ORIGINAL RESEARCH published: 22 February 2022 doi: 10.3389/fped.2021.780166 Fetal Tracheal Occlusion Increases Lung Basal Cells via Increased Yap Signaling Vincent Serapiglia 1† , Chad A. Stephens 2† , Rashika Joshi 3 , Emrah Aydin 4,5 , Marc Oria 5,6 , Mario Marotta 7 , Jose L. Peiro 5,6 and Brian M. Varisco 3,6* 1 School of Medicine, Northeast Ohio College of Medicine, Northeast Ohio Medical University, Rootstown Township, OH, United States, 2 Department of Psychiatry, University of Illinois at Chicago, Chicago, IL, United States, 3 Division of Critical Care Medicine, Department of Pediatrics, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, United States, 4 Department of Pediatric Surgery, Tekirdag Namik Kemal University School of Medicine, Tekirdag, Turkey, 5 Center for Fetal and Placental Research, Cincinnati Children’s Hospital Medical Center (CCHMC), Cincinnati, OH, United States, 6 Department of Pediatrics, College of Medicine, University of Cincinnati, Cincinnati, OH, United States, 7 Bioengineering, Cell Therapy and Surgery in Congenital Malformations Laboratory, Vall d’Hebron Hospital Research Institute, Universitat Autònoma de Barcelona, Barcelona, Spain Edited by: Fetal endoscopic tracheal occlusion (FETO) is an emerging surgical therapy for congenital Sule Cataltepe, Brigham and Women’s Hospital and diaphragmatic hernia (CDH). Ovine and rabbit data suggested altered lung epithelial cell Harvard Medical School, populations after tracheal occlusion (TO) with transcriptomic signatures implicating basal United States cells. To test this hypothesis, we deconvolved mRNA sequencing (mRNA-seq) data and Reviewed by: used quantitative image analysis in fetal rabbit lung TO, which had increased basal cells Augusto Schmidt, University of Miami, United States and reduced ciliated cells after TO. In a fetal mouse TO model, flow cytometry showed Rodney Britt, increased basal cells, and immunohistochemistry demonstrated basal cell extension to Nationwide Children’s Hospital, United States subpleural airways. Nuclear Yap, a known regulator of basal cell fate, was increased Florian Friedmacher, in TO lung, and Yap ablation on the lung epithelium abrogated TO-mediated basal cell University Hospital expansion. mRNA-seq of TO lung showed increased activity of downstream Yap genes. Frankfurt, Germany Human lung specimens with congenital and fetal tracheal occlusion had clusters of *Correspondence: Brian M. Varisco subpleural basal cells that were not present in the control. TO increases lung epithelial brian.varisco@cchmc.org cell nuclear Yap, leading to basal cell expansion. † These authors have contributed Keywords: fetal tracheal occlusion, congenital diaphragm hernia, lung development, basal cell, Yap (Hippo) equally to this work signaling, mechanotransduction Specialty section: This article was submitted to INTRODUCTION Neonatology, a section of the journal Congenital diaphragmatic hernia (CDH) results from the failed fusion of the pleuroperitoneal Frontiers in Pediatrics folds in the 6th to the 8th week of gestation in humans and occurs in 1:2,000 births (1). In CDH, Received: 20 September 2021 abdominal contents herniate into the thorax, causing ipsilateral and contralateral lung hypoplasia Accepted: 15 December 2021 and pulmonary vascular pruning. The morbidity and mortality associated with CDH are principally Published: 22 February 2022 determined by the size of the defect and the laterality, with right-sided hernia experiencing worse Citation: outcomes. The most severe defects account for 20–35% of all CDH and have 50% mortality with Serapiglia V, Stephens CA, Joshi R, severe respiratory morbidity in survivors (2). For this reason, a fetal surgical technique has been Aydin E, Oria M, Marotta M, Peiro JL and Varisco BM (2022) Fetal Tracheal developed to improve fetal lung growth in CDH. Occlusion Increases Lung Basal Cells Fetal endoscopic tracheal occlusion (FETO) was first described in animal models of CDH by via Increased Yap Signaling. Harrison et al. (3), and human clinical trials have been ongoing for 35 years (1). Typically, FETO Front. Pediatr. 9:780166. involves the placement of an endoluminal tracheal balloon at 26–29 weeks’ gestation with fetoscopic doi: 10.3389/fped.2021.780166 removal at 34 weeks. Tracheal occlusion (TO) prevents the egress of epithelial secretions and Frontiers in Pediatrics | www.frontiersin.org 1 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells causes lung expansion and growth. A meta-analysis of multiple MATERIALS AND METHODS small randomized trials indicated a survival benefit of FETO, but suggested continued respiratory morbidity (1). Larger Human Tissue Use randomized trials are ongoing. There are mixed reports of the Archived human tissues were used with a waiver from the impact of CDH on alveolar type II cell function (4–6) and a Cincinnati Children’s Hospital Institutional Review Board (2016- report that CDH reduces and TO increases TGF-β signaling 9641) and from the Vall d’Hebron Research Institute. and lung elastin synthesis (7). Two whole-lung gene expression profiling studies using a rabbit model reported complementary Animal Care and Maintenance findings of decreased cell proliferation in CDH with restoration All animal use was approved by the Cincinnati Children’s or excessive restoration of cell division with FETO (8, 9). Hospital IACUC (2019-0034). Timed pregnant C57BL/6 mice Our group reported preferential proliferation in the epithelial were housed in a barrier facility with chow and water ad libitum compartment in FETO that was potentially EGFR-related (9). and 12-h day/night cycles. Timed pregnant New Zealand white Proteomic analysis of tracheal fluid in an ovine model of CDH rabbits were purchased from Charles River (Wilmington, MA, and FETO found that PI3K and mTOR signaling was increased USA) and similarly housed. with reduced abundance of ciliated cells in FETO (10). Adding to the complexity of animal models is a recent finding that, in a Rabbit CDH and TO Model rabbit model of CDH and FETO, there are significant metabolic The fetal rabbit CDH and TO model has been previously differences between and within