Author response:
The following is the authors’ response to the original reviews.
Public Reviews:
Reviewer #1 (Public Review):
The manuscript investigates the role of the membrane-deforming cytoskeletal regulator protein Abba in cortical development and its potential implications for microcephaly. It is a valuable contribution to the understanding of Abba's role in cortical development. The strengths and weaknesses identified in the manuscript are outlined below:
Clinical Relevance:
The authors identified a patient with microcephaly and a patient with an intellectual disability harboring a mutation in the Abba variant (R671W) adding a clinically relevant dimension to the study.
Mechanistic Insights:
The study offers valuable mechanistic insights into the development of microcephaly by elucidating the role of Abba in radial glial cell proliferation, radial fiber organization, and the migration of neuronal progenitors. The identification of Abba's involvement in the cleavage furrow during cell division, along with its interaction with Nedd9 and positive influence on RhoA activity, adds depth to our understanding of the molecular processes governing cortical development. Though the reported results establish the novel interaction between Abba and Nedd9, the authors have not addressed whether the mutant protein loses this interaction and whether that results in the observed effects.
We appreciate the reviewer’s observation and fully agree that our study does not provide direct evidence that the phenotypes induced by the R671W mutant are mediated through NEDD9. We sincerely apologize if the manuscript inadvertently conveyed this impression.
While we show that the interaction with NEDD9 plays a role in the action of ABBA, our findings suggest that NEDD9 and RhoA activation have a minor influence on the phenotypes induced by this mutation, as highlighted by the evidence we presented.
We would like to point out that we have previously addressed this point in the discussion section of the manuscript. For clarity, below is an excerpt from that section:
“heterozygous expression of the human R671W variant would exert a dominant negative effect on ABBA's role in brain development, leading to microcephaly and cognitive delay. This notion is supported by recent work disclosing additional patient carrying the R671W variant42. In the same study the significant neurological phenotypes were observed in a drosophila model where the ortholog of human MTSS2 and MTSS1 mim was deleted. However, from a clinical genetics’ standpoint, it is unlikely to find patients with the recurrent R671W mutation without any homozygous or compound heterozygous loss-of-function mutations elsewhere in the ABBA gene. This could also suggest a gain-of-function effect of the R671W mutation. Supporting this notion, overexpressing ABBA-R671W in cells expressing the wild-type Abba in this study did not result in a dominant-negative decrease in RhoA activation, nor did it affect the expression of PH3 in vivo. These findings make it plausible to suggest that a mechanism responsible for the phenotype associated with overexpression of the human variant may primarily involve post-cell division processes, such as cell migration. “
We have made corrections to the new version of the manuscript to emphasize this further.
In Vivo Validation:
The overexpression of mutant Abba protein (R671W) resulting in phenotypic similarities to Abba knockdown effects supports the significance of Abba in cortical development.
Reviewer #2 (Public Review):
Summary:
Carabalona and colleagues investigated the role of the membrane-deforming cytoskeletal regulator protein Abba (MTSS1L/MTSS2) in cortical development to better understand the mechanisms of abnormal neural stem cell mitosis. The authors used short hairpin RNA targeting Abba20 with a fluorescent reporter coupled with in-utero electroporation of E14 mice to show changes to neural progenitors. They performed flow cytometry for in-depth cell cycle analysis of Abba-shRNA impact on neural progenitors and determined an accumulation in the S phase. Using culture rat glioma cells and live imaging from cortical organotypic slides from mice in utero electroporated with Abba-shRNA, the authors found Abba played a prominent role in cytokinesis. They then used a yeast-two-hybrid screen to identify three high-confidence interactors: Beta-Trcp2, Nedd9, and Otx2. They used immunoprecipitation experiments from E18 cortical tissue coupled with C6 cells to show Abba's requirement for Nedd9 localization to the cleavage furrow/cytokinetic bridge. The authors performed a shRNA knockdown of Nedd9 by in-utero electroporation of E14 mice and observed similar results as with the Abba-shRNA. They tested a human variant of Abba using in-utero electroporation of cDNA and found disorganized radial glial fibers and misplaced, multipolar neurons, but lacked the impact of cell division seen in the shRNA-Abba model.
Strengths:
A fundamental question in biology about the mechanics of neural stem cell division.
Directly connecting effects in Abba protein to downstream regulation of RhoA via Nedd9.
Incorporation of human mutation in ABBA gene.
Use of novel technologies in neurodevelopment and imaging.
