, where long, tortuous bundles frequently exit the plane of focus. These methodological considerations and potential biases are clearly described in the Materials and Methods section and the text now mentions the possible disorganization of the MT network in the mutant conditions (lane 272-273).

   In this revised version, we now provide complementary analyses of MT network organization.Beyond length measurements (and the mentioned limitations), we also quantified microtubule network orientation at stage 9, assessing whether cortical microtubules are preferentially oriented toward the posterior axis as observed in controls (revised Figure 3D and Figure 5B). While this analysis is also subject to the same technical limitations, it reveals a clear biological difference: microtubules exhibit posterior-biased orientation in control oocytes similar to a previous study (Parton et al., 2011) but adopt a randomized orientation in Nin-/-, ens, and Khc RNAi-depleted oocytes (revised Figure 3D and Figure 5B).

   Taken together, these complementary approaches, despite their technical constraints, provide convergent evidence for the role of the Khc/Ens complex in organizing cortical microtubule networks during oogenesis.

   1. "To investigate whether the presence of these short microtubules in ens and Khc RNAi oocytes is due to defects in microtubule anchoring or is also associated with a decrease in microtubule polymerization at their plus ends, we quantified the velocity and number of EB1comets, which label growing microtubule plus ends (Figure S3)." I do not understand how the anchoring or not of microtubule minus ends to the cortex determines how far their plus ends grow, and these measurements fall short of showing that plus end growth is unaffected. It has already been shown that the Kinesin-1-dependent transport of Dynactin to growing microtubule plus ends increases the length of microtubules in the oocyte because Dynactin acts as an anti-catastrophe factor at the plus ends. Thus, khc mutants should have shorter microtubules independently of any effects on ncMTOC anchoring. The measurements of EB1 comet speed and frequency in FigS2 will not detect this change and are not relevant for their claims about microtubule length. Furthermore, the authors measured EB1 comets at stage 9 (where they did not observe short MT) rather than at stage 10b. The authors' argument would be better supported if they performed the measurements at stage 10b.

   We thank the reviewer for raising this important point. The short microtubule (MT) length observed at stage 10B could indeed result from limited plus-end growth. Unfortunately, we were unable to test this hypothesis directly: strong endogenous yolk autofluorescence at this stage prevented reliable detection of Eb1-GFP comets, precluding velocity measurements.

   At least during stage 9, our data demonstrate that MT nucleation and polymerization rates are not reduced in both KhcRNAi and ens mutant conditions, indicating that the observed MT alterations must arise through alternative mechanisms.

   In the discussion, we propose the following interconnected explanations, supported by recent literature and the reviewers’ suggestions:

   1- Reduced MT rescue events. Two seminal studies from the Verhey and Aumeier laboratories have shown that constitutively active Kinesin-1 induces MT lattice damage (Budaitis et al., 2022), which can be repaired through GTP-tubulin incorporation into "rescue shafts" that promote MT rescue (Andreu-Carbo et al., 2022). Extrapolating from these findings, loss of Kinesin-1 activity could plausibly reduce rescue shaft formation, thereby decreasing MT stability. While challenging to test directly in our system, this mechanism provides a plausible framework for the observed phenotype.

   2- Impaired transport of stabilizing factors. As that reviewer astutely points out, Khc transports the dynactin complex, an anti-catastrophe factor, to MT plus ends (Nieuwburg et al., 2017). Loss of this transport could further compromise MT plus end stability. We now discuss this important mechanism in the revised manuscript.

   3- Loss of cortical ncMTOCs. Critically, our new quantitative analyses (revised Figure 3 and Figure 5) also reveal defective anteroposterior orientation of cortical MTs in mutant conditions. These experiments suggest that Ens/Khc-mediated localization of ncMTOCs to the cortex is essential for proper MT network organization, and possibly minus-end stabilization as suggested in several studies (Feng et al., 2019, Goodwin and Vale, 2011, Nashchekin et al., 2016).

   Altogether, we now propose an integrated model in which MT reduction and disorganization may result from multiple complementary mechanisms operating downstream of Kinesin-1/Ensconsin loss. While some aspects remain difficult to test directly in our in vivo system, the convergence of our data with recent mechanistic studies provides an interesting conceptual framework. The Discussion has been revised to reflect this comprehensive view in a dedicated paragraph (“A possible regulation of MT dynamics in the oocyte at both plus end minus MT ends by Ens and Khc” lane 415-432).

   1. The Shot overexpression experiments presented in Fig.3 E-F, Fig.4D and TableS1 are very confusing. Originally , the authors used Shot-GFP overexpression at stage 9 to show that there is a decrease of ncMTOCs at the cortex in ens mutants (Fig.3 E-F) and speculated that this caused the defects in MT length and cytoplasmic advection at stage 10B. However the authors later state on page 8 that : "Shot overexpression (Shot OE) was sufficient to rescue the presence of long cortical MTs and ooplasmic advection in most ens oocytes (9/14), resembling the patterns observed in controls (Figures 4B right panel and 4D). Moreover, while ens females were fully sterile, overexpression of Shot was sufficient to restore that loss of fertility (Table S1)". Is this the same UAS Shot-GFP and VP16 Gal4 used in both experiments? If so, this contradictions puts the authors conclusions in question.

   This is an important point that requires clarification regarding our experimental design.

   The Shot-YFP construct is a genomic insertion on chromosome 3. The ens mutation is also located on chromosome 3 and we were unable to recombine this transgene with the ens mutant for live quantification of cortical Shot. To circumvent this technical limitation, we used a UAS-Shot.L(C)-GFP transgenic construct driven by a maternal driver, expressed in both wild-type (control) and ens mutant oocytes. We validated that the expression level and subcellular localization of UAS-Shot.L(C)-GFP were comparable to those of the genomic Shot-YFP (new Figure S8 A and B).

   From these experiments, we drew two key conclusions. First, cortical Shot.L(C)-GFP is less abundant in ens mutant oocytes compared to wild-type (the quantification has been removed from this version). Second, despite this reduced cortical accumulation, Shot.L(C)-GFP expression partially rescues ooplasmic flows and microtubule streaming in stage 10B ens mutant oocytes, and restores fertility to ens mutant females.