lung lobes (11). A barrier to described (9). In brief, after anesthesia of the pregnant rabbit, understanding the mechanistic underpinning of changes in CDH in fetuses, a left-sided diaphragmatic defect was created at and TO has been the lack of a mouse TO model with high fetal embryonic day 25 (E25), TO was performed at E27, and fetuses survival rates. This obstacle has been recently overcome with the were collected at E30. The left lower lung was collected and development of a transuterine mouse TO model (12, 13). used for RNA sequencing and the left upper lobe collected The Yes-associated protein (Yap) signaling cascade was first for histology. descried in Drosophila as the Hippo signaling pathway and is a key mechanosensitive integrator of several developmental Mouse TO Model processes in the lung epithelium. Yap is phosphorylated The transuterine mouse TO model has been previously described by large tumor suppressor kinases 1 and 2 (Lats1 and (12, 13). Briefly, E16.5 C57BL/6 mice were anesthetized and the Lats2), which are phosphorylated by mammalian Ste20-like fetuses exposed by midline laparotomy. Anterior-facing fetuses serine/threonine kinases 1 and 2 (Mst1 and Mst2), which had ligation of the trachea and one carotid artery and jugular vein are in turn phosphorylated by cytoskeleton-coupled proteins by passing a non-cutting needle with 6–0 silk through the uterine at epithelial cell apical membranes. Phosphorylated Yap is wall and around those structures. Progesterone (2 mg) was transcriptionally inactive and sequestered in the cytoplasm. administered intramuscularly at E17.5 to prevent spontaneous Non-phosphorylated Yap is translocated into the nucleus, abortion. At E18.5, pregnant mice were anesthetized, the fetuses where it binds with TEAD (transcriptional enhancer factor collected, and both fetuses and dam sacrificed. domain proteins) factors, which are the enhancers of Yap target genes. The temporal and spatial roles of Yap in the developing lung are still emerging, but in general, in the Conditional Yap Deletion conducting airway epithelium, nuclear Yap is important for Based on previous studies (15) showing efficient depletion of Yap maintenance of the stem cell lineage and basal cell proliferation from the airway epithelium, we crossed mice with homozygous (14, 15); in the distal lung, shuttling of Yap between the floxed Yap1 exons (B6.Yap1tm1.Dupa , YapLoxp/LoxP ) with mice nuclear and cytoplasmic compartments is important in alveolar heterozygous for cre under the control of the Sonic hedgehog type 2 (AT2) to alveolar type 1 (AT1) cell differentiation promoter (B6.Shhtm1(EGFF/cre)Cjt , Shhcre ) and a single floxed Yap (16, 17). allele (Shhcre Yap+/1 ) to create ShhCre Yap1/1 fetuses. Genotyping The goal of this study was to characterize the impact of was performed as previously described (15). fetal TO on the lung epithelium. To do so, we first performed a post-hoc analysis of cell-specific genes derived from the left Tissue Processing lower lobes of fetal rabbits with CDH and/or TO and then Lung tissue was snap frozen on dry ice and stored at −80◦ C. validated these findings by histology of the left upper lobes. The lower left lobe in rabbits and the entire lung in mice After identifying basal cells as the most dysregulated cell type, were homogenized and the RNA extracted using RNEasy we analyzed the whole lungs of mice in a mouse model of kits (Qiagen, Hilden, Germany). Rabbit lung histology was FETO to identify these same changes and identify Hippo/Yap obtained from frozen left upper lung tissue in optimum cutting as key transcriptional regulators using a lung epithelial cell- temperature (OCT) compound (Tissue-Tek, Torrance, CA, USA) specific knockout approach. We lastly show that the lungs of and cryostat sectioning. Entire fetal mouse lungs were fixed human neonates with congenital TO have clusters of basal in 4% paraformaldehyde in phosphate-buffered saline (PBS), cells in the distal lung that are not present in the lung of dehydrated, paraffinized, and sections obtained. For lung lysates, control infants. frozen lung tissue was homogenized in RIPA buffer. Frontiers in Pediatrics | www.frontiersin.org 2 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells TABLE 1 | Primary antibodies. Flow Cytometry E18.5 mouse lungs were collected and a single cell suspension Antibody Manufacturer/source Use Dilution (catalog no.) created as previously described (18). Cells were stained for nerve growth factor receptor (NGFR) (ab8874, 1:200; Abcam) and NGFR Abcam (ab8874) IHC 1:1,000 EpCam (47-579180, 1:200; eBioscience, San Diego, CA, USA) Yap Cell Signaling (4912S) IHC 1:1,000 antibodies using Zombie Red (423109; BioLegend, San Diego, Phospho-Yap Cell Signaling (13008S) IHC 1:1,000 CA, USA) viability dye for dead cell exclusion on a FACSCanto Ki67 Abcam (ab16667) IHC 1:1,000 cell (BD Biosciences, Franklin Lakes, NJ, USA) flow cytometry Pro-surfactant protein C (guinea pig) Gift from Dr. Jeffrey IF 1:200 device and using FlowJo v10.7.1 software for quantification. Whitsett CGRP Abcam (ab360001) IF 1:200 PCR Acetylated alpha tubulin Millipore (MABT868) IF 1:200 The quantities of lung cell-specific messenger RNAs (mRNAs) Cytokeratin 14 Abcam (ab7800) IF 1:200 in rabbit lung was quantified by SYBR Green (204143; Qiagen) HopX Abcam (ab57832) IF 1:200 RT-PCR using the oligonucleotides in Supplementary Table 3. Mucin 5AC Abcam (ab3649) IF 1:200 Taqman PCR of mouse lung RNA was performed using Applied Scgb1a1 (guinea pig) Gift from Dr. Jeffrey IF 1:200 Biosystems proprietary primers for connective tissue growth Whitsett factor (CTGF) (Mm01192932_g1), Trp63 (Mm00495788_m1), Yap Cell Signaling (4912S) IF 1:200 and GAPDH (cat. no.4308313; Applied Biosystems, Waltham, Keratin-5 Abcam (ab53121) IF 1:200 MA, USA). mRNA-Seq and Bioinformatic Analysis Lung cell-specific mRNAs were extracted from LungGENS E16.5, Immunofluorescence, E18.5, postnatal day 1 (PND1), PND3, PND7, and PND28 Immunohistochemistry, and Image profiles (19). The FKPM (fragments per kilobase of transcript per million mapped reads) values of these genes were extracted from Acquisition a previously