CLASP1 is essential for neonatal lung function and survival in mice
This article has been Reviewed by the following groups
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
- Evaluated articles (Review Commons)
Abstract
The first breath of air at birth marks the beginning of extrauterine life, and breathing problems due to incomplete lung development or acute respiratory distress are common in premature babies and respiratory diseases. However, the underlying molecular mechanisms remain poorly understood. Here we show that the microtubule plus-end-tracking protein CLASP1 is required for neonatal lung function and survival. CLASP1 is expressed in the lungs and associated respiratory structures throughout embryonic development. Clasp1 disruption in mice caused intrauterine growth restriction and neonatal lethality due to acute respiratory failure. Knockout animals showed impaired lung inflation associated with smaller rib cage formation and abnormal diaphragm innervation. Live-cell analysis of microtubule dynamics in cultured hippocampal neurons revealed an increased catastrophe rate, consistent with a role of CLASP1 in neurite outgrowth. Histological and gene expression studies indicated that CLASP1 is required for normal pneumocyte differentiation and fetal lung maturation. Thus, CLASP1-mediated regulation of microtubule dynamics assists multiple systems essential for neonatal lung function and survival.
Article activity feed
-
-
-
-
-
Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
Manuscript number: RC-2022-01573
Corresponding author(s): Helder Maiato and Niels Galjart
1. General Statements
The murine microtubule (MT) plus-end tracking protein CLASP1 has been extensively examined in cultured cells, revealing an important function for this protein in mitosis and the regulation of MT dynamics. Here we describe a major in vivo phenotype of Clasp1 knockout (KO) mice: we find that these mice die at birth due to respiratory problems. In the first version of our manuscript we tried to link this in vivo phenotype of the KO mice to CLASP1’s major roles in cultured cells, including mitosis, and we therefore included multiple results, …
Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
Manuscript number: RC-2022-01573
Corresponding author(s): Helder Maiato and Niels Galjart
1. General Statements
The murine microtubule (MT) plus-end tracking protein CLASP1 has been extensively examined in cultured cells, revealing an important function for this protein in mitosis and the regulation of MT dynamics. Here we describe a major in vivo phenotype of Clasp1 knockout (KO) mice: we find that these mice die at birth due to respiratory problems. In the first version of our manuscript we tried to link this in vivo phenotype of the KO mice to CLASP1’s major roles in cultured cells, including mitosis, and we therefore included multiple results, obtained in cultured cells and in different organs.
We thank the reviewers for their thoughtful and constructive criticisms and for their judgment that our study is - in principle - worthy of publication. Based on suggestions by reviewers #2 and #3 we have decided to focus the revised manuscript on the lung phenotype of the Clasp1 KO mice, and on a possible cause for this phenotype. We believe that our new analysis, which was partly driven by the remarks of the reviewers, is revealing a mechanism for why the mice die at birth. This mechanism suggests a role for CLASP1 in controlling epithelial and endothelial cell differentiation in the neonatal lung, and in particular protein secretion in AT2 alveolar cells.
2. Description of the planned revisions
General remarks
We believe our new RNA-Seq analysis (explained in detail below, in point 3 “Description of incorporated revisions”) strongly suggests that four essential lung cell types (i.e. AT1 and AT2 cells, endothelial cells and immune cells) fail to properly differentiate in Clasp1 KO embryos. In particular AT2 cell differentiation and functioning are hampered in the KO mice.
Brief summary of planned experiments and table of old and new Figures
To support our new findings we will stain sections of wild type and KO lung with a selected set of antibodies and other reagents. To help the reviewers we have made a table with original Figures and Figures for the revision.
Figure Number
Original Figure
Fate of original
Revision Figure
1
Targeted inactivation of the Clasp1 allele
Remains
Targeted inactivation of the Clasp1 allele
2
Clasp1 KO mice show reduced rib-cage and delayed ossification
Minor revision
Clasp1 KO mice show reduced rib-cage and delayed ossification (Statistics will be added)
3
Innervation of the diaphragm is affected in Clasp1 KO mice from E14.5-E18.5
Moved to Supp
Newborn Clasp1 KO lungs show a drastic reduction in air inflation
4
Neurite outgrowth, branching capacity and microtubule dynamics are altered in Clasp1 KO neurons
Removed
Histological and immunological examination of the Clasp1 KO lungs demonstrating decreased air space
5
Histological and immunological examination of the Clasp1 KO lungs demonstrating decreased air space
Moved Up
(4)
Histo-morphological analysis of the developing lung throughout embryonic development (E14.5-PN1)
6
Transcriptome analysis of wild type and Clasp1 KO lungs
Major revision
Transcriptome analysis of wild type and Clasp1 knockout lungs reveals differentiation defects in four major lung cell types (New data added, old data moved to Supp)
7
Loss of Clasp1 alters the ratio of alveolar type I and type II cells in the lungs
Major revision
Cellular analysis of Clasp1 knockout lungs (New data will be added)
8
Role of Clasp1 in AT2 function (New data will be added)
S1
Incidental cell division defects in mouse embryonic fibroblasts derived from Clasp1 knockout mice
Removed
Innervation of the diaphragm is affected in Clasp1 knockout mice from E14.5-E18.5
S2
Ultra-structural analysis of diaphragms
Remains
Ultra-structural analysis of diaphragms
S3
Newborn Clasp1 knockout lungs show a drastic reduction in air inflation
Moved to Main (3)
Cellular analysis of late stage gestation mouse lungs
S4
Histo-morphological analysis of the developing lung throughout embryonic development (E14.5-PN1)
Moved to Main (5)
Exogenous administration of glucocorticoids promotes lung maturation and partially rescues postnatal lethality
S5
Cellular analysis of late stage gestation mouse lungs
Moved Up
(S3)
Analysis of signature genes and cell type signatures of the mouse and human lung
S6
Exogenous administration of glucocorticoids promotes lung maturation and partially rescues postnatal lethality
Moved Up (S4)
Transcriptome analysis of wild type, Clasp1, and Mll3 knockout E18.5 lungs
Below we react to specific comments of the reviewers, describing in more detail which experiments will be carried out and why we will do these experiments.
Specific remarks to the comments of the reviewers
Reviewer #1.
Comment:
p.17: Aqp5 expression was decreased in mutant lungs as shown by RNA-seq data and RT-qPCR. However, immunolabelling with T1a does not show a decrease in the number of Type I pneumocytes (Fig. 7D). According to the data presented, it is difficult to conclude that CLASP1 is involved in Type I pneumocyte differentiation.
A cell count should be done for Figure 7D. Immunolabeling with more markers for Type I pneumocytes, including AQP5 Ab, should be performed to determine if the decreased Aqp5 RNA expression correlates with less Type I cells. GSEA signature has to be confirmed by additional analyses.