   1. The authors based they conclusions about the involvement of Ens, Kinesin-1 and Ninein in ncMTOC anchoring on the decrease in cortical fluorescence intensity of Shot-YFP and Patronin-YFP in the corresponding mutant backgrounds. However, there is a large variation in average Shot-YFP intensity between control oocytes in different experiments. In Fig. 2F-G the average level of Shot-YFP in the control sis 130 AU while in Fig.3 G-H it is only 55 AU. This makes me worry about reliability of such measurements and the conclusions drawn from them.

   To clarify this point, we have harmonized the method used to quantify the Shot-YFP signals in Figure 4E with the methodology used in Figure 3B, based on the original images. The levels are not strictly identical (Control Figure 2 B: 132.7+/-36.2 versus Control Figure 4 E: 164.0+/- 37.7). These differences are usual when experiments are performed at several-month intervals and by different users.

   1. The decrease in the intensity of Shot-YFP and Patronin-YFP cortical fluorescence in ens mutant oocytes could be because of problems with ncMTOC anchoring or with ncMTOCs formation. The authors should find a way to distinguish between these two possibilities. The authors could express Ens-Mut (described in Sung et al 2008), which localises at the oocyte posterior and test whether it recruits Shot/Patronin ncMTOCs to the posterior.

   We tried to obtain the fly stocks described in the 2008 paper by contacting former members of Pernille Rørth's laboratory. Unfortunately, we learned that the lab no longer exists and that all reagents, including the requested stocks, were either discarded or lost over time. To our knowledge, these materials are no longer available from any source. We regret that this limitation prevented us from performing the straightforward experiments suggested by the reviewer using these specific tools.

   1. According to the Materials and Methods, the Shot-GFP used in Fig.3 E-F and Fig.4 was the BDSC line 29042. This is Shot L(C), a full-length version of Shot missing the CH1 actin-binding domain that is crucial for Shot anchoring to the cortex. If the authors indeed used this version of Shot-GFP, the interpretation of the above experiments is very difficult.

   The Shot.L(C) isoform lacks the CH1 domain but retains the CH2 actin-binding motif. Truncated proteins with this domain and fused to GST retains a weak ability to bind actin in vitro. Importantly, the function of this isoform is context-dependent: it cannot rescue shot loss-of-function in neuron morphogenesis but fully restores Shot-dependent tracheal cell remodeling (Lee and Kolodziej, 2002).

   In our experiments, when the Shot.L(C) isoform was expressed under the control of a maternal driver, its localization to the oocyte cortex was comparable to that of the genomic Shot-YFP construct (new Figure S8). This demonstrates unambiguously that the CH1 domain is dispensable for Shot cortical localization in oocytes, and that CH2-mediated actin binding is sufficient for this localization. Of note, a recent study showed that actin network are not equivalent highlighting the need for specific Shot isoforms harboring specialized actin-binding domain (Nashchekin et al., 2024).

   We note that the expression level of Shot.L(C)-GFP in the oocyte appeared slightly lower than that of Shot-YFP (expressed under endogenous Shot regulatory sequences), as assessed by Western blot (Figure S8 A).

   Critically, Shot.L(C)-GFP expression was substantially lower than that of Shot.L(A)-GFP (that harbored both the CH1 and CH2 domain). Shot.L(A)-GFP was overexpressed (Figure 8 A) and ectopically localized on MTs in both nurse cells and the ooplasm (Figure S8 B middle panel and arrow). These observations are in agreement that the Shot.L(C)-GFP rescue experiment was performed at near-physiological expression levels, strengthening the validity of our conclusions.

   1. Page 6 "converted in NCs, in a region adjacent to the ring canals, Dendra-Ens-labeled MTs were found in the oocyte compartment indicating they are able to travel from NC toward the oocyte through ring canals". I have difficulty seeing the translocation of MT through the ring canals. Perhaps it would be more obvious with a movie/picture showing only one channel. Considering that f Dendra-Ens appears in the oocyte much faster than MT transport through ring canals (140nm/s, Lu et al 2022), the authors are most probably observing the translocation of free Ens rather than Ens bound to MT. The authors should also mention that Ens movement from the NC to the oocyte has been shown before with Ens MBD in Lu et al 2022 with better resolution.

   We fully agree on the caveat mentioned by this reviewer: we may observe the translocation of free Dendra-Ensconsin. The experiment, was removed and replaced by referring to the work of the Gelfand lab. The movement of MTs that travel at ~140 nm/s between nurse cells toward the oocyte through the Ring Canals was reported before by Lu et al. (2022) with a very good resolution. Notably, this directional directed movement of MTs was measured using a fusion protein encompassing Ens MT-binding domain. We decided to remove this inclusive experiment and rather refer to this relevant study.

   1. Page 6: The co-localization of Ninein with Ens and Shot at the oocyte cortex (Figure 2A). I have difficulty seeing this co-localisation. Perhaps it would be more obvious in merged images of only two channels and with higher resolution images

   2. "a pool of the Ens-GFP co-localized with Ch-Patronin at cortical ncMTOCs at the anterior cortex (Figure 3A)". I also have difficulty seeing this.

   We have performed new high-resolution acquisitions that provide clearer and more convincing evidence for the localization cortical distribution of these proteins (revised Figure 2A-2C and Figure 4A). These improved images demonstrate that Ens, Ninein, Shot, and Patronin partially colocalize at cortical ncMTOCs, as initially proposed. Importantly, the new data also reveal a spatial distinction: while Ens localizes along microtubules extending from these cortical sites, Ninein appears confined to small cytoplasmic puncta adjacent but also present on cortical microtubules.

   1. "Ninein co-localizes with Ens at the oocyte cortex and partially along cortical microtubules, contributing to the maintenance of high Ens protein levels in the oocyte and its proper cortical targeting". I could not find any data showing the involvement of Ninein in the cortical targeting of Ens.

   We found decreased Ens localization to MTs and to the cell cortex region (new Figure S3 A-B).