published rabbit CDH and TO dataset (GSE84811). For immunohistochemistry (IHC), mouse lung sections were The median FKPM per group was used for the comparison of stained using the antibodies below (Table 1) with appropriate cell-specific mRNAs. For mouse TO experiments, RNA from secondary antibodies and the ABC Vectastain kit (Vector control and TO mouse lungs was barcoded using NexteraXT and Laboratories, Burlingame, CA, USA). For immunofluorescence Illumina oligos and sequenced at a depth of 10 million reads (IF), rabbit lung sections were stained with the antibodies per sample with 150-bp paired-end sequencing on a NovaSeq below and appropriate fluorophore-conjugated secondary sequencer (Illumina, San Diego, CA, USA). Reads were aligned antibodies. Brightfield imaging was performed on a Nikon NiE to GRCm38 using STAR (20) and count files were imported into upright microscope with ×10 tiled and ×40 magnification. DESeq2 [(21), p. 2] using a threshold of five reads per gene in Immunofluorescent images were obtained using a Nikon all samples. Differential expression by group was determined. SpectraX microscope. Deferentially expressed genes (DEGs) with an adjusted p-value of <0.1 and a fold change of two or greater were exported into Quantitative Image Analysis ToppGene (22) and Ingenuity Pathway Analysis (IPA) (23) for For rabbit lung specimens, images of the entire lung lobe gene set enrichment analysis (GSEA). at ×10 magnification were generated by tile scanning with autofocusing of each frame at ×10 magnification on a Nikon Statistical Analysis SpectraX microscope. For each analyte, a negative control R version 4.0.2 (24), ggpubr (25), and rstatix (26) were used image was used to determine an intensity threshold and for graphical and statistical comparisons using one-way ANOVA size thresholds determined based on the marker (nuclear vs. and Holm–Sidak post-hoc test for parametric and Kruskal–Wallis cytoplasmic, vs. apical membrane). Non-lung material was with Dunn’s post-hoc test for non-parametric comparisons. If manually excluded, and the percentage of cells expressing comparisons between more than four groups were performed, we each analyte was determined as nuclei associated with the applied Bonferroni correction for multiple comparisons. P < 0.05 marker divided by the total nuclei. These percentages were then considered significant. Line and whisker plots indicate medians compared using the Wilcoxon rank-sum test with Kruskal– and 25th and 75th percentiles. Wallis post-hoc comparisons. RESULTS Western Blot Mouse lung homogenates were electrophoretically separated and Increased Numbers of Lung Basal Cells in transferred to PVDF membranes and probed for cytokeratin-14 the Lung Tissue of the Rabbit FETO Model (ab7800, 1:1,000; Abcam, Cambridge, UK) with quantification To begin identifying the lung cells most impacted by FETO, we of total protein using REVERT total protein stain (cat. no. 923- generated lists of cell-specific mRNAs (Supplementary Table 1) 11011; LI-COR, Lincoln, NE, USA). using the Lung Gene Expression Analysis database (19). Frontiers in Pediatrics | www.frontiersin.org 3 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells FIGURE 1 | Increased basal cells in a rabbit model of fetal tracheal occlusion. (A) Lung cell-specific mRNAs and the relative expression of each in rabbit fetuses with sham surgery, diaphragmatic hernia creation (CDH), tracheal occlusion (TO), or both (CDHTO) were identified. Each data point represents the relative expression of one gene per group (n = 4 per group). This analysis identified a relative increase in basal cell- and a reduction in ciliated cell-associated genes. *p < 0.05, **p < 0.01, ****p < 0.0001 by Dunn’s post-hoc test with Bonferroni correction. (B) Immunostaining and quantitative image analysis of the upper left lobe rabbit lung showing a relative increase in the number of basal cells as a fraction of the total cells in TO lung (n = 4–6 per group). (C) The basal cell master regulator p63 was present in conducting airways of control rabbit fetuses largely below SCGB1A1-positive club cells. (D) The relative abundance of p63-positive cells was increased in the lungs of TO rabbit fetuses. (E) Apparent reductions of both alveolar type 1 and type 2 cells were not statistically significant. (F) Immunostaining for the alveolar type 2 cell marker pro-surfactant protein C (pro-SFTPC) and the alveolar type 1 cell marker HOPX in sham-treated fetal rabbit lung. (G) CDHTO lung appeared to have reductions in both markers. Normalizing the FKPM values of fetal rabbits undergoing This analysis of lung tissue from a previously published rabbit fetal creation of diaphragmatic hernia (CDH), FETO, or CDH/TO study showed that TO lung has increased numbers of both (CDH/FETO) to control, we identified increased relative basal cells compared to control. expressions of basal cell-specific mRNAs and decreased relative expressions of club, ciliated, and goblet cell mRNAs in FETO and CDH/FETO lungs. There were no substantial differences Increased Number of Basal Cells After in the relative abundances of AT1, AT2, lipofibroblast, matrix Fetal Mouse Tracheal Occlusion fibroblast, myofibroblast, epithelial, or myeloid cell mRNAs We next used a recently described mouse fetal TO model (12) (Figure 1A). As the focus of this project was TO, we extracted to test whether the same increase in lung basal cells was seen in a the 98 genes with adjusted p < 0.1 and 1.5-fold increased different animal model of TO. After TO at E16.5, E18.5 fetal lungs or decreased expression in control vs. TO and performed were collected and single cell suspensions were analyzed by flow GSEA. Cell cycle-related processes were enriched and basal cell cytometry. TO lungs had more than twice the number of basal genes were overrepresented (Table 2). Quantification of cell- cells of control lungs (Figures 2A,B). In a separate experiment, specific mRNAs was consistent with these bioinformatic findings the lungs of TO mice had ∼2.5-fold greater quantity of the (Supplementary Figure 1). On the protein level, cell-specific basal cell marker ketatin-14 (Figure 2C). Analysis of whole-lung