Answer:
Given the flat appearance of the T1a-positive cells (see old Figure 7E) it is difficult to carry out a quantification for T1a (which is Pdpn). We will perform new IF experiments to examine AT1/2 cell numbers using additional markers (e.g. Hopx for AT1).
Comment:
p.17: The same comments can be made for Type II pneumocytes and SpC expression.
Answer:
We actually did do an Sftpc (Pro-SPC) count (see old Figure 7E), which reveals that the number of Sftpc-expressing cells is up in the Clasp1 KO. At first sight this seems surprising, given that Chil1 (a top AT2 signature gene at E18.5) is virtually absent from Clasp1 KO lungs. However, our new GSEA analysis (shown in the new Figure 6) shows that of all the E18.5 AT2 signature genes (403 genes in total) the majority is down-regulated, including Chil1 and 4 other top signature genes, but some genes are up, including Sftpc (see new Figure 6). Combined with the fact that we observe more Pro-SPC-expressing cells in the Clasp1 KO lung we hypothesise that AT2 cell numbers are up compared to wild type, giving rise to higher mRNA counts of some genes in the RNA-Seq. Differentiation of AT2 cells is significantly hampered, giving rise to lower expression of many AT2 signature genes in the RNA-Seq. By contrast, all AT1 signature genes are either down or not affected (see new Figure 6). We interpret this as evidence that AT1 cell numbers are down. The same goes for endothelial cells (EC, see new Figure 6). We will perform additional IF experiments to examine this hypothesis.
Reviewer #3.
Comment:
T1α-positive cells should be quantified (Figure 7D). From the images, the number of T1α+ cells in Clasp1 KO is not consistent with the qPCR result showing markedly reduced Aqp5 transcript levels in Clasp1 KO. It is unclear whether the reduction in Aqp5 is due to impaired water channel function as the authors suggest or instead due to reduced number of AT1 cells, further investigation should be conducted.
Answer:
Please see our answer to reviewer #1 above. To summarise, we now have evidence that AT1 cell numbers are down. We will perform additional IF experiments to examine this hypothesis.
Comment:
Additional AT1 markers (Hopx, Ager, Clic5 and Rage) should be assessed by qPCR and immunostaining to determine the effect of Clasp1 knockout on AT1 cells.
Answer:
Please see our answer to reviewer #1 above. To summarise, we will perform new IF experiments to examine AT1/2 cell numbers using additional markers (e.g. Hopx for AT1).
3. Description of the revisions that have already been incorporated in the transferred manuscript
General remarks
As explained in detail below, we believe that our new RNA-Seq analysis has uncovered a mechanism underlying the severe lung phenotype of Clasp1 KO mice, and that it has revealed the major cell types affected in embryonic Clasp1 KO lungs.
Brief summary of experiments
In the first version of the manuscript we used Gene Set Enrichment Analysis (GSEA, see https://www.gsea-msigdb.org/gsea/index.jsp) to compare our RNA-seq results to publicly available scRNA-Seq datasets of cell type signature gene sets, which contain cluster marker genes for cell types identified in single-cell sequencing studies of human tissue. As stated in our manuscript, this revealed “enrichment of alveolar epithelial type I cells and lung capillary intermediate cells in WT lungs ….”. However, the analysis was restricted to what is available in the Gene Set Enrichment Analysis database of the University of San Diego. Thus, we could only compare our embryonic mouse lung data to adult human lung scRNA-Seq data.
We recently discovered publicly available scRNA-Seq datasets of the mouse lung (see https://research.cchmc.org/pbge/lunggens/mainportal.html and https://lungcells.app.vumc.org). The data in these portals are not part of the common GSEA sets of the University of San Diego. In particular the LGEA web portal is very easy to use and data can be downloaded for individual applications. In the new version of our manuscript we compared our RNA-Seq data to scRNA-Seq data of the embryonic mouse lung, focussing on E18.5. We first overlaid differentially expressed genes in Clasp1 KO lungs with LGEA E18.5 scRNA-seq gene signatures for different cell types, and we subsequently compared all the genes in our dataset with the gene signature lists, using custom-built gene signature sets and the GSEA software. In addition, we interrogated LGEA to find out which signature genes are specifically turned on from E16.5-E18.5 in the different cell types in the developing mouse lung. We found, for example, that Chil1, which is the most severely down-regulated gene in our Clasp1 KO RNA-Seq, is a very prominent AT2 signature gene; Chil1 is hardly expressed at E16.5 and prominently comes up at E18.5.
Our combined analysis strongly suggests that four cell types (AT1, AT2, endothelial cells (EC), and immune cells (IC)) are affected in their differentiation in the Clasp1 KO lung, and that this defect occurs in the later stages of lung development (from E16.5 onward). As the top five differentially down-regulated genes in KO lungs (including Chil1) are all top signature genes of AT2 cells, these data strongly suggest that it is this cell type that is most affected in the KO. A Metascape analysis (which includes a GO enrichment analysis, see also our specific answer to comments of reviewer #3 below) is consistent with the scRNA-Seq comparison and suggests, among others, that the secretory pathway might be hampered in the Clasp1 KO. This analysis furthermore indicates that cholesterol metabolism might be affected in the Clasp1 KO, which bears relevance to our dexamethasone rescue experiments.
Specific remarks to the comments of the reviewers
Reviewer #1.
Comment:
p.6: What is the justification to mention Nfib, Pdpn and Ndst1 mutant mice in the introduction? Do these genes have any cellular/molecular/functional relation with CLASP1?
Answer:
We initially wanted to provide examples of genes important for lung maturation, whose absence in knockout mice leads to lung collapse. Of the examples provided Pdpn (which is equal to the marker T1a) bears a relation with our data in that it is down-regulated in Clasp1 KO lungs (see Table S2, RNA-Seq); furthermore, we examined T1a localisation in IF stainings (see old Figure 7E). In the new version of the manuscript we modified this Introduction section, to better align with our recent results, and to introduce the papers mentioned by reviewer #3 (Nelson et al., 2017; doi:10.1242/dev.154823, Li, J. et al., Dev Cell, 44, 297-312 e5.), who points out that pressure plays an important role in lung development. In the Li et al manuscript Pdpn is mentioned as being expressed at E16.5 in so-called Id2+ cells, together with Sftpc. These cells are proposed to be the precursors of the AT1/2 epithelial cells that arise later.
Comment:
p.8: It is mentioned that CLASP1 is expressed in secretory cells of the lung. Which ones? Is CLASP1 expressed in nerves, muscle cells and/or fibroblasts of the diaphragm? These information are important according to the phenotypes described.
Co-immunolabelling experiments should be done.