   1. "our MT network analyses reveal the presence of numerous short MTs cytoplasmic clustered in an anterior pattern." "This low cortical recruitment of ncMTOCs is consistent with poor MT anchoring and their cytoplasmic accumulation." I could not find any data showing that short cortical MT observed at stage 10b in ens mutant and Khc RNAi were cytoplasmic and poorly anchored.

   The sentence was removed from the revised manuscript.

   1. "The egg chamber consists of interconnected cells where Dynein and Khc activities are spatially separated. Dynein facilitates transport from NCs to the oocyte, while Khc mediates both transport and advection within the oocyte." Dynein is involved in various activities in the oocyte. It anchors the oocyte nucleus and transports bcd and grk mRNA to mention a few.

   The text was amended to reflect Dynein involvement in transport activities in the oocyte, with the appropriate references (lane 105-107).

   1. The cartoons in Fig.2H and 3I exaggerate the effect of Ninein and Ens on cortical ncMTOCs. According to the corresponding graphs, there is a 20 and 50% decrease in each case.

   New cartoons (now revised Figure 3E and 4F), are amended to reflect the ncMTOC values but also MT orientation (Figure 3E).

   Significance

   Given the important concerns raised, the significance of the findings is difficult to assess at this stage.

   We sincerely thank the reviewer for their thorough evaluation of our manuscript. We have carefully addressed their concerns through substantial new experiments and analyses. We hope that the revised manuscript, in its current form, now provides the clarifications and additional evidence requested, and that our responses demonstrate the significance of our findings.

   Reviewer #4 (Evidence, reproducibility and clarity (Required)):

   Summary: This manuscript presents an investigation into the molecular mechanisms governing spatial activation of Kinesin-1 motor protein during Drosophila oogenesis, revealing a regulatory network that controls microtubule organization and cytoplasmic transport. The authors demonstrate that Ensconsin, a MAP7 family protein and Kinesin-1 activator, is spatially enriched in the oocyte through a dual mechanism involving Dynein-mediated transport from nurse cells and cortical maintenance by Ninein. This spatial enrichment of Ens is crucial for locally relieving Kinesin-1 auto-inhibition. The Ens/Khc complex promotes cortical recruitment of non-centrosomal microtubule organizing centers (ncMTOCs), which are essential for anchoring microtubules at the cortex, enabling the formation of long, parallel microtubule streams or "twisters" that drive cytoplasmic advection during late oogenesis. This work establishes a paradigm where motor protein activation is spatially controlled through targeted localization of regulatory cofactors, with the activated motor then participating in building its own transport infrastructure through ncMTOC recruitment and microtubule network organization.

   There's a lot to like about this paper! The data are generally lovely and nicely presented. The authors also use a combination of experimental approaches, combining genetics, live and fixed imaging, and protein biochemistry.

   We thank the reviewer for this enthusiastic and supportive review, which helped us further strengthen the manuscript.

   Concerns: Page 6: "to assay if elevation of Ninein levels was able to mis-regulate Ens localization, we overexpressed a tagged Ninein-RFP protein in the oocyte. At stage 9 the overexpressed Ninein accumulated at the anterior cortex of the oocyte and also generated large cortical aggregates able to recruit high levels of Ens (Figures 2D and 2H)... The examination of Ninein/Ens cortical aggregates obtained after Ninein overexpression showed that these aggregates were also able to recruit high levels of Patronin and Shot (Figures 2E and 2H)." Firstly, I'm not crazy about the use of "overexpressed" here, since there isn't normally any Ninein-RFP in the oocyte. In these experiments it has been therefore expressed, not overexpressed. Secondly, I don't understand what the reader is supposed to make of these data. Expression of a protein carrying a large fluorescent tag leads to large aggregates (they don't look cortical to me) that include multiple proteins - in fact, all the proteins examined. I don't understand this to be evidence of anything in particular, except that Ninein-RFP causes the accumulation of big multi-protein aggregates. While I can understand what the authors were trying to do here, I think that these data are inconclusive and should be de-emphasized.

   We have revised the manuscript by replacing overexpressed with expressed (lanes 211 and 212). In addition, we now provide new localization data in both cortical (new Figure S4 A, top) and medial focal planes (new Figure S4 A, bottom), demonstrating that Ninein puncta (the word used in Rosen et al, 2019), rather than aggregates are located cortically. We also show that live IRP-labelled MTs do not colocalize with Ninein-RFP puncta. In light of the new experiments and the comments from the other reviewers, the corresponding text has been revised and de-emphasized accordingly.

   Page 7: "Co-immunoprecipitations experiments revealed that Patronin was associated with Shot-YFP, as shown previously (Nashchekin et al., 2016), but also with EnsWT-GFP, indicating that Ens, Shot and Patronin are present in the same complex (Figure 3B)." I do not agree that association between Ens-GFP and Patronin indicates that Ens is in the same complex as Shot and Patronin. It is also very possible that there are two (or more) distinct protein complexes. This conclusion could therefore be softened. Instead of "indicating" I suggest "suggesting the possibility."

   We have toned down this conclusion and indicated “suggesting the possibility” (lane 238-239).

   Page 7: "During stage 9, the average subcortical MT length, taken at one focal plane in live oocytes (see methods)..." I appreciate that the authors have been careful to describe how they measured MT length, as this is a major point for interpretation. I think the reader would benefit from an explanation of why they decided to measure in only one focal plane and how that decision could impact the results.

   We appreciate this helpful suggestion. Cortical microtubules are indeed highly dynamic and extend in multiple directions, including along the Z-axis. Moreover, their diameter is extremely small (approximately 25 nm), making it technically challenging to accurately measure their full length with high resolution using our Zeiss Airyscan confocal microscope (over several, microns): the acquisition of Z-stacks is relatively slow and therefore not well suited to capturing the rapid dynamics of these microtubules. Consequently, our length measurements represent a compromise and most likely underestimate the actual lengths of microtubules growing outside the focal plane. We note that other groups have encountered similar technical limitations (Parton et al., 2011).

   Page 7: "... the MTs exhibited an orthogonal orientation relative to the anterior cortex (Figures 4A left panels, 4C and 4E)." This phenotype might not be obvious to readers. Can it be quantified?