markers identified an increased number of basal cells in TO sections of control and FETO mice revealed increased abundance lung (Figures 1B–D) and reduced numbers of AT1 cells in CDH, and more distal extension of basal cells in TO lung compared TO, and CDH/TO lung compared to control (Figures 1E–G). to control (Figures 2D,E). These fetal mouse data corroborate Frontiers in Pediatrics | www.frontiersin.org 4 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells findings in a fetal rabbit model showing that TO increases the 6.43E-24 2.83E-19 2.09E-18 3.55E-18 1.17E-16 1.77E-16 1.77E-16 5.64E-14 2.24E-13 2.41E-13 q-value number of basal cells. Activation of Yap Signaling by FETO No. Genes % Genes 20% The Yap pathway is both a regulator of epithelial cell proliferation 17% 17% 17% 16% 16% 16% 14% 14% 14% and organ size and a direct regulator of p63, a necessary transcription factor for basal cells (15). Yap signaling is regulated by a series of kinases, with phosphorylated Yap (phospho-Yap) Topp cell atlas 20 17 17 17 16 16 16 14 14 14 being targeted for degradation and non-phosphorylated Yap translocating to the nucleus where TEAD factors are recruited, activating Yap-regulated genes (27). Thus, non-phosphorylated 1.86E-09 Early Lung Fibroblast Progenitor Yap is transcriptionally active, while phospho-Yap is sequestered 4.08E-10 Dividing Alveolar Macrophage 1.34E-10 Basal Cell (Mouse Trachea) in the cytoplasm. The immunohistochemistry for Yap and 1.34E-10 Cycling Basal Cell (Mouse 1.73E-09 Developing Plasmablast phospho-Yap identified grouped Yap-positive and phospho-Yap- 4.03E-09 AT1-AT2 progenitors 6.62E-10 Neuroepithelial cell positive cells in E18.5 lung (Figures 3A,B). In contrast, in E18.5 3.83E-10 Activated T cell TO lungs, most nuclei were Yap-positive, and phospho-Yap 1.34E-10 Dividing T-cell staining was nearly absent (Figures 3C,D). The mRNA levels Trachea) of the Yap target genes connective tissue growth factor (CTGF) 4.44E-09 B-cell Term and Trp63 (p63) were both increased in TO lung (Figures 3E,F). These data demonstrate the activation and transcriptional q-value activity of Yap following TO. To test whether TO activation of Yap was directly responsible for basal cell expansion, we conditionally deleted Yap from the No. Genes % Genes epithelium of fetuses using a Sonic hedgehog driver and floxed 33% 27% 27% 21% 28% 28% 5% 5% 5% 5% Yap alleles. There was a non-significant reduction in the number of basal cells between YapLoxP/LoxP or Yap+/1 mice and Yap1/1 GO biological function at baseline and a significant reduction in the number of post-TO 32 26 26 21 27 27 basal cells in Yap1/1 (Figures 3G–I). While these data cannot 5 5 5 5 differentiate whether the lack of TO-induced basal cell expansion is from the absence of a necessary progenitor or the lack of a assembly involved in nuclear cell 1.92E-02 circulatory system development assembly involved in cell cycle 4.59E-02 double-strand break repair via necessary transcriptional driver, they do show that Yap is essential 1.65E-02 regulation of cell population for basal cell expansion in TO. break-induced replication 3.88E-02 vasculature development 3.65E-05 pre-replicative complex 2.68E-03 pre-replicative complex 1.60E-02 pre-replicative complex cycle DNA replication Pathways and Processes Impacted by 5.79E-02 biological adhesion 1.73E-02 tube development DNA replication Tracheal Occlusion 5.59E-02 cell adhesion proliferation We examined the relative weight of Yap transcriptional activity assembly compared to other impacted signaling pathways by performing Term mRNA sequencing (mRNA-seq) and differential gene expression analysis in three control and three TO E18.5 lungs. TO lungs q-value TABLE 2 | Gene set enrichment analysis of fetal rabbit to vs. Control. were most similar to each other (Figure 4A), although the separation between control and TO specimens was incomplete when assessed by the first two principal components (Figure 4B). No. genes % Genes 15% 11% After filtering for genes that contained at least five reads in 5% 8% 7% 5% 6% 3% 3% 7% three or more of the specimens, 17,374 genes were used in the analysis. Among these genes, only 42 had adjusted p-values GO molecular function of <0.1. GSEA of these genes identified inflammation-related, 15 11 5 8 7 5 6 3 3 7 tube morphogenesis, and blood vessel morphogenesis processes as differentially regulated (Supplementary Table 2). For further analysis, we analyzed genes that were 2-fold increased (626 Catalytic activity, acting on DNA Cell adhesion molecule binding Single-stranded DNA helicase DNA replication origin binding Single-stranded DNA binding genes) or decreased (207 genes). Among the upregulated genes, Extracellular matrix structural Glycosaminoglycan binding 3’-5’ DNA helicase activity the biological processes that were activated were cardiac and striated muscle function-related, and there were no significant DNA helicase activity Calcium ion binding molecular functions identified among the downregulated genes (Supplementary Figure 2). Upstream analysis of genes with a constituent 2-fold change in expression identified interferons and Toll-like activity receptors as activated, but it also identified regulators important Term in cell cycle and lung development, such as TBX5, SPI1, CEBPA, Frontiers in Pediatrics | www.frontiersin.org 5 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells TGFB1, AREG, FOXM1, and CCND1 (Figure 4C). Yap/Taz in the nucleus) leads to the disorganization of the conducting phosphorylation was predicted to be decreased (activation z- airway epithelium without evidence of increased proliferation score = −0.82, adjusted p = 1.49 × 10−7 ) and TEAD2 activity and with evidence of ectopic AT2 and AT1 cell markers (17). increased (activation z-score = 1.63, adjusted p = 2.7 × 10−7 ) However, Yap has also been recently shown to mediate AT1- in this analysis, but the relative weights of these differences were to-AT2 cell conversion with an important role in post-injury smaller than those of other regulators. Clustering of specimens repair (29). Mst1 and Mst2 phosphorylate Yap, leading to using genes known to be regulated by Yap