Answer:
We apologize for our incorrect phrasing. With respect to the lung, we now state that “CLASP1 is expressed in the endothelium of blood vessels, as well as in all cells lining the airways of mouse lungs at E18.5 (Fig. 1A)”.
Comment:
p.11: To identify the cause of the respiratory failure, the authors looked at the innervation pattern of the phrenic nerve in the diaphragm. Mutants present decreased branching but larger nerve extensions covering a wider innervated area and less neuromuscular junctions. Despite the decreased innervation of the diaphragm, its morphology is normal as well as the ultra-structure of the sarcomeres suggesting a mild phenotype rather than the cause of death of the mutants as suggested by the authors (p.20).
Diaphragmatic muscle activity should be measured to establish if the contractile activity of the diaphragm is affected. This might support the statement of the authors.
Answer:
We thank the reviewer for these observations. We agree with the reviewer and have toned down our conclusions in this section. We now simply describe the innervation pattern because we believe it is interesting, and we tentatively conclude that it may contribute to the severe respiratory phenotype which is primarily due to impaired AT1/2, EC, and IC differentiation.
Comment:
p.13: The authors examined lung from mutants. Mutant lungs do not float and they are collapsed at birth. However, lung morphology appears normal and myofibroblasts, ciliated cells and Club cells are present as shown by IHC labeling. No difference in proliferation and apoptosis was reported.
It would have been more informative to do BrdU/EdU immunolabeling for proliferation in order to see if differences occur in specific cell types of the lung. It is not clear why the authors have limited their IHC analysis to these three specific cell types. A complete analysis should be done.
Answer:
As described above (general remarks), we compared our RNA-Seq data to publicly available scRNA-Seq data from the developing mouse lung (see new Figure 6). These comparisons reveal which cell types are affected in the Clasp1 KO lung (AT1/2, EC, IC), and which process might be hampered.
Comment:
p.14: The authors proposed a delay in lung development according to lung morphology that appears more collapsed starting at E15.5.
Measurement of branching would allow to quantify this delay. Since cell differentiation occurs ~E16.5, analysis of the onset of cell types can also support a delay in lung development.
Answer:
As described above (general remarks), we compared our RNA-Seq data to publicly available scRNA-Seq data from the developing mouse lung (see new Figure 6). This not only revealed which cell types are affected in the Clasp1 KO lung, but also suggest that a differentiation block occurs at E16.5 to E18.5. For example, Chil1, a top AT2 signature gene of E18.5, is hardly expressed at E16.5 and is strongly upregulated at E18.5. This gene fails to become up-regulated in the Clasp1 KO, indicating that epithelial precursor cells have problems differentiating to AT2 type cells. By contrast, Id2, a marker of precursor epithelial cells, is normally expressed in the Clasp1 KO, and two genes that are co-expressed with Id2 in these precursor cells (Pdpn and Sftpc) are slightly down and up, respectively, in the Clasp1 KO. Thus, while our lung morphology studies might suggest early defects, our RNA-Seq indicates that specific defects occur during the late terminal saccular stage, i.e. from E16.5 onward. We therefore agree with with Negretti et al (2021, doi: 10.1242/dev.199512, Discussion section) who state: the developmental stages of the lung are largely founded on histologically descriptive features. While this is important, such a categorization often results in debate regarding the function and identity of cell types within the boundaries of each stage. By contrast transcriptome analysis suggests that different cell types commit to change asynchronously during development, suggesting that the timing of the saccular-to-alveolar transition is fluid and highly cell-type specific.
As shown by Li et al (2018, doi.org/10.1016/j.devcel.2018.01.008) mechanical forces contribute to embryonic lung alveolar epithelial cell differentiation. Interestingly, RNA-Seq data from Nelson et al (2017; doi:10.1242/dev.154823) suggest that CLASP1 is a “pressure sensing gene” (see also below, our answer to comments of reviewer #3). Thus, Clasp1 KO lungs might fail to properly sense pressure, which could explain, at least in part, the observed failure in epithelial differentiation.
Comment:
p.15: Finally, the authors conclude this section by "these data support a direct role for CLASP1 in lung maturation".
Which direct role? How? This sentence appears premature according to the data presented. The authors should look at microtubule dynamics in lung cells from mutant embryos to see if a link exists between the proposed role of the protein and the lung phenotype observed.
Answer:
The reviewer is correct, knockout studies can not demonstrate a direct role of a protein in a perturbed process. We have therefore removed the word “direct” from this phrase.
Comment:
p.15: The authors attempted to rescue the defective lung maturation phenotype by treating pregnant females with dexamethasone at late gestational stages. Around 10% of mutants survive for more than 45 minutes to 2 hrs compared to 20-30 minutes for mutants obtained from untreated mothers (p.9). Even though it is an intriguing result, the very small numbers of "survivors" makes very difficult to reach a conclusion.
This section should be shortened.
Answer:
Our new Metascape analysis, which will be presented in the new Figure 8, suggests that cholesterol metabolism is affected in the Clasp1 KO mice. Cholesterol is an important component of mammalian cell membranes, of both alveolar and lamellar body surfactant, and it is a precursor of vitamin D and steroid hormones. A cholesterol defect would explain the partial rescue by dexamethasone in the Clasp1 KO, i.e. dexamethasone can rescue a steroid hormone defect but it cannot rescue other defects (e.g. surfactant production). Given these new results we decided to leave the section on glucocorticoids as it is and come back to it when we discuss the Metascape result in the revised manuscript.
Comment:
p.16: To determine which molecular mechanisms are responsible for the lung defect, the authors performed RNA-seq analysis on E18.5 lung specimens. The number of genes with significant differential expression was low and the highest scores were cathepsin E for the upregulated gene and chitinase-like 1 for the downregulated gene.
Are these two genes known for their role in lung development? Please describe.
Answer:
The Ctse gene, which encodes Cathepsin E, is indeed the most upregulated gene in the Clasp1 KO. Although it is up-regulated in all three KO mice, Ctse expression is quite low (normalised counts: ~2 in KO, up from ~0.2 in WT). Based on the comment of this reviewer we examined Ctse expression in the scRNA-Seq lung repositories, but we could not find any description, presumably because its expression is too low (scRNA-Seq has difficulty catching low abundance genes), consistent with our data. Furthermore, there is not much literature on the role of Cathepsin E in the lung. We therefore decided to remove any mention of Ctse in the manuscript. By contrast, the expression and function of Chil1 are described in detail.
Comment:
p.16: Except for the fact that Chil1 is also downregulated in mutant lungs for the H3K4 methyltransferase Mll3 gene, it is not clear why the authors compared these 2 sets of data.