   We have now analyzed the orientation of microtubules (MTs) along the dorso-ventral axis. Our analysis shows that ens, Khc RNAi oocytes (new Figure 5B), and, to a lesser extent, Nin mutant oocytes (new Figure 3D), display a more random MT orientation compared to wild-type (WT) oocytes. In WT oocytes, MTs are predominantly oriented toward the posterior pole, consistent with previous findings (Parton et al., 2011).

   Page 8: "Altogether, the analyses of Ens and Khc defective oocytes suggested that MT organization defects during late oogenesis (stage 10B) were caused by an initial failure of ncMTOCs to reach the cell cortex. Therefore, we hypothesized that overexpression of the ncMTOC component Shot could restore certain aspects of microtubule cortical organization in ens-deficient oocytes. Indeed, Shot overexpression (Shot OE) was sufficient to rescue the presence of long cortical MTs and ooplasmic advection in most ens oocytes (9/14)..." The data are clear, but the explanation is not. Can the authors please explain why adding in more of an ncMTOC component (Shot) rescues a defect of ncMTOC cortical localization?

   We propose that cytoplasmic ncMTOCs can bind the cell cortex via the Shot subunit that is so far the only component that harbors actin-binding motifs. Therefore, we propose that elevating cytoplasmic Shot increase the possibility of Shot to encounter the cortex by diffusion when flows are absent. This is now explained lane 282-285.

   I'm grateful to the authors for their inclusion of helpful diagrams, as in Figures 1G and 2H. I think the manuscript might benefit from one more of these at the end, illustrating the ultimate model.

   We have carefully considered and followed the reviewer’s suggestions. In response, we have included a new figure illustrating our proposed model: the recruitment of ncMTOCs to the cell cortex through low Khc-mediated flows at stage 9 enhances cortical microtubule density, which in turn promotes self-amplifying flows (new Figure 7, panels A to C). Note that this Figure also depicts activation of Khc by loss of auto-inhibition (Figure 7, panel D).

   I'm sorry to say that the language could use quite a bit of polishing. There are missing and extraneous commas. There is also regular confusion between the use of plural and singular nouns. Some early instances include:

   1. Page 3: thought instead of "thoughted."
   2. Page 5: "A previous studies have revealed"
   3. Page 5: "A significantly loss"
   4. Page 6: "troughs ring canals" should be "through ring canals"
   5. Page 7: lives stage 9 oocytes
   6. Page 7: As ens and Khc RNAi oocytes exhibits
   7. Page 7: we examined in details
   8. Page 7: This average MT length was similar in Khc RNAi and ens mutant oocyte..

   We apologize for errors. We made the appropriate corrections of the manuscript.

   Reviewer #4 (Significance (Required)):

   This work makes a nice conceptual advance by showing that motor activation controls its own transport infrastructure, a paradigm that could extend to other systems requiring spatially regulated transport.

   We thank the reviewers for their evaluation of the manuscript and helpful comments.

   Read the original source
2. 

   Review Commons

   Nov 25, 2025

   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: This manuscript presents an investigation into the molecular mechanisms governing spatial activation of Kinesin-1 motor protein during Drosophila oogenesis, revealing a regulatory network that controls microtubule organization and cytoplasmic transport. The authors demonstrate that Ensconsin, a MAP7 family protein and Kinesin-1 activator, is spatially enriched in the oocyte through a dual mechanism involving Dynein-mediated transport from nurse cells and cortical maintenance by Ninein. This spatial enrichment of Ens is crucial for locally relieving Kinesin-1 auto-inhibition. The Ens/Khc complex promotes cortical recruitment …

   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: This manuscript presents an investigation into the molecular mechanisms governing spatial activation of Kinesin-1 motor protein during Drosophila oogenesis, revealing a regulatory network that controls microtubule organization and cytoplasmic transport. The authors demonstrate that Ensconsin, a MAP7 family protein and Kinesin-1 activator, is spatially enriched in the oocyte through a dual mechanism involving Dynein-mediated transport from nurse cells and cortical maintenance by Ninein. This spatial enrichment of Ens is crucial for locally relieving Kinesin-1 auto-inhibition. The Ens/Khc complex promotes cortical recruitment of non-centrosomal microtubule organizing centers (ncMTOCs), which are essential for anchoring microtubules at the cortex, enabling the formation of long, parallel microtubule streams or "twisters" that drive cytoplasmic advection during late oogenesis. This work establishes a paradigm where motor protein activation is spatially controlled through targeted localization of regulatory cofactors, with the activated motor then participating in building its own transport infrastructure through ncMTOC recruitment and microtubule network organization.

   There's a lot to like about this paper! The data are generally lovely and nicely presented. The authors also use a combination of experimental approaches, combining genetics, live and fixed imaging, and protein biochemistry.

   Concerns:

   Page 6: "to assay if elevation of Ninein levels was able to mis-regulate Ens localization, we overexpressed a tagged Ninein-RFP protein in the oocyte. At stage 9 the overexpressed Ninein accumulated at the anterior cortex of the oocyte and also generated large cortical aggregates able to recruit high levels of Ens (Figures 2D and 2H)... The examination of Ninein/Ens cortical aggregates obtained after Ninein overexpression showed that these aggregates were also able to recruit high levels of Patronin and Shot (Figures 2E and 2H)." Firstly, I'm not crazy about the use of "overexpressed" here, since there isn't normally any Ninein-RFP in the oocyte. In these experiments it has been therefore expressed, not overexpressed. Secondly, I don't understand what the reader is supposed to make of these data. Expression of a protein carrying a large fluorescent tag leads to large aggregates (they don't look cortical to me) that include multiple proteins - in fact, all the proteins examined. I don't understand this to be evidence of anything in particular, except that Ninein-RFP causes the accumulation of big multi-protein aggregates. While I can understand what the authors were trying to do here, I think that these data are inconclusive and should be de-emphasized.

   Page 7: "Co-immunoprecipitations experiments revealed that Patronin was associated with Shot-YFP, as shown previously (Nashchekin et al., 2016), but also with EnsWT-GFP, indicating that Ens, Shot and Patronin are present in the same complex (Figure 3B)." I do not agree that association between Ens-GFP and Patronin indicates that Ens is in the same complex as Shot and Patronin. It is also very possible that there are two (or more) distinct protein complexes. This conclusion could therefore be softened. Instead of "indicating" I suggest "suggesting the possibility."