showed that TO and cytoplasmic sequestration. Mst1 and Mst2 deletion also impairs control specimens were distinct from one another (Figure 4D). the differentiation of conducting airway epithelial cells, but In performing analysis of lung cell-specific mRNAs similar to that increases the proliferation of these undifferentiated epithelial done in fetal rabbit lungs, it was found that basal cells, club/goblet cells (14). In the study of Lange et al. increased total and nuclear cells, AT2, and myeloid cell mRNAs were increased in TO lung Yap led to the expansion of lung basal cells at the expense of club compared to control (Figure 4E). cells. We did not observe any increases in AT1, as previously described in mice with deletions of Lats1 and Lats2 (which Human Tracheal Occlusion Results in are downstream of Mst1 and Mst2, also leading to increased Increased Yap and Extension of Basal Cells nuclear Yap) and in a recent study of ex vivo inflation of fetal mouse lungs (30). Interestingly, ectopic p63-positive cells were Distally observed in Mst1- and Mst2-deficient mice (16). Our finding that To determine whether Yap activation and distal lung basal a TO-mediated increase in nuclear Yap mediates the expansion cell expansion also occurred in human TO, we analyzed lung of the lung basal cell population is largely consistent with the specimens from two infants who died shortly after birth with literature on the role of Yap in lung epithelial cell differentiation congenital high airway obstruction (CHAOS) (gestational ages, and function, with discrepant findings likely attributable to the 34 and 36 weeks), three infants who died either before or shortly differences in the five models (constitutive nuclear Yap, Mst1 and after delivery following FETO for CDH, and two lung specimens Mst2 deletion, Lats1 and Lats2 deletion, ex vivo lung inflation, of infants who died of non-pulmonary causes at those same ages. and TO with lung epithelial cell Yap deletion). All lung specimens had the typical position and spacing of basal Clinically, little is known about the post-FETO lungs of infants cells in the major conducting airways (Figure 5A), and basal with CDH. While FETO improves the survival of fetuses with cells were not identified apart from these conducting airways the most severe forms of CDH (1), impaired lung function is a in control lung specimens (Figure 5B). CHAOS lung specimens substantial problem in survivors. The importance of basal cell (Figures 5C,D) had clusters of basal cells in the airspaces of the clusters in the distal lung is also unclear. Basal cells or basal cell distal lung with Yap-positive nuclei close to these cells. FETO progenitors have been shown to be contributors to distal lung lung specimens (Figures 5E,F) showed expansion of basal cells regeneration following influenza infection (28), and keratin-14- in the distal airways, although nuclear Yap staining was not positive cell “pods” can be identified in normal and regenerating prominent. These data are in agreement with the findings in adult distal lungs (31). Increased percentages of basal cells a mouse model of TO with regard to Yap activation and the are associated with several non-malignant pulmonary diseases presence of basal cells in the distal lung airspaces. when assessed by single-cell mRNA-seq (32). Our studies are not comprehensive enough to describe when keratin-14-positive DISCUSSION pods may appear in the human lung as we did not observe any in control lung specimens. Whether these cells persist or are This is the first study to show changes in the distribution of adaptive or deleterious in the context of subsequent lung injury lung cell populations in response to FETO and is also the first can now be assessed by removal of the tracheal ligature at E17.5 to study the role of a specific signaling cascade in TO. While with subsequent parturition and raising of post-TO mice (33). the role of Yap in specifying lung epithelial cell populations is Although we focused on Yap and basal cells, our bioinformatic well established, this study is the first to show its relevance in data suggest that other signaling cascades are more significantly human lung development. To the best of our knowledge, this is impacted than this one. The most notable difference between also the first study using transgenic mice as fetal TO models. To TO and control mouse lungs was the increased inflammation date, animal studies have been hampered by the need to use large in TO lung. The role of different lung resident macrophages in animal models that are less amenable to genetic manipulation. lung development is being increasingly appreciated (34), and The development of a minimally invasive mouse TO model (12) how TO alters the relative abundance and activity of these allowed us to use a lung epithelial cell-specific deletion of Yap populations is likely important in post-TO lung. TO lung also to mechanistically examine its role in the observed expansion of has activation of several cardiac and skeletal muscle-related basal cells in TO mouse lung. pathways. This could be from three non-exclusive sources: Yap has well-established roles in defining the proximal lung myofibroblasts, vascular smooth muscle cells, and airway conducting airway epithelium in early lung development (15) smooth muscle cells. Lung mesenchymal cells in general and lung and in maintaining lung basal cells in more mature lung (28). myofibroblasts in particular play an important role in saccular In the distal lung epithelium, cytoplasmic Yap plays an important and alveolar lung development (35, 36). Pulmonary vascular role in AT2 cell and eventual AT1 cell differentiation. Failure of pruning, vascular smooth muscle hypertrophy, and pulmonary Yap to undergo nuclear–cytoplasmic shuttling (i.e., sequestered hypertension are substantial problems for infants with CDH Frontiers in Pediatrics | www.frontiersin.org 6 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells FIGURE 2 | Fetal endoscopic tracheal occlusion (FETO) increases the number and distalization of lung basal cells. (A) The percentage of lung basal cells as defined by the co-expression of EpCam and