Weaknesses:
Unexplored components of the pathway (such as what neurogenic populations are impacted by Abba mutation) and unleveraged aspects of their data (such as the live imaging) limit the scope of their findings and leave significant questions about the effect of ABBA on radial glia development.
(1) The claim of disorganized radial glial fibers lacks quantifications.
On page 11, the authors claim that knockdown of Abba leads to changes in radial glial morphology observed with vimentin staining. Here they claim misoriented apical processes, detached end feet, and decreased number of RGP cells in the VZ. However, they do not provide quantification of process orientation to better support their first claim. Measurements of radial glia fiber morphology (directionality, length) and angle of division would be metrics that can be applied to data.
In the corrected version of the manuscript, we provide new qualification of changes in dispersion of vimentin immunostaining (Supplementary Figure 1).
Some of these analyses could be done in their time-lapse microscopy images, such as to quantify the number of cell divisions during their period of analysis (though that is short-15 hours).
This is indeed a very good idea. We have reanalyzed the recordings to follow cell division. Unfortunately, the number of cells that we were able to follow was low, making statistical analysis of the data unreliable. As the reviewer alluded in the comment longer recording times than 15h are required to make reliable conclusion. Instead, we have performed live-cell imaging using Aniling-GFP coelectroporeted with RFP as a marker of mitotic progression . We monitored the distribution of cells showing accumulation of Anillin-GFP in control (Scramble) and ABBA-shRNA3 conditions (this data was added to new Supplementary Figure 3). Anillin has been shown to be an efficient tool to monitor cell division in vivo as in particular as it displays accumulation and correlated increase intensity of Anillin-GFP ((Hesse et al Nature Com. 2012, DOI: 10.1038/ncomms2089).
(2) It is unclear where the effect is:
-In RG or neuroblasts? Is it in cell cleavage that results in the accumulation of cells at VZ (as sometimes indicated by their data like in Figure 2A or 4D)?
The data suggest that radial glial (RG) cells are indeed blocked prior to abscission. This phenomenon might contribute to the accumulation of cells at the ventricular zone (VZ), as indicated by observations such as those in Figure 2A and 4D. The interruption in cell cleavage likely prevents the proper progression of division, causing RG cells to remain at the VZ rather than proceeding with their normal differentiation or migration processes. This finding highlights a potential mechanistic link between disrupted abscission and cell accumulation in the VZ.
Interrogation of cell death (such as by cleaved caspase 3) would also help.
Caspase-3 cleavage is widely used as a marker for apoptosis; however, it may not be the most reliable tool for monitoring apoptosis during brain cortical development. The developing brain is a highly dynamic environment where caspase-3 activation can be transient and involved in non-apoptotic processes, such as synaptic pruning and neuronal remodeling. This makes it challenging to distinguish caspase-3 activity associated with apoptosis from its roles in physiological processes.
In contrast, monitoring overall cell survival provides a more reliable measure of developmental outcomes, as it reflects the net balance of cell death and survival mechanisms. By focusing on cell survival e.g. quantification of number of RGP, we can better assess the functional consequences of apoptosis and its interplay with neurogenesis and other developmental processes. In line with this we have added more data on the quantification of RGPC as well as their distribution in new Supplementary Figure 3.
Given their time-lapse, can they identify what is happening to the RG fiber?
Both apical and basal endfeet appear to detach and retract prior to radial glial (RG) cell death. This is evident in Figure 1D, as well as from our observation of cellular bodies located far from the ventricular surface (VS), as demonstrated in the new Supplementary Figure 3.
The authors describe a change in "migration" but do not show evidence for this for either progenitor or neuroblast populations. Given they have nice time-lapse imaging data, could they visualize progenitor versus young neuron migration? Analysis of neuroblasts (such as with doublecortin expression in the tissue) would also help understand any issues in migration (of neurons v stem cells).
This is an excellent question that arises from the extensive data presented in this study. Addressing it would require repeating a significant portion of the experiments. We fully agree with the reviewer that these are important and obvious questions that warrant a dedicated study to answer them thoroughly. Additionally, we believe that the data showing the accumulation of migrating electroporated cells in the ventricular (V) and subventricular (SV) zones provide compelling evidence of abnormal migration in ABBA-shRNA electroporated cells.
-At cleavage furrow? In abscission? There is high-resolution data that highlights the cleavage furrow as the location of interest (Figure 3A), however, there is also data (Figure 3B) to suggest Abba is expressed elsewhere as well and there is an overall soma decrease. More detail of the localization of Abba during the division process would be helpful for example, could cleavage furrow proteins, such as Aurora B, co-localization (and potentially co-IP) help delineate subpopulations of Abba protein? Furthermore, the FRET imaging is a unique way to connect their mutation with function - could they measure/quantify differences at furrow compared to the rest of soma to further corroborate that the Abba-associated RhoA effect was furrow-enriched?