Can CLASP1 and MLL3 interact together? How? Did the authors looked at the list of genes that are commonly differentially expressed? Does it provide some clues on the mechanisms? The RNA-seq data should be analyzed more deeply.
Answer:
The reviewer is correct, we compared the Mll3 (i.e. Kmt2c) RNA-Seq dataset because Chil1 is down-regulated in the Mll3 KO lung at E18.5, like in the Clasp1 KO. To examine a possible relation between Mll3 and Clasp1 in more detail, we overlaid the differentially expressed genes from the Mll3 dataset with the custom-built gene signature dataset of E18.5 lung (described above). The data suggest that Mll3 knockout affects AT1 differentiation (see new Supplementary Figure S6C). This mode of action is clearly different from that of CLASP1, and since Mll3 is nuclear and CLASP1 is cytoplasmic we do not believe these proteins interact. Given our new and exciting data on the Clasp1 KO lung phenotype, we moved the Mll3 data to the new Supplementary Figure 6, and only briefly we touch upon these data in the manuscript.
Comment:
p.16: There is also a Clasp2 gene with a more restricted expression pattern. Clasp2 mutant mice either die from hemorrhages or survive. It is not clear why the RNA-seq data of the lungs from Clasp2-/- mice are presented since no lung phenotype is mentioned for these mice. How the lack of change in Chil1 expression in Clasp2 mutant lungs is informative?
This should be clarified or the data should be removed.
Answer:
The reviewer is correct, i.e. in light of our new findings (Chil1 is a top signature gene of E18.5 AT2 cells) it makes little sense to include the Clasp2 KO RNA-Seq data, as these were generated in adult mouse lungs. We therefore removed these data from the manuscript.
Comment:
p.31: The authors mentioned a role for CLASP1 in the mesenchyme.
What are the experiments and data that support this sentence?
Answer:
We thank the reviewer for this remark, we have no evidence for a role of CLASP1 in the mesenchyme and have removed this phrase.
Comment:
How do the authors reconcile their observation of CLASP1 expression in lung secretory cells (p.8) with their conclusion of defective Type I cell differentiation (p.17)?
Answer:
We apologize for our incorrect phrasing. With respect to the lung, we now state that “CLASP1 is expressed in the endothelium of blood vessels, as well as in all cells lining the airways of mouse lungs at E18.5 (Fig. 1A)”.
Reviewer #2.
Comment:
Fig. 3. There is not a lot of detail how the analysis in B-E was done, and no primary data for the synaptic defects.
Answer:
We have removed these data from the manuscript.
Reviewer #3.
Comment:
- The authors showed significant reduction in the rib cage size and abnormal diaphragm innervation in Clasp1 KO. Mechanical properties play a crucial role in regulating lung development and maturation. So changes in intrathoracic space and pressure are a major limiting factor that impairs lung development and maturation (Nelson et al., 2017; doi:10.1242/dev.154823, Li, J. et al., Dev Cell, 44, 297-312 e5.). Answer:
We thank the reviewer for these interesting papers and observations.
Nelson et al (2017; doi:10.1242/dev.154823) devised a method to culture lung-on-a-chip where they can induce pressure in culture. They apply this to examine lung development and they also do RNA-Seq. Interestingly, they find that Clasp1 is down-regulated at high pressure compared to low pressure (log2FC 0.5, Clasp1 goes down ~1.5 fold in high pressure). Thus Clasp1 appears to be a “pressure-responsive gene”. However, Nelson et al examine gene expression at much earlier time points than we do (E12-14 versus E18.5). In our view it therefore makes little sense to compare RNA-Seq data.
Li et al (2018 doi.org/10.1016/j.devcel.2018.01.008) show that mechanical forces help to control embryonic lung alveolar epithelial cell differentiation. More specifically, mechanical force from amniotic fluid inhalation ensures AT1 cell differentiation, whereas FGF10-mediated ERK1/2 signaling induces a protrusive structure in some cells that protects from mechanical force-caused flattening to specify AT2 fate. They conclude that future AT2 cells can “embed” into mesenchyme by exerting an acto-myosin based force and hence they can keep their cuboidal shape. The differentiation of the two cell types occurs at different time points, E16.5 for AT2, and E17.5 for AT1. In this manuscript they also mention that Id2+ tip cells express pro-SPC and Pdpn (which are up and down, respectively, in Clasp1 KO). These Id2+ cells would be the AT1/2 progenitors.
We believe that a smaller ribcage in the Clasp1 KO does not necessarily have to be a cause of increased pressure on the lung, if the lung is also smaller. Nonetheless, since CLASP1 is a “pressure-responsive gene”, Clasp1 KO lungs might experience aberrant pressure sensing (in addition to a possible pressure difference due to a smaller ribcage). This different sensing predicts altered differentiation pathways, which is exactly what we see. We have modified the revised version of the manuscript to reflect these thoughts and observations.
Comment:
Since CLASP1 was found to be highly expressed in the lung endothelium (Figure 1A), this suggests the importance of CLASP1 in the lung vasculature. GSEA analysis also showed significant downregulation of genes from the lung capillary intermediate 1 cell signature gene set in Clasp1 KO (Figure 7G). Extensive crosstalk between the lung endothelium and other lung cell types is critical for the regulation of lung development. However, no further investigation was carried out to elucidate this.
Answer:
We have performed a new comparison, which is extensively discussed above and shows that EC are affected in the Clasp1 KO lungs, as predicted by this reviewer. We will discuss crosstalk between cell types in the new version of the manuscript.
Comment:
Analysis of RNA-Seq data needs to be re-written. Pathway or GO enrichment was not performed. Although the authors have identified a number of key DEGs, only Chil1 was investigated. It is also unclear how it led the authors to identify Mll3 KO experiment on the Omnibus repository. A list of overlapped genes between Mll3 KO dataset and Clasp1 KO dataset were not provided. Aqp5 (AT1 marker gene) that authors claimed to be significantly reduced in Clasp1 KO is not on the DEGs list (Table S2).
Answer:
We initially focused on Chil1 because its expression is almost completely abrogated in all three Clasp1 KO lungs. The identification of the Mll3 dataset was coincidental; we mentioned it because Chil1 is also affected in these KO mice. A Venn diagram of overlapping significantly deregulated genes in both datasets is shown in the new Figure S6 of the revised manuscript. However, this analysis has been superseded by the new comparison with scRNA-Seq data from the LGEA web portal. As extensively explained above this new analysis provides a satisfying explanation for the lack of Chil1 in Clasp1 KO lungs. We also performed a Metascape analysis (which includes pathway and GO enrichment analyses), which will be included in the revised version of this manuscript. Finally, the reviewer is correct that Aqp5 is not in the DEGs list, this is because the adjusted p-value did not reach the required significance. We nevertheless showed its RNA-Seq values, first because the p-value is significant, second, because RT-PCR experiments confirm it to be down-regulated, and third, because Aqp1 (another AT1 marker) is also deregulated (with an adjusted p-value that is significant). In the revised manuscript we will examine Aqp5 levels by IF staining.