NGFR increased from ∼1.5 to ∼2.5% at E18.5 following tracheal occlusion (TO) at E16.5. (B) The increased percentage of basal cells in TO (n = 6) compared to control (n = 7) was consistent and statistically significant. (C) The whole-lung protein content of the basal cell marker keratin-14 was also increased (n = 5 per group). (D) Immunohistochemistry of control lung for NGFR showed restriction of basal cells to the basement membrane of large conducting airways. (E) FETO lungs had increased cellularity and had extension of basal cells beyond the central conducting airways into the more distal conducting airways. (D′ ), Magnified Image of D; (E′ ), Magnified Image of E. Frontiers in Pediatrics | www.frontiersin.org 7 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells FIGURE 3 | Increased lung epithelial Yap signaling is responsible for basal cell expansion. (A,B) In control E18.5 fetuses, non-phosphorylated (nuclear, active) Yap is present in small clusters of distal lung cells (A) and phosphorylated (cytosolic, non-active) Yap is widely present (B). (C) In E18.5 TO lung, non-phosphorylated Yap is present more homogenously throughout the lung and can be detected more in nuclei than in control, with denser abundance of nuclear-Yap-containing cells round airways. (D) In TO lung, phosphorylated Yap is less abundant than in control lung. (E,F) PCR of lung homogenates showed increased levels of connective tissue (Continued) Frontiers in Pediatrics | www.frontiersin.org 8 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells FIGURE 3 | growth factor (CTGF) (E) and p63 (F). Both are Yap target genes. Comparison is by Welch’s t-test. (G) Mice with hemizygous deletion of lung epithelial cell Yap at baseline (n = 9) and after TO (n = 5) showed similar numbers of basal cells to wild-type (WT) mice. Lung epithelial cell-specific deletion of Yap without TO (n = 4) or with TO (n = 5) both had basal cell frequencies comparable to control hemizygous and WT mice. *p < 0.05, **p < 0.01 by Holm–Sidak post-hoc test. (H,I) Compared to wild-type (H), mice with conditional deletion of Yap from the lung epithelium (I) had reduced abundance of basal cells as assessed by nerve growth factor receptor (NGFR) immunohistochemistry. Scale bars, 500 and 100 µm. (A′ ), Magnified Image of A; (C′ ), Magnified Image of C; (H′ ), Magnified Image of H; (I′ ), Magnified Image of I. FIGURE 4 | Yap in mouse tracheal occlusion (TO) lung. (A) mRNAs from the lungs of three control and three TO fetuses analyzed by mRNA-seq. As assessed by sample distances, TO and control lungs were most similar to one another. (B) Principal component plot showing incomplete separation of control from TO specimens. (C) Ingenuity Pathway Analysis of upstream regulators of genes that were 2-fold changed between control and TO lungs showing that the regulators with the greatest predicted activation (red) were largely related to inflammation and those with the greatest predicted inactivation (blue) were mostly related to cell cycle. Yap is transcriptionally inactive when phosphorylated, and non-phosphorylated YAP recruits TEAD factors to response elements. YAP and TEAD factors were not among the top 20 activated and inactivated regulators, but were up- and downregulated consistent with the increased nuclear Yap in TO. (D) The mRNA levels of genes known to be regulated by Yap were consistent with the increased nuclear Yap in TO lung specimens. (E) Assessment of lung cell-specific mRNAs showing increased basal cell (1.17), club/goblet cell (1.15), AT2 (1.21), and myeloid cell (1.45) mRNAs in TO compared to control lungs. (2). In rabbit models of CDH and TO, TO enhanced lung mouse TO models. Furthermore, TGF-β has been consistently perfusion (37) and reduced arterial wall thickness after surgical reported as increased after TO in rabbit and ovine models (7, diaphragmatic hernia (38). The potential impact of TO in airway 9, 10), and our study shows this to be true in the mouse model smooth muscle hypertrophy and lung fibroblast populations as well. While caution is required when making cross-species merits further investigation, which is now more achievable with comparisons (9), consistent findings across multiple studies Frontiers in Pediatrics | www.frontiersin.org 9 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells FIGURE 5 | Basal cells are present in the distal lung of infants with congenital high airway obstruction (CHAOS). (A) Staining was performed on lung specimens from infants of gestational age 34–36 weeks who died within 48 h of birth with CHAOS, being a human analog to mouse tracheal occlusion. The lungs of both CHAOS and control infants had keratin-5-positive cells (green arrows) beneath the pseudostratified epithelium of the large conducting airways. (B) The subpleural lung of control specimens had no keratin-5-positive cells. (C) CHAOS lung had regions with clusters of keratin-5-positive cells. None of these cells were Yap-positive (red arrows), although cells containing both cytoplasmic and nuclear Yap were observed near keratin-5-positive cells. (D) Another CHAOS lung specimen showing clusters of keratin-5-positive cells just below the pleura. (E) Specimen of an infant with congenital diaphragmatic hernia treated with fetal endoscopic tracheal occlusion who died shortly after birth. In airways just below the pleura, clusters of keratin-5-positive basal cells can be identified. (A′ ), Magnified Image of A; (C′ ), Magnified Image of C-box 1; (C′′ ), Magnified Image of C-box 2; (D′ ), Magnified Image of D; (E′ ), Magnified Image of E. should provide confidence that the observed alterations are likely errors due to differences in tissue sections, sectioning plane, operative in humans. and imaging plane. The staining of all sections together and Several study limitations are noted. For the quantification imaging, thresholding, and quantification using identical settings of the lung cell populations in fetal rabbit lungs, quantitative somewhat mitigate these concerns, and the correlation of image image analysis of lung sections is subject to measurement analysis and mRNA data provides confidence with regard to