In the corrected version of the manuscript, we include new quantification of RhoA activity in the region corresponding to the cleavage furrow (New Figure 5), This new data show similar results as the previous and indicate that the changes observed are primarily derived from the cleavage furrow region. In the future a detailed dissection of the molecules involved in the mechanism would be highly desirable. These notions are now included in the discussion.
-The data highlights nicely that a furrow doesn't clearly form when ABBA expression and subsequent RhoA activity are decreased (in Figure 3 or 5A). Does this lead to cells that can't divide because of poor abscission, especially since "rounding" still occurs? Or abnormal progenitors (with loss of fiber or inability to support neuroblast migration)? Or abnormal progression of progenitors to neuroblasts?
Our findings, combined with previous results, suggest multiple mechanisms through which ABBA depletion and subsequent Nedd9 and RhoA signaling disruptions could impact progenitor cells and neuroblasts. Below is a detailed response to each question:
(1) Do cells fail to divide due to poor abscission?
Nedd9 is a key regulator of RhoA signaling, which could be essential for cleavage furrow ingression and abscission. Reduced Nedd9 expression may leads to non-activation of RhoA, thereby impairing cleavage furrow ingression. Furthermore, since RhoA deactivation is critical for successful abscission, any disruption in this signaling pathway could compromise the final stages of cytokinesis. While we do not directly observe failed abscission, the impaired furrow formation in Figure 3 and 5A aligns with the hypothesis that some cells may struggle to complete division due to defects in RhoA-mediated abscission.
(2) Are abnormal progenitors generated (e.g., loss of fiber or inability to support neuroblast migration)?
Disrupted Nedd9 expression not only affects cell cycle progression but also influences the structural integrity of radial glial progenitors (RGPs). RGPs with impaired cleavage furrow ingression may exhibit detachment of apical and basal endfeet (Supplementary Figure 3), leading to abnormalities in their scaffold function. This structural disruption likely contributes to the accumulation of electroporated cells in the ventricular (V) and subventricular (SV) zones (Figure 5A), supporting the idea that abnormal progenitors fail to support proper neuroblast migration.
(3) Is there abnormal progression of progenitors to neuroblasts?
Given that Nedd9 triggers cells to enter mitosis, its impaired function may prevent progenitors from properly progressing through the cell cycle, causing cell cycle arrest and eventual decrease survival. This would directly impact the ability of progenitors to transition into neuroblasts. Moreover, the abnormal membrane composition and PI(4,5)P2 enrichment we hypothesize during cytokinesis could disrupt ABBA recruitment and its interaction with Nedd9. This disruption would impair RhoA activation, further compromising the progression of progenitors to neuroblasts.
In conclusion, our findings suggest that impaired ABBA expression disrupts Nedd9 and RhoA signaling, leading to poor cleavage furrow ingression, abnormal progenitor structure, and defective neuroblast migration. These processes collectively contribute to developmental defects in the cortex. Future studies focusing on live imaging of cytokinesis and cell fate mapping will help elucidate better these mechanisms further.
(3) Limited to a singular time point of mouse cortical development
On page 13, the authors outline the results of their Y2H screen with the identification of three high-confidence interactors. Notably, they used an E10.5-E12.5 mouse brain embryo library rather than one that includes E14, the age of their in-utero electroporation mice. Many of the authors' claims focus on in-utero electroporation of shRNA-Abba of E14 mice that are then evaluated at E16-18. Justification for the focus on this age range should be included to support that their findings can then be applied to all mouse corticogenesis.
We thank the reviewer to point this out. Indeed, the data suggest that the interaction between ABBA and Nedd9 occurs before E14. The reason to address the questions at E14 is that in earlier work, we have shown that ABBA is mainly expressed through E10.5-12.5 in the floorplate structure formed by radial glia. The radial glia-specific expression was confirmed through double staining with radial glial (RC2) and neuronal (Tuj1) markers at E12.5 (see Saarikangas et al. J. Cell Sci. 121:1444-1454, 2008). Thus, we consider the Y2H library relevant for identifying ABBA's interactors within radial glia. We have specified this better in the corrected manuscript.
(4) Detail of the effect of the human variant of the ABBA mutation in mice is lacking.
Their identification of the R671W mutation is interesting and the IUE model warrants more characterization, as they did with their original KD experiments.