Comment:
There is a lack of cohesion between the experimental findings presented in the paper and the RNA Seq data analysis. Pathway or GO enrichment was not performed for the DEGs the authors identified. This would help identify the key functions of the deregulated genes in Clasp1 KOs and provide a fuller picture of what pathways/biological processes are dysregulated in the absence CLASP1. Instead, the authors have focused on one single gene, Chil1 in the subsequent analysis. The authors infer that overlapped DEGs between Mll3 KO and Clasp1 KO mean that same cell types or signalling pathways are affected in embryonic lungs of Mll3 and Clasp1 KO, this is an overinterpretation. A list showing the overlap in DEGs between Mll3 KO dataset and Clasp1 KO dataset should be provided.
Answer:
We have improved our RNA-Seq analysis and we have performed a Metascape analysis, which includes pathway and GO enrichment analyses. Results are shown in the new Figures 6 and 8. The Metascape analysis indicates which pathways/biological processes are deregulated in the absence CLASP1. We observe, for example, defects in endocytosis, and cholesterol metabolism. Given the new data, we decided to pay less attention to the Mll3-CLASP1 comparison.
Minor comments:
Figure 1A - please label the specific cell types to aid visualisation.
Figure 6B - present the Log2FC for KO vs WT instead of WT vs KO to facilitate data visualisation and interpretation
Figure 6E - provide the overlapping genes in a list and include it as a supplementary table
Figure 7D and 7F - Quantification is needed
The statistical tests used should be added to the figure legends.
There is some wording in the manuscript that is either unclear or inaccurate, please carefully check the manuscript. e.g. manuscript refers to alveolization- I would recommend changing this to the more widely used terms alveolarization or alveologenesis. The manuscript refers to 'catastrophe rate'- this term needs to be defined. Answers:
This has been done.
This has been done.
This panel has been moved to a Supplementary Figure, as the analysis is less relevant now we will not provide the list.
This will be done.
This has been/will be done.
This has been done. The term “catastrophe rate” has been removed.
4. Description of analyses that authors prefer not to carry out
General remarks
Based on the comments of reviewers #2 and #3 we have decided to fully focus our revised manuscript on the lung phenotype of the Clasp1 KO mice. We still do show the results on the ribcage (Figure 2) and diaphragm (Figure S2) because they might enhance the severity of the lung phenotype. We have decided not to carry out extra “non-lung” experiments.
Specific remarks to the comments of the reviewers
Reviewer #1.
Comment
p.10: Homozygous mutants are smaller. The authors reported minor skeletal phenotypes small rib cage and delayed ossification in sternum and occipital bone.
The number of specimens analyzed was not mentioned rendering difficult to establish if these observations are important or not. Stats should be included.
Answer:
Whereas the results of Figure 1H, I (growth deficits at E15.5 and PN1) are based on analysis of multiple animals, the embryonic skeleton data presented in Figure 2 are based on single mouse comparisons, i.e. one WT and one KO. Given the obvious growth deficit in the KO (Figure 1H, I) and the fact that gross morphological observation did not reveal a specific body part in the KO mice that is affected (Figure 1G), we were of the opinion that a representative comparison of the skeleton is allowed and we therefore kept Figure 2 intact. Since we focus in the revision on the lung phenotype, we have decided against examining the skeletons of more mice. We are willing to remove Figure 2, or make it Supplemental, if the reviewer feels that the skeletal phenotype is too prominently displayed.
Comment:
p.10: The authors established MEF used to study cell division. Multipolar spindles and additional centrosomes were detected in mutant cells.
No stats were provided to establish if the differences in numbers are significant. According to the authors, the cell division defects may explain the smaller size of mutants. The authors should check proliferation in MEF. The sentence of conclusion is not well supported according to the data presented.
Answer:
Based on the advice of reviewer #2, who states “I think it would be best to better focus the paper on the lung phenotype”, we have decided to remove the mitotic data on MEFs.
Comment:
p.12: The authors looked at the growth capacity of motor neurons and dorsal root ganglion neurons and showed a reduced growth in both cases.
How do the authors reconcile the observation made in the diaphragm in which nerve extensions are larger with the reduced growth capacity of neurons?
Answer:
We thank the reviewer for this remark, which is difficult to address, as CLASPs are expressed at different levels in neurons and as different isoforms, which may even have antagonistic functions. For example, in our recent publication (Sayas et al, 2019, DOI: 10.3389/fncel.2019.00005) we find through RNA-Seq that in cultured hippocampal neurons (3DIV) Clasp2β/γ levels are increased compared to Clasp2α-mRNA and that both in hippocampal and in DRG neurons Clasp2 mRNA levels are higher than Clasp1. As CLASP2b/g have a different function compared to CLASP2a, it is conceivable that absence of CLASP1 leads to different effects due to different CLASP2 activities. However, we recognize that these are speculations. Because of this and because reviewer #2 advices against inserting the neuronal data, we have decided to completely remove these results from the manuscript.
Comment:
p.12: The authors used cultured hippocampal neurons for imaging microtubule growth. According to the authors, the loss of CLASP1 deregulates microtubule dynamics.
No explanation was provided to justify the use of hippocampal neurons. What is a catastrophe rate? What is the justification to study this parameter? What does it tell us about microtubule dynamics?
Answer:
Although we have decided to remove the neuronal data from the revised manuscript, we would like to address this comment nonetheles. Hippocampal neurons are often used in the field, hence they represent a “golden standard”. Furthermore, the techniques to examine microtubule dynamics are well established in this system. Dynamic microtubule behaviour is described using five parameters: growth rate of microtubules, shrinkage rate of microtubules, catastrophe and rescue frequencies (the conversion of growth to shrinkage or from shrinkage to growth, respectively), and pauzing times. The marker used in our studies (EB3-GFP) accumulates at the ends of growing microtubules, allowing us to measure growth rate and the duration of a growth event. The latter is the inverse of the catastrophe frequency. Hence, using EB3-GFP we are able to examine two of the five parameters. Although this is not complete the parameters do allow us to draw (speculative) conclusions. For example, a higher growth rate indicates that free tubulin concentration is higher, as tubulin concentration is a main determinant of growth rate. This in turn means that there are less microtubules (tubulin must come from somewhere). If this correlates with the catastrophe frequency (which should be higher) than one can conclude that CLASP1 is a microtubule-stabilising protein.
Reviewer #2.