   Page 7: "During stage 9, the average subcortical MT length, taken at one focal plane in live oocytes (see methods)..." I appreciate that the authors have been careful to describe how they measured MT length, as this is a major point for interpretation. I think the reader would benefit from an explanation of why they decided to measure in only one focal plane and how that decision could impact the results.

   Page 7: "... the MTs exhibited an orthogonal orientation relative to the anterior cortex (Figures 4A left panels, 4C and 4E)." This phenotype might not be obvious to readers. Can it be quantified?

   Page 8: "Altogether, the analyses of Ens and Khc defective oocytes suggested that MT organization defects during late oogenesis (stage 10B) were caused by an initial failure of ncMTOCs to reach the cell cortex. Therefore, we hypothesized that overexpression of the ncMTOC component Shot could restore certain aspects of microtubule cortical organization in ens-deficient oocytes. Indeed, Shot overexpression (Shot OE) was sufficient to rescue the presence of long cortical MTs and ooplasmic advection in most ens oocytes (9/14)..." The data are clear, but the explanation is not. Can the authors please explain why adding in more of an ncMTOC component (Shot) rescues a defect of ncMTOC cortical localization?

   I'm grateful to the authors for their inclusion of helpful diagrams, as in Figures 1G and 2H. I think the manuscript might benefit from one more of these at the end, illustrating the ultimate model.

   I'm sorry to say that the language could use quite a bit of polishing. There are missing and extraneous commas. There is also regular confusion between the use of plural and singular nouns. Some early instances include:

   1. Page 3: thought instead of "thoughted."
   2. Page 5: "A previous studies have revealed"
   3. Page 5: "A significantly loss"
   4. Page 6: "troughs ring canals" should be "through ring canals"
   5. Page 7: lives stage 9 oocytes
   6. Page 7: As ens and Khc RNAi oocytes exhibits
   7. Page 7: we examined in details
   8. Page 7: This average MT length was similar in Khc RNAi and ens mutant oocyte..

   Significance

   This work makes a nice conceptual advance by showing that motor activation controls its own transport infrastructure, a paradigm that could extend to other systems requiring spatially regulated transport.

   Read the original source
3. 

   Review Commons

   Nov 25, 2025

   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

   The manuscript of Berisha et al., investigates the role of Esconsin (Ens), Kinesin-1 and Ninein in organisation of microtubules (MT) in Drosophila oocyte. At stage 9 oocytes Kinesin-1 transports oskar mRNA, a posterior determinant, along MT that are organised by ncMTOCs. At stage 10b, Kinesin-1 induces cytoplasmic advection to mix the contents of the oocyte. Ensconsin/Map7 is a MT associated protein (MAP) that uses its MT-binding domain (MBD) and kinesin binding domain (KBD) to recruit Kinesin-1 to the microtubules and to stimulate the motility of MT-bound Kinesin-1. Using various new Ens transgenes, the authors demonstrate the …

   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

   The manuscript of Berisha et al., investigates the role of Esconsin (Ens), Kinesin-1 and Ninein in organisation of microtubules (MT) in Drosophila oocyte. At stage 9 oocytes Kinesin-1 transports oskar mRNA, a posterior determinant, along MT that are organised by ncMTOCs. At stage 10b, Kinesin-1 induces cytoplasmic advection to mix the contents of the oocyte. Ensconsin/Map7 is a MT associated protein (MAP) that uses its MT-binding domain (MBD) and kinesin binding domain (KBD) to recruit Kinesin-1 to the microtubules and to stimulate the motility of MT-bound Kinesin-1. Using various new Ens transgenes, the authors demonstrate the requirement of Ens MBD and Ninein in Ens localisation to the oocyte where Ens activates Kinesin-1 using its KBD. The authors also claim that Ens, Kinesin-1 and Ninein are required for the accumulation of ncMTOCs at the oocyte cortex and argue that the detachment of the ncMTOCs from the cortex accounts for the reduced localisation of oskar mRNA at stage 9 and the lack of cytoplasmic streaming at stage 10b.

   Although the manuscript contains several interesting observations, the authors' conclusions are not sufficiently supported by their data. The structure function analysis of Ensconsin (Ens) is potentially publishable, but the conclusions on ncMTOC anchoring and cytoplasmic streaming not convincing