Frontiers in Pediatrics | www.frontiersin.org 10 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells the relative changes in cell populations between experimental ETHICS STATEMENT groups. However, the comparison of the relative abundance of one cell population to another should be made with caution. The animal study was reviewed and approved by Cincinnati We quantified only the relative percentage of basal cells in TO Children’s Hospital Medical Center, IACUC 2019-0034. lungs. This is relevant because hypercellularity is known to occur following TO (9, 12). We did not determine whether there were increased quantities of basal cell-associated mRNAs and AUTHOR CONTRIBUTIONS proteins in TO basal cells, or if this increased expression was due solely to an increased number of basal cells. We did not explore VS, CS, RJ, EA, and BV contributed to the conception and design the impact of TO on mesenchymal or immune cells, although of the study. EA, MO, MM, and JP all developed the assays and these signatures were more highly altered than that of Yap and surgical techniques and identified clinical specimens required basal cells. Of particular note, we did not deeply investigate or for the study. BV wrote the first draft of the manuscript. All differentiate changes in gene expression that may have arisen authors contributed to manuscript revision, read, and approved from the pulmonary vasculature, which is known to be impacted the submitted version. by both CDH and TO (38). Finally, we could not obtain any lung tissues to directly assess basal cells in post-TO human lung, FUNDING but we were able to secure a small number of specimens from infants with congenital occlusion of the trachea. Development of Funding was obtained from NIH/NHLBI R01HL141229 (to BV). a biobank for CDH and FETO-treated lung specimens of expired infants would greatly enhance the translatability of animal work being done in this area. ACKNOWLEDGMENTS In summary, we have shown that Yap mediates the expansion of the basal cell population in mouse and human lungs in We would like to acknowledge the Confocal Imaging Core and response to tracheal occlusion. Research Flow Cytometry Core at Cincinnati Children’s Hospital Medical Center for technical assistance. DATA AVAILABILITY STATEMENT The datasets presented in this study can be found in online SUPPLEMENTARY MATERIAL repositories. The names of the repository/repositories and accession number(s) can be found at: GSE184865, https://www- The Supplementary Material for this article can be found ncbi-nlm-nih-gov.ezproxy.u-pec.fr/geo/query/acc.cgi?acc= online at: https://www.frontiersin.org/articles/10.3389/fped. GSE184865. 2021.780166/full#supplementary-material REFERENCES experimental congenital diaphragmatic hernia? Am J Physiol Lung Cell Mol Physiol. (2016) 310:L311–27. doi: 10.1152/ajplung.00122.2015 1. Al-Maary J, Eastwood MP, Russo FM, Deprest JA, Keijzer R. Fetal 8. Engels AC, Brady PD, Kammoun M, Finalet Ferreiro J, DeKoninck P, Endo Tracheal occlusion for severe pulmonary hypoplasia in isolated congenital M, et al. Pulmonary transcriptome analysis in the surgically induced rabbit diaphragmatic hernia: a systematic review and meta-analysis of survival. Ann model of diaphragmatic hernia treated with fetal tracheal occlusion. DMM. Surg. (2016) 264:929–33. doi: 10.1097/SLA.0000000000001675 (2016) 9:221–8. doi: 10.1242/dmm.021626 2. Coughlin MA, Werner NL, Gajarski R, Gadepalli S, Hirschl R, Barks J, et al. 9. Varisco BM, Sbragia L, Chen J, Scorletti F, Joshi R, Wong HR, et al. Excessive Prenatally diagnosed severe CDH: mortality and morbidity remain high. J reversal of epidermal growth factor receptor and ephrin signaling following Pediatr Surg. (2015) 82. tracheal occlusion in rabbit model of congenital diaphragmatic hernia. Mol 3. Harrison MR, Jester JA, Ross NA. Correction of congenital diaphragmatic Med. (2016) 22. doi: 10.2119/molmed.2016.00121 hernia in utero. I The model: intrathoracic balloon produces fatal pulmonary 10. Peiro JL, Oria M, Aydin E, Joshi R, Cabanas N, Schmidt R, et al. Proteomic hypoplasia. Surgery. (1980) 88:174–82. profiling of tracheal fluid in an ovine model of congenital diaphragmatic 4. Boucherat O, Benachi A, Chailley-Heu B, Franco-Montoya ML, hernia and fetal tracheal occlusion. Am J Physiol Lung Cell Mol Physiol. (2018) Elie C, Martinovic J, et al. Surfactant maturation is not delayed 315:L1028–41. doi: 10.1152/ajplung.00148.2018 in human fetuses with diaphragmatic hernia. PLoS Med. (2007) 11. Dobrinskikh E, Al-Juboori SI, Oria M, Reisz JA, Zheng C, Peiro JL, et al. 4:e237. doi: 10.1371/journal.pmed.0040237 Heterogeneous response in rabbit fetal diaphragmatic hernia lungs after 5. Chapin CJ, Ertsey R, Yoshizawa J, Hara A, Sbragia L, Greer JJ, et al. Congenital tracheal occlusion. J Surg Res. (2019) 250:23–38. doi: 10.1016/j.jss.2019.12.025 diaphragmatic hernia, tracheal occlusion, thyroid transcription factor-1, and 12. Aydin E, Joshi R, Oria M, Varisco BM, Lim FY, Peiro JL. Fetal tracheal fetal pulmonary epithelial maturation. Am J Physiol Lung Cell Mol Physiol. occlusion in mice: a novel transuterine method. J Surg Res. (2018) 229:311– (2005) 289:L44–52. doi: 10.1152/ajplung.00342.2004 5. doi: 10.1016/j.jss.2018.04.028 6. Davey MG, Biard JM, Robinson L, Tsai J, Schwarz U, Danzer E, et al. Surfactant 13. Aydin E, Joshi R, Oria M, Lim F-Y, Varisco BM, Peiro JL. Transuterine fetal protein expression is increased in the ipsilateral but not contralateral lungs tracheal occlusion model in mice. JoVE. (2021). doi: 10.3791/61772-v of fetal sheep with left-sided diaphragmatic hernia. Pediatr Pulmonol. (2005) 14. Lange AW, Sridharan A, Xu Y, Stripp BR, Perl AK, Whitsett JA. 39:359–67. doi: 10.1002/ppul.20175 Hippo/Yap signaling controls epithelial progenitor cell proliferation and 7. Vuckovic A, Herber-Jonat S, Flemmer AW, Ruehl IM, Votino C, Segers V, differentiation in the embryonic and adult lung. J Mol Cell Biol. (2015) et al. Increased TGF-beta: a drawback of tracheal occlusion in human and 7:35–47. doi: 10.1093/jmcb/mju046 Frontiers in Pediatrics | www.frontiersin.org 11 February 2022 | Volume 9 | Article 780166 Serapiglia et al. FETO and Basal Cells 15. Mahoney JE, Mori M, Szymaniak AD, Varelas X, Cardoso 31. Vaughan AE, Brumwell AN, Gotts Xi Y, Brownfield JE, Treutlein DG, Tan B, WV. The hippo pathway effector Yap controls patterning and et al. HA. (2015). Lineage-negative progenitors mobilize to regenerate lung differentiation of airway epithelial progenitors. Dev Cell. (2014) epithelium after major injury. Nature. 