Comment:
Fig. S1. It would be good to indicate the number of cells / experiments analyzed. In panel D, there is only one multi-nucleated cell, which without further analysis does not mean much. The authors correlate this mitotic defect with smaller animal size although this connection is not at all conclusive. If both CLASPs are important for mitosis, do CLASP2 KOs have similar size defects? It is also mentioned above that CLASP1 KOs show microcephaly. Are there fewer neurons that might also be linked to a stem cell division defect? I understand that this is not the central point of the paper and important to include given previous work on CLASPs, but it would be good to discuss a little clearer. It seems the authors do not think this is the/a cause of the lung phenotype, but can that be completely excluded?
Answer:
Based upon suggestions of this reviewer (for example: “I think it would be best to better focus the paper on the lung phenotype”) we will not address this comment beyond a statement that Clasp2 knockout mice are indeed also smaller.
Fig. 4. Please indicate n of cells / experiments and statistics in the figure legend. In panel B and C, it would help to include the time on the figure itself and to scale the y-axis the same to better illustrate differences. It is very hard to see much in panel D. The quantifications in E and F do not make sense. How can the total neurite length (average of many neurons?) be larger than the longest neurite length?
The switch to MT dynamics in Fig. 4 is very abrupt and the relevance is unclear. Where were these kymographs located in the neuron (growth cones or neurites)? Primary data needs to shown here. The changes in catastrophe frequency are not that large and I doubt this can be accurately measured from kymographs as shown. Yes, MTs are important in neurite growth, but the potential link here is very vague. Are similar changes in MT dynamics also seen in the MEFs?
Minor:
Answer:
See above, we will not address these comments, since we will remove these data.
Reviewer #3.
Comment:
The lung morphological difference and disrupted lung cell differentiation in Clasp1 KO could be secondary to the biomechanical defects. This is crucially important but is not addressed in this study, ex vivo lung culture may help to answer this question.
Answer:
While the experiments suggested by this reviewer are interesting, we do not have sufficient expertise (nor the equipment) to carry out such specialised experiments.
Comment:
CLASPs are known to regulate directed cell migration (Myer and Myers 2017, doi: 10.1242/bio.028571) and this is a key process required for lung morphogenesis. Experiments to address whether directed cell migration is affected should be conducted in Clasp1 KO mice.
Answer:
We agree that migration assays would be interesting to perform. However, again, we do not have the expertise to do such assays in the developing lung. Experiments in MEFs are possible, and indeed, we previously showed a role for CLASP2 in directed cel migration in MEFs (DOI: 10.1016/j.cub.2006.09.065). However, lung epithelial cells are different from MEFs, and we have shown that CLASPs have cell type- (and isoform-)specific functions. Reviewer #2 actually advised us to focus on the lung phenotype.
Comment:
Higher magnification images of staining for microtubule associated proteins in neurons is required to show the details of the defects.
Answer:
Based on the reviewers’ advice we decided to take out the neuronal data and focus the manuscript on the lung phenotype.
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #3
Evidence, reproducibility and clarity
In the manuscript, the authors used a Clasp1 KO mouse to investigate the roles of CLASP1, a microtubule-associated protein. The manuscript presents a number of phenotypes found in the homozygous knockouts including reduced intrauterine growth, altered respiratory muscle innervation, and perturbed lung maturation. The loss of CLASP1 leads to neonatal lethality due to breathing defects however, whilst the manuscript shows a number of different phenotypes in the mutants, The underlying mechanism of how disrupted CLASP1-mediated microtubule dynamics causes the phenotypes observed is not clear. The authors propose some mechanisms for 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
In the manuscript, the authors used a Clasp1 KO mouse to investigate the roles of CLASP1, a microtubule-associated protein. The manuscript presents a number of phenotypes found in the homozygous knockouts including reduced intrauterine growth, altered respiratory muscle innervation, and perturbed lung maturation. The loss of CLASP1 leads to neonatal lethality due to breathing defects however, whilst the manuscript shows a number of different phenotypes in the mutants, The underlying mechanism of how disrupted CLASP1-mediated microtubule dynamics causes the phenotypes observed is not clear. The authors propose some mechanisms for the phenotypes but these are largely speculative and additional experiments are required to substantiate them.
Some major and minor comments are detailed below which we hope will be useful.
Major comments:
- The authors showed significant reduction in the rib cage size and abnormal diaphragm innervation in Clasp1 KO. Mechanical properties play a crucial role in regulating lung development and maturation. So changes in intrathoracic space and pressure are a major limiting factor that impairs lung development and maturation (Nelson et al., 2017; doi:10.1242/dev.154823, Li, J. et al., Dev Cell, 44, 297-312 e5.).
- The lung morphological difference and disrupted lung cell differentiation in Clasp1 KO could be secondary to the biomechanical defects. This is crucially important but is not addressed in this study, ex vivo lung culture may help to answer this question.
- CLASPs are known to regulate directed cell migration (Myer and Myers 2017, doi: 10.1242/bio.028571) and this is a key process required for lung morphogenesis. Experiments to address whether directed cell migration is affected should be conducted in Clasp1 KO mice.
- Since CLASP1 was found to be highly expressed in the lung endothelium (Figure 1A), this suggests the importance of CLASP1 in the lung vasculature. GSEA analysis also showed significant downregulation of genes from the lung capillary intermediate 1 cell signature gene set in Clasp1 KO (Figure 7G). Extensive crosstalk between the lung endothelium and other lung cell types is critical for the regulation of lung development. However, no further investigation was carried out to elucidate this.
- Higher magnification images of staining for microtubule associated proteins in neurons is required to show the details of the defects.
- Analysis of RNA-Seq data needs to be re-written. Pathway or GO enrichment was not performed. Although the authors have identified a number of key DEGs, only Chil1 was investigated. It is also unclear how it led the authors to identify Mll3 KO experiment on the Omnibus repository. A list of overlapped genes between Mll3 KO dataset and Clasp1 KO dataset were not provided. Aqp5 (AT1 marker gene) that authors claimed to be significantly reduced in Clasp1 KO is not on the DEGs list (Table S2).
- T1α-positive cells should be quantified (Figure 7D). From the images, the number of T1α+ cells in Clasp1 KO is not consistent with the qPCR result showing markedly reduced Aqp5 transcript levels in Clasp1 KO. It is unclear whether the reduction in Aqp5 is due to impaired water channel function as the authors suggest or instead due to reduced number of AT1 cells, further investigation should be conducted.
- Additional AT1 markers (Hopx, Ager, Clic5 and Rage) should be assessed by qPCR and immunostaining to determine the effect of Clasp1 knockout on AT1 cells.