   1. The main conclusion of the manuscript is that "MT advection failure in Khc and ens in late oogenesis stems from defective cortical ncMTOCs recruitment". This completely overlooks the abundant evidence that Kinesin-1 directly drives cytoplasmic streaming by transporting vesicles and microtubules along microtubules, which then move the cytoplasm by advection (Palacios et al., 2002; Serbus et al, 2005; Lu et al, 2016). Since Kinesin-1 generates the flows, one cannot conclude that the effect of khc and ens mutants on cortical ncMTOC positioning has any direct effect on these flows, which do not occur in these mutants.
   2. The authors claim that streaming phenotypes of ens and khs mutants are due to a decrease in microtubule length caused by the defective localisation of ncMTOCs. In addition to the problem raised above, However, I am not convinced that they can make accurate measurements of microtubule length from confocal images like those shown in Figure 4. Firstly, they are measuring the length of bundles of microtubules and cannot resolve individual microtubules. This problem is compounded by the fact that the microtubules do not align into parallel bundles in the mutants. This will make the "microtubules" appear shorter in the mutants. In addition, the alignment of the microtubules in wild-type allows one to choose images in which the microtubule lie in the imaging plane, whereas the more disorganised arrangement of the microtubules in the mutants means that most microtubules will cross the imaging plane, which precludes accurate measurements of their length.
   3. "To investigate whether the presence of these short microtubules in ens and Khc RNAi oocytes is due to defects in microtubule anchoring or is also associated with a decrease in microtubule polymerization at their plus ends, we quantified the velocity and number of EB1comets, which label growing microtubule plus ends (Figure S3)." I do not understand how the anchoring or not of microtubule minus ends to the cortex determines how far their plus ends grow, and these measurements fall short of showing that plus end growth is unaffected. It has already been shown that the Kinesin-1-dependent transport of Dynactin to growing microtubule plus ends increases the length of microtubules in the oocyte because Dynactin acts as an anti-catastrophe factor at the plus ends. Thus, khc mutants should have shorter microtubules independently of any effects on ncMTOC anchoring. The measurements of EB1 comet speed and frequency in FigS2 will not detect this change and are not relevant for their claims about microtubule length. Furthermore, the authors measured EB1 comets at stage 9 (where they did not observe short MT) rather than at stage 10b. The authors' argument would be better supported if they performed the measurements at stage 10b.
   4. The Shot overexpression experiments presented in Fig.3 E-F, Fig.4D and TableS1 are very confusing. Originally , the authors used Shot-GFP overexpression at stage 9 to show that there is a decrease of ncMTOCs at the cortex in ens mutants (Fig.3 E-F) and speculated that this caused the defects in MT length and cytoplasmic advection at stage 10B. However the authors later state on page 8 that : "Shot overexpression (Shot OE) was sufficient to rescue the presence of long cortical MTs and ooplasmic advection in most ens oocytes (9/14), resembling the patterns observed in controls (Figures 4B right panel and 4D). Moreover, while ens females were fully sterile, overexpression of Shot was sufficient to restore that loss of fertility (Table S1)". Is this the same UAS Shot-GFP and VP16 Gal4 used in both experiments? If so, this contradictions puts the authors conclusions in question.
   5. The authors based they conclusions about the involvement of Ens, Kinesin-1 and Ninein in ncMTOC anchoring on the decrease in cortical fluorescence intensity of Shot-YFP and Patronin-YFP in the corresponding mutant backgrounds. However, there is a large variation in average Shot-YFP intensity between control oocytes in different experiments. In Fig. 2F-G the average level of Shot-YFP in the control sis 130 AU while in Fig.3 G-H it is only 55 AU. This makes me worry about reliability of such measurements and the conclusions drawn from them.
   6. The decrease in the intensity of Shot-YFP and Patronin-YFP cortical fluorescence in ens mutant oocytes could be because of problems with ncMTOC anchoring or with ncMTOCsformation. The authors should find a way to distinguish between these two possibilities. The authors could express Ens-Mut (described in Sung et al 2008), which localises at the oocyte posterior and test whether it recruits Shot/Patronin ncMTOCs to the posterior.
   7. According to the Materials and Methods, the Shot-GFP used in Fig.3 E-F and Fig.4 was the BDSC line 29042. This is Shot L(C), a full-length version of Shot missing the CH1 actin-binding domain that is crucial for Shot anchoring to the cortex. If the authors indeed used this version of Shot-GFP, the interpretation of the above experiments is very difficult.
   8. Page 6 "converted in NCs, in a region adjacent to the ring canals, Dendra-Ens-labeled MTs were found in the oocyte compartment indicating they are able to travel from NC toward the oocyte trough ring canals". I have difficulty seeing the translocation of MT through the ring canals. Perhaps it would be more obvious with a movie/picture showing only one channel. Considering that f Dendra-Ens appears in the oocyte much faster than MT transport through ring canals (140nm/s, Lu et al 2022) , the authors are most probably observing the translocation of free Ens rather than Ens bound to MT. The authors should also mention that Ens movement from the NC to the oocyte has been shown before with Ens MBD in Lu et al 2022 with better resolution.
   9. Page 6: The co-localization of Ninein with Ens and Shot at the oocyte cortex (Figure 2A). I have difficulty seeing this co-localisation. Perhaps it would be more obvious in merged images of only two channels and with higher resolution images
   10. "a pool of the Ens-GFP co-localized with Ch-Patronin at cortical ncMTOCs at the anterior cortex (Figure 3A)". I also have difficulty seeing this.
   11. "Ninein co-localizes with Ens at the oocyte cortex and partially along cortical microtubules, contributing to the maintenance of high Ens protein levels in the oocyte and its proper cortical targeting". I could not find any data showing the involvement of Ninein in the cortical targeting of Ens.
   12. "our MT network analyses reveal the presence of numerous short MTs cytoplasmic clustered in an anterior pattern." "This low cortical recruitment of ncMTOCs is consistent with poor MT anchoring and their cytoplasmic accumulation." I could not find any data showing that short cortical MT observed at stage 10b in ens mutant and Khc RNAi were cytoplasmic and poorly anchored.
   13. "The egg chamber consists of interconnected cells where Dynein and Khc activities are spatially separated. Dynein facilitates transport from NCs to the oocyte, while Khc mediates both transport and advection within the oocyte." Dynein is involved in various activities in the oocyte. It anchors the oocyte nucleus and transports bcd and grk mRNA to mention a few.
   14. The cartoons in Fig.2H and 3I exaggerate the effect of Ninein and Ens on cortical ncMTOCs. According to the corresponding graphs, there is a 20 and 50% decrease in each case.

   Significance

   Given the important concerns raised, the significance of the findings is difficult to assess at this stage.

   Read the original source
4. 

   Review Commons

   Nov 25, 2025

   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, Berisha et al. investigate how microtubule (MT) organization is spatially regulated during Drosophila oogenesis. The authors identify a mechanism in which the Kinesin-1 activator Ensconsin/MAP7 is transported by dynein and anchored at the oocyte cortex via Ninein, enabling localized activation of Kinesin-1. Disruption of this pathway impairs ncMTOC recruitment and MT anchoring at the cortex. The authors combine genetic manipulation with high-resolution microscopy and use three key readouts to assess MT organization during mid-to-late oogenesis: cortical MT formation, localization of posterior determinants, and …

   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, Berisha et al. investigate how microtubule (MT) organization is spatially regulated during Drosophila oogenesis. The authors identify a mechanism in which the Kinesin-1 activator Ensconsin/MAP7 is transported by dynein and anchored at the oocyte cortex via Ninein, enabling localized activation of Kinesin-1. Disruption of this pathway impairs ncMTOC recruitment and MT anchoring at the cortex. The authors combine genetic manipulation with high-resolution microscopy and use three key readouts to assess MT organization during mid-to-late oogenesis: cortical MT formation, localization of posterior determinants, and ooplasmic streaming. Notably, Kinesin-1, in concert with its activator Ens/MAP7, contributes to organizing the microtubule network it travels along. Overall, the study presents interesting findings, though we have several concerns we would like the authors to address.

