PI3K signaling specifies proximal-distal fate by driving a developmental gene regulatory network in SOX9+ mouse lung progenitors
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This manuscript is one of the first broad epigenetic analyses in lung development and will be of interest to not only lung biologists but also to the field of epithelial developmental biology. Using paired transcriptomic and epigenetic data, they have uncovered a vast repertoire of signaling mechanisms underlying lung development. These findings have opened up the field's opportunities to understand and study novel pathways and have further defined a role for PI3 kinase signaling in lung development.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Developmental Biology (eLife)
Abstract
The tips of the developing respiratory buds are home to important progenitor cells marked by the expression of SOX9 and ID2. Early in embryonic development (prior to E13.5), SOX9+progenitors are multipotent, generating both airway and alveolar epithelium, but are selective progenitors of alveolar epithelial cells later in development. Transcription factors, including Sox9, Etv5 , Irx , Mycn , and Foxp1/2 interact in complex gene regulatory networks to control proliferation and differentiation of SOX9+progenitors. Molecular mechanisms by which these transcription factors and other signaling pathways control chromatin state to establish and maintain cell-type identity are not well-defined. Herein, we analyze paired gene expression (RNA-Seq) and chromatin accessibility (ATAC-Seq) data from SOX9+ epithelial progenitor cells (EPCs) during embryonic development in Mus musculus . Widespread changes in chromatin accessibility were observed between E11.5 and E16.5, particularly at distal cis-regulatory elements (e.g. enhancers). Gene regulatory network (GRN) inference identified a common SOX9+ progenitor GRN, implicating phosphoinositide 3-kinase (PI3K) signaling in the developmental regulation of SOX9+ progenitor cells. Consistent with this model, conditional ablation of PI3K signaling in the developing lung epithelium in mouse resulted in an expansion of the SOX9+ EPC population and impaired airway epithelial cell differentiation. These data demonstrate that PI3K signaling is required for epithelial patterning during lung organogenesis, and emphasize the combinatorial power of paired RNA and ATAC seq in defining regulatory networks in development.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
In this article the authors use paired gene expression and chromatin accessibility data on isolated Sox9 positive progenitors to identify a role for Pi3K in lung epithelial differentiation and branching. The authors show some intriguing findings.
We appreciate the generally positive and very constructive comments from both reviewers, which are highly aligned and helped us to improve the manuscript in revision. We have focused our time and attention on evaluation of the in vivo model of Shh-Cre/Pik3caf/f knockout animals, and have reorganized several figures to highlight these expanded data.
However, some additional experiments are necessary to confirm/validate their conclusions. Some issues with the current experiments make them hard to interpret.
- While the experiments in Fig.6 show …
Author Response
Reviewer #1 (Public Review):
In this article the authors use paired gene expression and chromatin accessibility data on isolated Sox9 positive progenitors to identify a role for Pi3K in lung epithelial differentiation and branching. The authors show some intriguing findings.
We appreciate the generally positive and very constructive comments from both reviewers, which are highly aligned and helped us to improve the manuscript in revision. We have focused our time and attention on evaluation of the in vivo model of Shh-Cre/Pik3caf/f knockout animals, and have reorganized several figures to highlight these expanded data.
However, some additional experiments are necessary to confirm/validate their conclusions. Some issues with the current experiments make them hard to interpret.
- While the experiments in Fig.6 show an increase in branching morphogenesis after treatment with different inhibitors, it is unclear whether this is because inhibition of Pi3K in the epithelium or mesenchyme.
We agree that one of the major limitations of the lung explant model is the inability to isolate the effects of pharmacologic treatments to specific cell-types within the lung. We think that the data itself is robust in that it is highly repeatable and reproducible with two unrelated pan-class I PI3K inhibitors. However, given the significantly increased in vivo data present in this revision, we decided to omit these data in the current version of the manuscript in order to focus on the epithelial specific roles appreciable in the Shh-Cre knockout animals.
- Similarly it is difficult to assess whether the effect on Sox9+ EPCs is due to the inhibitor acting on the epithelium or mesenchyme.
See response to #1 above.
- In the abstract the authors mention that prior to E13.5, SOX9+ progenitors are multipotent, generating both airway and alveolar epithelium, but are selective alveolar progenitors later in development. To further investigate this the authors isolated Sox9 positive progenitors at 11.5 and 16.5. The authors then as expected find some genes being differentially expressed in the progenitors at these different time points. However, while these changes in expression likely reflect the narrowed differentiation potential of the Sox9+ EPCs at E16.5 it is unclear whether this really helps to explain how Sox9+EPCs at E11.5 differentiate into proximal epithelium.
We agree that a significant number of the differentially expressed genes and differentially accessible chromatin regions observed between E11.5 and E16.5 reflect the “narrowed differentiation potential” of the lung epithelium, but the regulators of this differentiation potential are incompletely understood. We specifically chose to further interrogate the role of Pi3K signaling in proximal-distal patterning of the lung epithelium both based on these data, as well as from our prior work on congenital pulmonary airway malformations (a disease most commonly characterized by expansion of cystic airway structures). Our added data in revision, evaluating the contribution of Pik3ca signaling to epithelial maturation, validate the idea that these data can be used to identify new regulators of distal progression of differentiation at minimum.