- There is a lack of cohesion between the experimental findings presented in the paper and the RNA Seq data analysis. Pathway or GO enrichment was not performed for the DEGs the authors identified. This would help identify the key functions of the deregulated genes in Clasp1 KOs and provide a fuller picture of what pathways/biological processes are dysregulated in the absence CLASP1. Instead, the authors have focused on one single gene, Chil1 in the subsequent analysis. The authors infer that overlapped DEGs between Mll3 KO and Clasp1 KO mean that same cell types or signalling pathways are affected in embryonic lungs of Mll3 and Clasp1 KO, this is an overinterpretation. A list showing the overlap in DEGs between Mll3 KO dataset and Clasp1 KO dataset should be provided.
Minor comments:
- Figure 1A - please label the specific cell types to aid visualisation.
- Figure 6B - present the Log2FC for KO vs WT instead of WT vs KO to facilitate data visualisation and interpretation
- Figure 6E - provide the overlapping genes in a list and include it as a supplementary table
- Figure 7D and 7F - Quantification is needed
- The statistical tests used should be added to the figure legends.
- There is some wording in the manuscript that is either unclear or inaccurate, please carefully check the manuscript. e.g. manuscript refers to alveolization- I would recommend changing this to the more widely used terms alveolarization or alveologenesis. The manuscript refers to 'catastrophe rate'- this term needs to be defined.
Significance
The authors have carefully documented a variety of phenotypes that occur in Clasp1 knockout mice, this is novel because this is the first report of genetic manipulation of Clasp1 in an animal model. It is clear that the homozygotes die because they cannot breathe properly once they transition to air breathing at birth. However, it is not clear in the current manuscript what the underlying reasons for the breathing defects are. The manuscript shows a number of respiration- related deficiencies including small rib cage, disrupted diaphragm innervation and lack of alveolar maturation but the manuscript documents a series of phenotypes rather than pulling together a hypothesis about the role of Clasp1 in the respiratory system.
Foetal breathing movements are essential for normal lung development and the maturation of cells into their differentiated phenotypes e.g. ATII to ATI cells in the alveoli. It could be that the reduced thoracic space, coupled with the diaphragm deficiencies are the underlying cause of the alveolar cell maturation defects and failure of normal breathing, due to impaired biomechanics. The authors could conduct further experiments to explore this avenue e.g. ex vivo lung culture to see if there are still developmental deficiencies in the absence of reduced intrathoracic space (small rib cage).
The manuscript details some interesting findings but in its current form, it lacks a coherent story. I am not convinced that all the details of the effects on DRG and motor neurons is required in the same manuscript as the analysis of lung biology. It may be clearer to split the findings into separate manuscripts.
-
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
The manuscript by Pereira et al. investigates the phenotype of loss of the MT-associated plus-end-bidding protein CLASP1 in mouse. They find that CLASP1 KO pups are not viable due to rapid respiratory failure and the paper presents an in-depth analysis of the lung phenotype that is quite striking. However, the mechanistic links to previously proposed cellular functions of CLASP1 in mitosis and MT dynamics are weak and confusing.
For example, the analysis of mitotic defects in MEFs or of MT dynamics in neurons is not convincing and does not really explain anything; if or if not these cellular phenotypes are related to the observed …
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
The manuscript by Pereira et al. investigates the phenotype of loss of the MT-associated plus-end-bidding protein CLASP1 in mouse. They find that CLASP1 KO pups are not viable due to rapid respiratory failure and the paper presents an in-depth analysis of the lung phenotype that is quite striking. However, the mechanistic links to previously proposed cellular functions of CLASP1 in mitosis and MT dynamics are weak and confusing.
For example, the analysis of mitotic defects in MEFs or of MT dynamics in neurons is not convincing and does not really explain anything; if or if not these cellular phenotypes are related to the observed lung defect. (in fact, the discussion of the paper does not even mention again these functions of CLASP1). So, in the end, the reader is left with a menu of choice of whether CLASP1 is directly involved in lung development, required for innervation or for something else. Many other questions are left unanswered: Is innervation defective because lung development is abnormal, or does innervation control development (as it has been proposed in other organs)? How is CLASP1 controlling the lung transcriptome; does this have anything to do with its cellular functions or is this a completely indirect effect, again stemming from a deeper developmental defect? Overall, I think the lung phenotype is interesting and worth publishing. However, I do not exactly know how to resolve the mechanistic questions, but I think it would be best to better focus the paper on the lung phenotype and maybe rearrange the order data are presented (Fig. 4 seems oddly plopped in the middle of the lung analysis).
Specific comments:
Fig. S1. It would be good to indicate the number of cells / experiments analyzed. In panel D, there is only one multi-nucleated cell, which without further analysis does not mean much. The authors correlate this mitotic defect with smaller animal size although this connection is not at all conclusive. If both CLASPs are important for mitosis, do CLASP2 KOs have similar size defects? It is also mentioned above that CLASP1 KOs show microcephaly. Are there fewer neurons that might also be linked to a stem cell division defect? I understand that this is not the central point of the paper and important to include given previous work on CLASPs, but it would be good to discuss a little clearer. It seems the authors do not think this is the/a cause of the lung phenotype, but can that be completely excluded?
Fig. 3. There is not a lot of detail how the analysis in B-E was done, and no primary data for the synaptic defects.
Fig. 4. Please indicate n of cells / experiments and statistics in the figure legend. In panel B and C, it would help to include the time on the figure itself and to scale the y-axis the same to better illustrate differences. It is very hard to see much in panel D. The quantifications in E and F do not make sense. How can the total neurite length (average of many neurons?) be larger than the longest neurite length? The switch to MT dynamics in Fig. 4 is very abrupt and the relevance is unclear. Where were these kymographs located in the neuron (growth cones or neurites)? Primary data needs to shown here . The changes in catastrophe frequency are not that large and I doubt this can be accurately measured from kymographs as shown. Yes, MTs are important in neurite growth, but the potential link here is very vague. Are similar changes in MT dynamics also seen in the MEFs?
Minor:
Fig. 1A please indicate in legend what is CLASP staining (suppose the brown stuff).
Define HT in text.
Again, please include statistics in figure legends (and indicate n and p values)
Significance
Findings presented in regard to CLASP1 role in lung development are interesting and significant, also as a potential novel model system of newborn respiratory failure. The mechanistic link to known functions of CLASP1 however remains vague and would need substantial additional work to address properly.
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #1
Evidence, reproducibility and clarity
The manuscript entitled "CLASP1 is essential for neonatal lung function and survival in mice" by Pereira et al. reports the characterization of the phenotype of a Clasp1 null mutant mouse line. CLASP1 is a microtubule plus-end tracking protein involved in the regulation of microtubule dynamics broadly expressed in the organism. All Clasp1-/- mice die at birth from respiratory failure.