   Ensconsin enrichment in the oocyte

   1. Enrichment in the oocyte
      • Ensconsin is a MAP that binds MTs. Given that microtubule density in the oocyte significantly exceeds that in the nurse cells, its enrichment may passively reflect this difference. To assess whether the enrichment is specific, could the authors express a non-Drosophila MAP (e.g., mammalian MAP1B) to determine whether it also preferentially localizes to the oocyte?
      • The ability of ens-wt and ens-LowMT to induce tubulin polymerization according to the light scattering data (Fig. S1J) is minimal and does not reflect dramatic differences in localization. The authors should verify that, in all cases, the polymerization product in their in vitro assays is microtubules rather than other light-scattering aggregates. What is the control in these experiments? If it is just purified tubulin, it should not form polymers at physiological concentrations.
   2. Photoconversion caveats MAPs are known to dynamically associate and dissociate from microtubules. Therefore, interpretation of the Ens photoconversion data should be made with caution. The expanding red signal from the nurse cells to the oocyte may reflect a any combination of dynein-mediated MT transport and passive diffusion of unbound Ensconsin. Notably, photoconversion of a soluble protein in the nurse cells would also result in a gradual increase in red signal in the oocyte, independent of active transport. We encourage the authors to more thoroughly discuss these caveats. It may also help to present the green and red channels side by side rather than as merged images, to allow readers to assess signal movement and spatial patterns better.
   3. Reduction of Shot at the anterior cortex
      • Shot is known to bind strongly to F-actin, and in the Drosophila ovary, its localization typically correlates more closely with F-actin structures than with microtubules, despite being an MT-actin crosslinker. Therefore, the observed reduction of cortical Shot in ens, nin mutants, and Khc-RNAi oocytes is unexpected. It would be important to determine whether cortical F-actin is also disrupted in these conditions, which should be straightforward to assess via phalloidin staining.
      • MTs are barely visible in Fig. 3A, which is meant to demonstrate Ens-GFP colocalization with tubulin. Higher-quality images are needed.
   4. MT gradient in stage 9 oocytes In ens-/-, nin-/-, and Khc-RNAi oocytes, is there any global defect in the stage 9 microtubule gradient? This information would help clarify the extent to which cortical localization defects reflect broader disruptions in microtubule polarity.
   5. Role of Ninein in cortical anchoring The requirement for Ninein in cortical anchorage is the least convincing aspect of the manuscript and somewhat disrupts the narrative flow. First, it is unclear whether Ninein exhibits the same oocyte-enriched localization pattern as Ensconsin. Is Ninein detectable in nurse cells? Second, the Ninein antibody signal appears concentrated in a small area of the anterior-lateral oocyte cortex (Fig. 2A), yet Ninein loss leads to reduced Shot signal along a much larger portion of the anterior cortex (Fig. 2F)-a spatial mismatch that weakens the proposed functional relationship. Third, Ninein overexpression results in cortical aggregates that co-localize with Shot, Patronin, and Ensconsin. Are these aggregates functional ncMTOCs? Do microtubules emanate from these foci?
   6. Inconsistency of Khc^MutEns rescue The Khc^MutEns variant partially rescues cortical MT formation and restores a slow but measurable cytoplasmic flow yet it fails to rescue Staufen localization (Fig. 5). This raises questions about the consistency and completeness of the rescue. Could the authors clarify this discrepancy or propose a mechanistic rationale?

   Minor points:

   1. The pUbi-attB-Khc-GFP vector was used to generate the Khc^MutEns transgenic line, presumably under control of the ubiquitous ubi promoter. Could the authors specify which attP landing site was used? Additionally, are the transgenic flies viable and fertile, given that Kinesin-1 is hyperactive in this construct?
   2. On page 11 (Discussion, section titled "A dual Ensconsin oocyte enrichment mechanism achieves spatial relief of Khc inhibition"), the statement "many mutations in Kif5A are causal of human diseases" would benefit from a brief clarification. Since not all readers may be familiar with kinesin gene nomenclature, please indicate that KIF5A is one of the three human homologs of Kinesin heavy chain.
   3. On page 16 (Materials and Methods, "Immunofluorescence in fly ovaries"), the sentence "Ovaries were mounted on a slide with ProlonGold medium with DAPI (Invitrogen)" should be corrected to "ProLong Gold."

   Significance

   This study shows that enrichment of MAP7/ensconsin in the oocyte is the mechanism of kinesin-1 activation there and is important for cytoplasmic streaming and localization non-centrosomal microtubule-organizing centers to the oocyte cortex

   Read the original source
5. 

   Review Commons

   Nov 25, 2025

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

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   This paper addresses a very interesting problem of non-centrosomal microtubule organization in developing Drosophila oocytes. Using genetics and imaging experiments, the authors reveal an interplay between the activity of kinesin-1, together with its essential cofactor Ensconsin, and microtubule organization at the cell cortex by the spectraplakin Shot, minus-end binding protein Patronin and Ninein, a protein implicated in microtubule minus end anchoring. The authors demonstrate that the loss of Ensconsin affects the cortical accumulation non-centrosomal microtubule organizing center (ncMTOC) proteins, microtubule length and vesicle …

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

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   This paper addresses a very interesting problem of non-centrosomal microtubule organization in developing Drosophila oocytes. Using genetics and imaging experiments, the authors reveal an interplay between the activity of kinesin-1, together with its essential cofactor Ensconsin, and microtubule organization at the cell cortex by the spectraplakin Shot, minus-end binding protein Patronin and Ninein, a protein implicated in microtubule minus end anchoring. The authors demonstrate that the loss of Ensconsin affects the cortical accumulation non-centrosomal microtubule organizing center (ncMTOC) proteins, microtubule length and vesicle motility in the oocyte, and show that this phenotype can be rescued by constitutively active kinesin-1 mutant, but not by Ensconsin mutants deficient in microtubule or kinesin binding. The functional connection between Ensconsin, kinesin-1 and ncMTOCs is further supported by a rescue experiment with Shot overexpression. Genetics and imaging experiments further implicate Ninein in the same pathway. These data are a clear strength of the paper; they represent a very interesting and useful addition to the field.