517:621–5. doi: 10.1038/nature14112 30:137–50. doi: 10.1016/j.devcel.2014.06.003 32. Zuo W, Rostami MR, Shenoy SA, LeBlanc MG, Salit J, Strulovici- 16. Nantie LB, Young RE, Paltzer WG, Zhang Y, Johnson RL, Verheyden Barel Y, et al. Cell-specific expression of lung disease risk-related JM, et al. Lats inactivation reveals hippo function in alveolar type I cell genes in the human small airway epithelium. Respir Res. (2020) differentiation during lung transition to air breathing. Development. (2018) 21:200. doi: 10.1186/s12931-020-01442-9 145:dev163105. doi: 10.1242/dev.163105 33. Aydin E, Torlak N, Yildirim A, Bozkurt EG. Reversible 17. Soldt BJ, van Qian J, Li J, Tang N, Lu J, Cardoso WV. Yap and its subcellular fetal tracheal occlusion in mice: a novel minimal invasive localization have distinct compartment-specific roles in the developing lung. technique. J Surg Res. (2021) 260:278–83. doi: 10.1016/j.jss.2020.1 Development. (2019) 146:dev175810. doi: 10.1242/dev.175810 1.080 18. Joshi R, Liu S, Brown MD, Young SM, Batie M, Kofron JM, et al. Stretch 34. Tan SY, Krasnow MA. Developmental origin of lung macrophage diversity. regulates expression and binding of chymotrypsin-like elastase 1 in the Development. (2016) 143:1318–27. doi: 10.1242/dev.129122 postnatal lung. FASEB. (2016) 30:590–600. doi: 10.1096/fj.15-277350 35. Endale M, Ahlfeld S, Bao E, Chen X, Green J, Bess Z, et al. Temporal, spatial, 19. Du Y, Guo M, Whitsett JA, Xu Y. “LungGENS”: a web-based tool for and phenotypical changes of PDGFRalpha expressing fibroblasts during late mapping single-cell gene expression in the developing lung. Thorax. 70:109–4. lung development. Dev Biol. (2017) 425:161–75. doi: 10.1016/j.ydbio.2017.0 (2015). doi: 10.1136./thoraxjnl-2015-207035 3.020 20. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. 36. Riccetti M, Gokey JJ, Aronow B, Perl A-KT. The elephant in the lung: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. (2013) 29:15– Integrating lineage-tracing, molecular markers, and single cell sequencing 21. doi: 10.1093/bioinformatics/bts635 data to identify distinct fibroblast populations during lung development and 21. Love MI, Huber W, Anders S. Moderated estimation of fold change regeneration. Matrix Biol. (2020) 91–2:51–74. doi: 10.1016/j.matbio.2020.0 and dispersion for RNA-seq data with DESeq2. Genome Biol. (2014) 5.002 15:550. doi: 10.1186/s13059-014-0550-8 37. Cruz-Martinez R, Moreno-Alvarez O, Prat J, Krauel L, Tarrado X, Castanon 22. Chen J, Bardes EE, Aronow BJ, Jegga AG. ToppGene Suite for gene list M, et al. Lung tissue blood perfusion changes induced by in utero tracheal enrichment analysis and candidate gene prioritization. Nucleic Acids Res. occlusion in a rabbit model of congenital diaphragmatic hernia. Fetal Diagn (2009) 37:W305–11. doi: 10.1093/nar/gkp427 Ther. (2009) 26:137–42. doi: 10.1159/000254485 23. Krämer A, Green J, Pollard J, Tugendreich S. Causal analysis 38. Prat Ortells J, Albert A, Tarrado X, Krauel L, Cruz R, Moreno-Alvarez O, approaches in Ingenuity Pathway Analysis. Bioinformatics. (2014) et al. Airway and vascular maturation stimulated by tracheal occlusion do 30:523–30. doi: 10.1093/bioinformatics/btt703 not correlate in the rabbit model of diaphragmatic hernia. Pediatr Res. (2014) 24. R Core Team. R: A Language and Environment for Statistical Computing. 75:487–92. doi: 10.1038/pr.2013.244 Vienna, Austria: R Foundation for Statistical Computing (2019). 25. Kassambara, A. ggpubr: Publication Ready Plots-Articles-STHDA (2020). Conflict of Interest: The authors declare that the research was conducted in the Available online at: http://www.sthda.com/english/articles/24-ggpubr- absence of any commercial or financial relationships that could be construed as a publication-ready-plots/ potential conflict of interest. 26. Kassambara A. rstatix: Pipe-Friendly Framework for Basic Statistical Tests (2020). Publisher’s Note: All claims expressed in this article are solely those of the authors 27. Meng Z, Moroishi T, Guan KL. Mechanisms of Hippo pathway regulation. and do not necessarily represent those of their affiliated organizations, or those of Genes Dev. (2016) 30:1–17. doi: 10.1101/gad.274027.115 the publisher, the editors and the reviewers. Any product that may be evaluated in 28. Yang Y, Riccio P, Schotsaert M, Mori M, Lu J, Lee DK, et al. Spatial- this article, or claim that may be made by its manufacturer, is not guaranteed or temporal lineage restrictions of embryonic p63(+) progenitors establish distinct stem cell pools in adult airways. Dev Cell. (2018) 44:752– endorsed by the publisher. 61e4. doi: 10.1016/j.devcel.2018.03.001 29. Penkala IJ, Liberti DC, Pankin J, Sivakumar A, Kremp MM, Copyright © 2022 Serapiglia, Stephens, Joshi, Aydin, Oria, Marotta, Peiro Jayachandran S, et al. Age-dependent alveolar epithelial plasticity and Varisco. This is an open-access article distributed under the terms of orchestrates lung homeostasis and regeneration. Cell Stem Cell. (2021) the Creative Commons Attribution License (CC BY). The use, distribution 28:1775–89.e5. doi: 10.1016/j.stem.2021.04.026 or reproduction in other forums is permitted, provided the original author(s) 30. Nguyen TM, van der Merwe J, Elowsson Rendin L, Larsson-Callerfelt A-K, and the copyright owner(s) are credited and that the original publication in Deprest J, et al. Stretch increases alveolar type 1 cell number in fetal lungs this journal is cited, in accordance with accepted academic practice. No use, through ROCK-Yap/Taz pathway. Am J Physiol Lung Cell Mol Physiol. (2021) distribution or reproduction is permitted which does not comply with these 321:L814–26. doi: 10.1152/ajplung.00484.2020 terms. Frontiers in Pediatrics | www.frontiersin.org 12 February 2022 | Volume 9 | Article 780166