General comment:
This is an interesting study about the characterization of a Clasp1 mutant mouse line. The manuscript is clear and well-written. The analysis is descriptive. Many aspects are studied but unfortunately they are covered superficially. …
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #1
Evidence, reproducibility and clarity
The manuscript entitled "CLASP1 is essential for neonatal lung function and survival in mice" by Pereira et al. reports the characterization of the phenotype of a Clasp1 null mutant mouse line. CLASP1 is a microtubule plus-end tracking protein involved in the regulation of microtubule dynamics broadly expressed in the organism. All Clasp1-/- mice die at birth from respiratory failure.
General comment:
This is an interesting study about the characterization of a Clasp1 mutant mouse line. The manuscript is clear and well-written. The analysis is descriptive. Many aspects are studied but unfortunately they are covered superficially. However, they open the way to more deepened analyses.
Specific comments:
p.6: What is the justification to mention Nfib, Pdpn and Ndst1 mutant mice in the introduction? Do these genes have any cellular/molecular/functional relation with CLASP1?
p.8: It is mentioned that CLASP1 is expressed in secretory cells of the lung. Which ones? Is CLASP1 expressed in nerves, muscle cells and/or fibroblasts of the diaphragm? These information are important according to the phenotypes described.
- Co-immunolabelling experiments should be done.
p.10: Homozygous mutants are smaller. The authors reported minor skeletal phenotypes small rib cage and delayed ossification in sternum and occipital bone.
- The number of specimens analyzed was not mentioned rendering difficult to establish if these observations are important or not. Stats should be included.
p.10: The authors established MEF used to study cell division. Multipolar spindles and additional centrosomes were detected in mutant cells.
- No stats were provided to establish if the differences in numbers are significant. According to the authors, the cell division defects may explain the smaller size of mutants. The authors should check proliferation in MEF. The sentence of conclusion is not well supported according to the data presented.
p.11: To identify the cause of the respiratory failure, the authors looked at the innervation pattern of the phrenic nerve in the diaphragm. Mutants present decreased branching but larger nerve extensions covering a wider innervated area and less neuromuscular junctions. Despite the decreased innervation of the diaphragm, its morphology is normal as well as the ultra-structure of the sarcomeres suggesting a mild phenotype rather than the cause of death of the mutants as suggested by the authors (p.20).
- Diaphragmatic muscle activity should be measured to establish if the contractile activity of the diaphragm is affected. This might support the statement of the authors.
p.12: The authors looked at the growth capacity of motor neurons and dorsal root ganglion neurons and showed a reduced growth in both cases.
- How do the authors reconcile the observation made in the diaphragm in which nerve extensions are larger with the reduced growth capacity of neurons?
p.12: The authors used cultured hippocampal neurons for imaging microtubule growth. According to the authors, the loss of CLASP1 deregulates microtubule dynamics.
- No explanation was provided to justify the use of hippocampal neurons. What is a catastrophe rate? What is the justification to study this parameter? What does it tell us about microtubule dynamics?
p.13: The authors examined lung from mutants. Mutant lungs do not float and they are collapsed at birth. However, lung morphology appears normal and myofibroblasts, ciliated cells and Club cells are present as shown by IHC labeling. No difference in proliferation and apoptosis was reported.
- It would have been more informative to do BrdU/EdU immunolabeling for proliferation in order to see if differences occur in specific cell types of the lung. It is not clear why the authors have limited their IHC analysis to these three specific cell types. A complete analysis should be done.
p.14: The authors proposed a delay in lung development according to lung morphology that appears more collapsed starting at E15.5.
- Measurement of branching would allow to quantify this delay. Since cell differentiation occurs ~E16.5, analysis of the onset of cell types can also support a delay in lung development.
p.15: Finally, the authors conclude this section by "these data support a direct role for CLASP1 in lung maturation".
- Which direct role? How? This sentence appears premature according to the data presented. The authors should look at microtubule dynamics in lung cells from mutant embryos to see if a link exists between the proposed role of the protein and the lung phenotype observed.
p.15: The authors attempted to rescue the defective lung maturation phenotype by treating pregnant females with dexamethasone at late gestational stages. Around 10% of mutants survive for more than 45 minutes to 2 hrs compared to 20-30 minutes for mutants obtained from untreated mothers (p.9). Even though it is an intriguing result, the very small numbers of "survivors" makes very difficult to reach a conclusion.
- This section should be shortened.
p.16: To determine which molecular mechanisms are responsible for the lung defect, the authors performed RNA-seq analysis on E18.5 lung specimens. The number of genes with significant differential expression was low and the highest scores were cathepsin E for the upregulated gene and chitinase-like 1 for the downregulated gene.
- Are these two genes known for their role in lung development? Please describe.
p.16: Except for the fact that Chil1 is also downregulated in mutant lungs for the H3K4 methyltransferase Mll3 gene, it is not clear why the authors compared these 2 sets of data.
- Can CLASP1 and MLL3 interact together? How? Did the authors looked at the list of genes that are commonly differentially expressed? Does it provide some clues on the mechanisms? The RNA-seq data should be analyzed more deeply.
p.16: There is also a Clasp2 gene with a more restricted expression pattern. Clasp2 mutant mice either die from hemorrhages or survive. It is not clear why the RNA-seq data of the lungs from Clasp2-/- mice are presented since no lung phenotype is mentioned for these mice. How the lack of change in Chil1 expression in Clasp2 mutant lungs is informative?
- This should be clarified or the data should be removed.
p.17: Aqp5 expression was decreased in mutant lungs as shown by RNA-seq data and RT-qPCR. However, immunolabelling with Ti does not show a decrease in the number of Type I pneumocytes (Fig. 7D). According to the data presented, it is difficult to conclude that CLASP1 is involved in Type I pneumocyte differentiation.
- A cell count should be done for Figure 7D. Immunolabeling with more markers for Type I pneumocytes, including AQP5 Ab, should be performed to determine if the decreased Aqp5 RNA expression correlates with less Type I cells. GSEA signature has to be confirmed by additional analyses.
p.17: The same comments can be made for Type II pneumocytes and SpC expression.
p.31: The authors mentioned a role for CLASP1 in the mesenchyme.
What are the experiments and data that support this sentence?
How do the authors reconcile their observation of CLASP1 expression in lung secretory cells (p.8) with their conclusion of defective Type I cell differentiation (p.17)?
Minor comment:
Legend of Figure S4 should be for Figure S5 and vice-versa.
Significance
In summary: a descriptive characterization of a Clasp1 mutant mouse line but no real clue on how this microtubule-associated protein acts to produce the phenotypes observed that likely cause the death of the mutant newborns.
This manuscript should interest researchers in lung developmental biology and cell biology,
My expertise: mouse models, lung development, gene regulation and networks
-