   The weaknesses of the study are two-fold. First, the paper seems to lack a clear molecular model, uniting the observed phenomenology with the molecular functions of the studied proteins. Most importantly, it is not clear how kinesin-based plus-end directed transport contributes to cortical localization of ncMTOCs and regulation of microtubule length.

   Second, not all conclusions and interpretations in the paper are supported by the presented data. Below is a list of specific comments, outlining the concerns, in the order of appearance in the paper/figures.

   1. Figure 1. The statement: "Ens loading on MTs in NCs and their subsequent transport by Dynein toward ring canals promotes the spatial enrichment of the Khc activator Ens in the oocyte" is not supported by data. The authors do not demonstrate that Ens is actually transported from the nurse cells to the oocyte while being attached to microtubules. They do show that the intensity of Ensconsin correlates with the intensity of microtubules, that the distribution of Ensconsin depends on its affinity to microtubules and that an Ensconsin pool locally photoactivated in a nurse cell can redistribute to the oocyte (and throughout the nurse cell) by what seems to be diffusion. The provided images suggest that Ensconsin passively diffuses into the oocyte and accumulates there because of higher microtubule density, which depends on dynein. To prove that Ensconsin is indeed transported by dynein in the microtubule-bound form, one would need to measure the residence time of Ensconsin on microtubules and demonstrate that it is longer than the time needed to transport microtubules by dynein into the oocyte; ideally, one would like to see movement of individual microtubules labelled with photoconverted Ensconsin from a nurse cell into the oocyte. Since microtubules are not enriched in the oocyte of the dynein mutant, analysis of Ensconsin intensity in this mutant is not informative and does not reveal the mechanism of Ensconsin accumulation.
   2. Figure 2. According to the abstract, this figure shows that Ensconsin is "maintained at the oocyte cortex by Ninein". However, the figure doesn't seem to prove it - it shows that oocyte enrichment of Ensonsin is partially dependent on Ninein, but this applies to the whole cell and not just to the cell cortex. Furthermore, it is not clear whether Ninein mutation affects microtubule density, which in turn would affect Ensconsin enrichment, and therefore, it is not clear whether the effect of Ninein loss on Ensconsin distribution is direct or indirect. The observation that the aggregates formed by overexpressed Ninein accumulate other proteins, including Ensconsin, supports, though does not prove their interactions. Furthermore, there is absolutely no proof that Ninein aggregates are "ncMTOCs". Unless the authors demonstrate that these aggregates nucleate or anchor microtubules (for example, by detailed imaging of microtubules and EB1 comets), the text and labels in the figure would need to be altered.

   Minor comment: Note that a "ratio" (Figure 2C) is just a ratio, and should not be expressed in arbitrary units.

   1. Figure 3B: immunoprecipitation results cannot be interpreted because the immunoprecipitated proteins (GFP, Ens-GFP, Shot-YFP) are not shown. It is also not clear that this biochemical experiment is useful. If the authors would like to suggest that Ensconsin directly binds to Patronin, the interaction would need to be properly mapped at the protein domain level.
   2. One of the major phenotypes observed by the authors in Ens mutant is the loss of long microtubules. The authors make strong conclusions about the independence of this phenotype from the parameters of microtubule plus-end growth, but in fact, the quality of their data does not allow to make such a conclusion, because they only measured the number of EB1 comets and their growth rate but not the catastrophe, rescue or pausing frequency. Note that kinesin-1 has been implicated in promoting microtubule damage and rescue (doi: 10.1016/j.devcel.2021). In the absence of such measurements, one cannot conclude whether short microtubules arise through defects in the minus-end, plus-end or microtubule shaft regulation pathways. It is important to note in that a spectraplakin, like Shot, can potentially affect different pathways, particularly when overexpressed. Unjustified conclusions should be removed: the authors do not provide sufficient data to conclude that "ens and Khc oocytes MT organizational defects are caused by decreased ncMTOC cortical anchoring", because the actual cortical microtubule anchoring was not measured.

   Minor comment: Microtubule growth velocity must be expressed in units of length per time, to enable evaluating the quality of the data, and not as a normalized value.

   1. A significant part of the Discussion is dedicated to the potential role of Ensconsin in cortical microtubule anchoring and potential transport of ncMTOCs by kinesin. It is obviously fine that the authors discuss different theories, but it would be very helpful if the authors would first state what has been directly measured and established by their data, and what are the putative, currently speculative explanations of these data.

   Minor comment: The writing and particularly the grammar need to be significantly improved throughout, which should be very easy with current language tools. Examples: "ncMTOCs recruitment" should be "ncMTOC recruitment"; "Vesicles speed" should be "Vesicle speed", "Nin oocytes harbored a WT growth,"- unclear what this means, etc. Many paragraphs are very long and difficult to read. Making shorter paragraphs would make the authors' line of thought more accessible to the reader.

   Significance

   This paper represents significant advance in understanding non-centrosomal microtubule organization in general and in developing Drosophila oocytes in particular by connecting the microtubule minus-end regulation pathway to the Kinesin-1 and Ensconsin/MAP7-dependent transport. The genetics and imaging data are of good quality, are appropriately presented and quantified. These are clear strengths of the study which will make it interesting to researchers studying the cytoskeleton, microtubule-associated proteins and motors, and fly development.

   The weaknesses of this study are due to the lack of clarity of the overall molecular model, which would limit the impact of the study on the field. Some interpretations are not sufficiently supported by data, but this can be solved by more precise and careful writing, without extensive additional experimentation.

   My expertise is cell biology and biochemistry of the microtubule cytoskeleton, including both microtubule-associated proteins and microtubule motors.

   Read the original source
6. 

   Version published to 10.1101/2025.04.15.648882 on bioRxiv

   Apr 16, 2025