- qPCR in Fig.8 reflects the lack of airways but doesn't reflect their differentiation, it appears differentiation in club and ciliated cells still occurs but appears delayed. Differentiation of the bronchial epithelium occurs after Sox9+ EPCs have differentiated into Sox2+ airway cells.
It is unclear if the differentiation of the Sox2+ airway epithelium is delayed or whether Pik3ca plays a role in the differentiation of these Sox2+ airway epithelial cells.
While we agree that there is not a complete impairment of airway epithelial differentiation (i.e. there are still club and ciliated cells present), our data imply that differentiation of airway epithelium into mature secretory and ciliated cells appears to be more heavily impacted than the generation of Sox2+ epithelium itself. We have added additional data and restructured Figure 8 to hopefully make this more clear. Immunofluorescence microscopy for Sox2 from E12.5-E18.5 (new Figure 8, panels A-X) and quantification of Sox2 mRNA at E18.5 (new Figure 8, panel NN) are now shown. Although there does appear to be a decrease in the number of airway branches (consistent with what is also seen in the H&E times series and data shown in Figure 6), the airways are still lined with Sox2+ epithelium and the reduction in Sox2 mRNA transcript at E18.5 is relatively small (and doesn’t reach statistical significance). In contrast, there is a dramatic reduction in the secretoglobin proteins Scgb1a1 and Scgb3a2 at both the mRNA and protein level, and the ciliated cell marker Foxj1 (at mRNA level). Moreover, there is a dramatic reduction in the percentage of secretory and ciliated cells in the Pik3caShhCre lungs. Thus, although we cannot exclude that there is a smaller but significant impairment in the generation of Sox2+ epithelium, it appears that the more significant phenotype present is the differentiation of airway epithelium into mature airway epithelium. We anticipate that our follow-up studies will refine the precise molecular mechanisms by which PI3K signaling directs differentiation of the lung epithelium.
-
Evaluation Summary:
This manuscript is one of the first broad epigenetic analyses in lung development and will be of interest to not only lung biologists but also to the field of epithelial developmental biology. Using paired transcriptomic and epigenetic data, they have uncovered a vast repertoire of signaling mechanisms underlying lung development. These findings have opened up the field's opportunities to understand and study novel pathways and have further defined a role for PI3 kinase signaling in lung development.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
In this article the authors use paired gene expression and chromatin accessibility data on isolated Sox9 positive progenitors to identify a role for Pi3K in lung epithelial differentiation and branching. The authors show some intriguing findings.
However, some additional experiments are necessary to confirm/validate their conclusions.
Some issues with the current experiments make them hard to interpret.- While the experiments in Fig.6 show an increase in branching morphogenesis after treatment with different inhibitors, it is unclear whether this is because inhibition of Pi3K in the epithelium or mesenchyme.
- Similarly it is difficult to assess whether the effect on Sox9+ EPCs is due to the inhibitor acting on the epithelium or mesenchyme.
- In the abstract the authors mention that prior to E13.5, SOX9+ …
Reviewer #1 (Public Review):
In this article the authors use paired gene expression and chromatin accessibility data on isolated Sox9 positive progenitors to identify a role for Pi3K in lung epithelial differentiation and branching. The authors show some intriguing findings.
However, some additional experiments are necessary to confirm/validate their conclusions.
Some issues with the current experiments make them hard to interpret.- While the experiments in Fig.6 show an increase in branching morphogenesis after treatment with different inhibitors, it is unclear whether this is because inhibition of Pi3K in the epithelium or mesenchyme.
- Similarly it is difficult to assess whether the effect on Sox9+ EPCs is due to the inhibitor acting on the epithelium or mesenchyme.
- In the abstract the authors mention that prior to E13.5, SOX9+ progenitors are multipotent, generating both airway and alveolar epithelium, but are selective alveolar progenitors later in development. To further investigate this the authors isolated Sox9 positive progenitors at 11.5 and 16.5. The authors then as expected find some genes being differentially expressed in the progenitors at these different time points. However, while these changes in expression likely reflect the narrowed differentiation potential of the Sox9+ EPCs at E16.5 it is unclear whether this really helps to explain how Sox9+EPCs at E11.5 differentiate into proximal epithelium.
4)qPCR in Fig.8 reflects the lack of airways but doesn't reflect their differentiation, it appears differentiation in club and ciliated cells still occurs but appears delayed. Differentiation of the bronchial epithelium occurs after Sox9+ EPCs have differentiated into Sox2+ airway cells.
It is unclear if the differentiation of the Sox2+ airway epithelium is delayed or whether Pik3ca plays a role in the differentiation of these Sox2+ airway epithelial cells.