We have now included addition data in the corrected manuscript showing R671W dependent changes in INM (Supplementary Figure 3 )
Could they show that Abba protein levels are decreased (in either cell lines or electroporated tissue)?
Estimation of ABBA expression in cell expressing ABBA R671W as in Supplemental Figure 5 did not show significant change.
-While time-lapse morphology might not have been performed, more analysis on cell division phenotype (such as plane of division and radial glia morphology) would be helpful.
This would be indeed very informative, but we were not able to perform these analysis in the existing dataset.
Recommendations for the authors:
Reviewer #1 (Recommendations For The Authors):
Here are some suggestions for targeting some of the weaknesses by additional experiments:
Regional Demarcation in Radial Glial Cell Population:
While the authors demonstrate a decrease in overall RFP-positive cells in response to Abba knockdown, the distinction between different regions should be demarcated using cortical layer-specific markers (e.g., CUX1/BRN2 for the upper layer and CTIP2/FOXP2). Quantification based on regional markers would enhance accuracy and meaningful interpretation.
In order to harmonize the quantification during the different developmental stages we have used a broader definition of the cortical regions that may not be entirely fitting with the regions identified with the staining of Cux1 and CTIP2. We have now however included in the supplementary figure 1 with the staining for Cux1 and CTIP2 showing the corresponding regions defined in the manuscript. Supplementary Figure 1.
Mitotic Stage Marker and BrdU Staining:
The discrepancy between no changes in staining with the mitotic stage marker PH3 and a reported decrease in Ki67 staining calls for further clarification. Additionally, the use of BrdU staining could distinguish the effects on dividing cells after Abba knockdown. The authors are encouraged to explore these aspects further, including their applicability to NEDD9 knockdown and Abba mutant overexpression.
As suggested by the reviewer elsewhere, we made use of life imaging. We monitored the distribution of cells showing accumulation of Anillin-GFP in control (Scramble) and ABBA-shRNA3 conditions (this data has been added to the new Supplementary Figure 3). Anillin has been shown to be an efficient tool for monitoring cell cycle stages in vivo (Hesse et al Nature Com. 2012, DOI: 10.1038/ncomms2089). Interestingly, we observed an increase in cells displaying accumulated Anillin in ABBA-shRNA3 treated cells, which is consistent with an arrest of progression of mitosis.
Quantification of Cytokinesis Effects:
The brain slices illustrating the effects of Abba knockdown on cytokinesis would benefit from a quantification depicting changes in interkinetic nuclear migration and the number of successful mitosis events. This would enhance the clarity and interpretation of the observed effects.
In the revised manuscript we have included new data in Supplementary Figure 3 were we report the quantification of the distance of the RGC from the ventricle to address the reviewer’s comments. We were not entirely sure about comment about quantification of successful mitosis events, but as specified above, we have included new data from the monitoring of anillin. We hope to perform more detailed experiments and analysis in future studies.
Loss of Interaction and NEDD9 Localization:
The manuscript lacks an exploration of the loss or decrease in interaction between Abba and NEDD9 in the case of the pathogenic patient-derived mutation in Abba. Addressing this aspect is crucial, as it may shed light on the underlying causes of the observed effects. Furthermore, investigating changes in NEDD9 localization following overexpression of the Abba mutant would provide additional insights.
We fully agree with the reviewer’s comment. Unfortunately the anti NEDD9 antibody had a poor performance in slice immunohistochemistry, which hampered further reliable investigation of expression and distribution changes in vivo. Resolving this issue and providing a more detailed characterization of the mechanism of Abba-NEDD9 interaction will be important in future studies.
Overall, I believe that with minor revisions and additional contextualization, the manuscript has the potential to make a significant contribution to the field. I recommend acceptance pending the incorporation of the suggested revisions.
Reviewer #2 (Recommendations For The Authors):
The manuscript is generally well-organized. We hope that given their nice experimental systems, many of the comments and questions can be addressed with their data already on hand.
Minor Comments
• For Figure 6E A closeup of the vimentin would be helpful - hard to visualize radial glia morphology at the current magnification.
This has been corrected in the new version of the manuscript
• For the in utero electroporation what was their rationale for 2-4 day interval before evaluation? For example, waiting for more cortical plate development to be able to manifest long-term effects.
We observed a massive cell death at E18, in only few of those brains we were able to still observe RFP cells. We have also tried P6 animals but none of them had significant reminding electroporated cells that’s why we have decided to focus at E17, 3 days after the electroporation to have still enough expression of the shRNA.
• Figure 4E-F lacks images of controls for comparison of effect.
This has been corrected in the revised version of the manuscript