-
Reviewer #2 (Public Review):
In the current manuscript, Khattar et al., have employed a novel analysis using a combinatorial approach composed of the epigenetic and transcriptomic assessment of the developing lung landscape. With this method, they have been able to reduce the noise in data that frequently complicates single assessment of open chromatin or differential gene expression only. In this instance, they were able to ascertain only active transcription factors based on the gene regulatory networks they opened in their analysis. From these data, they discovered a network of PI3 kinase signaling that has been previously implicated in lung development but now confirmed and complemented with their ex vivo and in vivo experiments.
The strengths of this study are the novelty of their epigenetic and transcriptomic combinatorial …
Reviewer #2 (Public Review):
In the current manuscript, Khattar et al., have employed a novel analysis using a combinatorial approach composed of the epigenetic and transcriptomic assessment of the developing lung landscape. With this method, they have been able to reduce the noise in data that frequently complicates single assessment of open chromatin or differential gene expression only. In this instance, they were able to ascertain only active transcription factors based on the gene regulatory networks they opened in their analysis. From these data, they discovered a network of PI3 kinase signaling that has been previously implicated in lung development but now confirmed and complemented with their ex vivo and in vivo experiments.
The strengths of this study are the novelty of their epigenetic and transcriptomic combinatorial analysis in the developing lung. The data they have generated will be of significant interest to the lung field, and it has provided a platform for future studies. Their vast bioinformatics analysis supports their conclusions on the first part of this study.
From their bioinformatic analysis, the authors identified a gene regulator network implicating PI3 kinase signaling in branching and proximal-distal axis formation. While much of their data for the remainder of this study supports this pathway's involvement, there remains many unanswered questions to fully support their conclusions.
First and foremost, the authors implicate that there is a gradient of PI3K signaling in the developing lung epithelium based on expression of phospho-AKT. This conclusion would be better supported with confirmation of expression (by qPCR) of several of the previously identified genes in the PI3K regulatory network they identified earlier. In addition, AKT can be modulated by other pathways. The authors should consider examining an additional read out of PI3K signaling. Furthermore, the current imaging has not delineated the proximal-distal axis (Sox2-Sox9), which develops and is sustained from E12.5 to the saccular stage. Additional imaging showing that gradient exists along the proximal-distal axis as well as other time points during development would be critical for their conclusion.
The authors used both ex vivo and in vivo modeling to support a role for PI3K in lung development. Inhibition of PI3K in both systems results in expansion of SOX9+ EPCs. However, the mechanism is quite unclear here, especially in the presence of decreased proliferation of SOX9+ EPCs. Is the expansion of SOX9 EPCs at the expense of proximal SOX2+ epithelial progenitor cells during development. The mechanism would be better supported by assessment of SOX2 EPCs during early development with particular attention to the proximal-distal axis and appropriate branching at early time points. In addition, previous studies demonstrated that the branching defects such as the authors demonstrated result in alveolar specification defects and explain their findings of differentiation defects of both proximal and distal lineages.
Finally, additional controls are need to support their conclusions. In particular, deletion of PI3Kca should be confirmed by its gene expression and pAKT during development. As mentioned earlier, additional confirmation of some of PI3K signaling component and target gene upregulation by RNAseq should be confirmed by qPCR and/or RNAscope/in situ or IHC. These studies would help integrate previous findings in PI3K signaling and lung development and provide a more comprehensive understanding of the mechanism.
-
Reviewer #3 (Public Review):
This study addresses the gene regulatory network that controls lung branching and proximal/distal patterning. Focusing on E11.5 versus E16.5 epithelium, Monocle analysis of transcriptome signatures revealed 26 gene expression modules. This was followed by RNA-seq and ATAC-seq analysis of Sox9+ cells from these same two stages. Integration of the two datasets, and comparison to ENCODE histone data revealed links from differential accessible chromatin peaks to gene expression differences, to likely transcription factors that may bind to CREs. These analysis revealed a role of PI3K in Sox9+ cell differentiation. Pharmacological inhibition of PI3k led to increase in branching. In vivo inactivation of Pi3kca led to persistence of Sox9+ cells, alveolar cysts and reduction of airway cell differentiation.
This is a …
Reviewer #3 (Public Review):
This study addresses the gene regulatory network that controls lung branching and proximal/distal patterning. Focusing on E11.5 versus E16.5 epithelium, Monocle analysis of transcriptome signatures revealed 26 gene expression modules. This was followed by RNA-seq and ATAC-seq analysis of Sox9+ cells from these same two stages. Integration of the two datasets, and comparison to ENCODE histone data revealed links from differential accessible chromatin peaks to gene expression differences, to likely transcription factors that may bind to CREs. These analysis revealed a role of PI3K in Sox9+ cell differentiation. Pharmacological inhibition of PI3k led to increase in branching. In vivo inactivation of Pi3kca led to persistence of Sox9+ cells, alveolar cysts and reduction of airway cell differentiation.
This is a stellar study that integrates transcriptomic and epigenomic signatures of the same lung cell population. The authors used innovative algorithms and the ENCODE dataset to tease out key drivers in GRN. This is effectively coupled with in vitro and in vivo functional tests. The findings advance current understanding of lung development. Deeper interrogation of the predicted link between transcription factors and PI3K pathway in the mutant would strengthen the overall cohesiveness of the message from this study.
-
