Epithelial Wntless regulates postnatal alveologenesis

Efficient gas exchange requires the coordinated development of the alveolar epithelium and mesenchyme during the neonatal stage (Hogan et al., 2014; Mund et al., 2008; Schittny et al., 2008; Yang et al., 2016). For example, mesenchymal cells differentiate into myofibroblasts (α-SMA⁺), which synergize with epithelial differentiation and subdivide saccules into alveoli (Boström et al., 1996; Chao et al., 2016; Lindahl et al., 1997). Meanwhile, myofibroblasts produce the elastin-based extracellular matrix (ECM) to provide a complex and net-like structure for alveolar formation. The presence of myofibroblasts in interstitial regions is transient, present at around postnatal (P) day 3 to P14 in mice (McGowan et al., 2008). The persistent presence of myofibroblasts during the alveologenesis stage has been found in individuals with bronchopulmonary dysplasia (BPD), which is characterized by alveolar simplification (Hilgendorff and O'Reilly, 2015; Husain et al., 1998; Jobe and Bancalari, 2001; Northway et al., 1967). In line with this, a hyperoxia-induced mouse model of BPD shows enrichment of myofibroblasts in the enlarged air sac post-alveologenesis (Branchfield et al., 2016). The epithelium seems to play an important role in the control of myofibroblasts during alveologenesis (Bourbon et al., 2005; Zhang et al., 2020). Consistently, loss of Etv4/5 in epithelial cells leads to epithelial defects and increased expression of α-SMA in the simplified alveoli (Branchfield et al., 2016). However, it remains unknown what signal(s) from the epithelium are needed for regulating the transient presence of myofibroblasts.

Multiple signaling pathways, including the Wnt pathway, have been shown to regulate lung development (Bellusci et al., 1996; Colvin et al., 2001; Goss et al., 2009; Harris-Johnson et al., 2009; Kishimoto et al., 2020; Mucenski et al., 2003; Shu et al., 2005; Snowball et al., 2015; Yin et al., 2008). Ablation of Wnt2/2b or the signaling mediator β-catenin results in the loss of lung progenitors and complete lung agenesis (Goss et al., 2009; Harris-Johnson et al., 2009). Moreover, overexpression of the Wnt inhibitor Dkk1 leads to disrupted proximal-distal patterning with proximalized lung epithelium in the distal domain (Shu et al., 2005). More recently, the protein Wntless homolog (Wls), a chaperone protein responsible for the transport and secretion of Wnt ligands, has been shown to regulate lung branching morphogenesis, epithelial differentiation and vascular development (Cornett et al., 2013; Jiang et al., 2013). However, in both studies, Shh-Cre was used to delete Wls and mutants die at birth. Therefore, the function of Wls in postnatal lung development remains unclear.

Here, we use Sftpc-Cre and Nkx2.1-Cre mouse lines to delete Wls and observed BPD-like emphysema with severely enlarged saccules. Although epithelial commitment is not affected, extensive myofibroblasts persist in the adult mutant lungs with the presence of α-SMA⁺ endothelial cells. These mutants also display more severe pulmonary fibrosis than the controls following bleomycin challenge. Although neonatal deletion of Wls in the alveolar type II (AT2) cells with Sftpc-CreER does not impact homeostasis or exacerbate bleomycin-induced fibrosis, Wls ablation using Ager-CreER leads to the extensive presence of α-SMA⁺ endothelial cells.

Shh-Cre activities are detected in the early foregut at approximately embryonic day (E) 8.75 (Harris-Johnson et al., 2009; Rodriguez et al., 2010), before lung budding. The resultant Shh-Cre; Wls^loxp/loxp mutants succumbed to death at P0 (Jiang et al., 2013). We then used the transgenic Sftpc-Cre mouse line, which is active at ∼E10.5 (Okubo and Hogan, 2004). Unlike Shh-Cre; Wls^loxp/loxp mutants, Sftpc-Cre; Wls^loxp/loxp mutants are viable and survive to adulthood. As expected, deletion of epithelial Wls leads to reduced Wnt signaling activities, as evidenced by the Axin2-LacZ reporter allele at P30 (Fig. S1A). Histological examination reveals progressively enlarged air sac, as observed in individuals with BPD and mouse models (Fig. 1A) (Hilgendorff and O'Reilly, 2015; Jobe and Bancalari, 2001). The mutant lungs exhibit normal alveologenesis when examined at P0. However, air sac enlargement becomes apparent at P10 and more severe at P30, although proliferation and apoptosis are comparable with littermate controls (Fig. S1B-E). The number of alveolar saccules decreases from 209.67±4.33 to 103.5±5.9 per mm² (P<0.0001) at P10, and 226.67±3.71 to 73±5.58 per mm² (P<0.0001) at P30 (Fig. 1B). Conversely, the mean linear intercept (MLI) increases from 31.41±0.43 µm to 51.79±2.26 µm (P<0.001) at P10 and from 26.83±1.08 µm to 62.86±2.51 µm (P<0.0001) at P30 (Fig. 1C). These emphysema-like phenotypes remain in the adult mutant lungs (Fig. 1D). Similar enlarged air sacs are also present in the lungs of Nkx2.1-Cre; Wls^loxp/loxp mutants at P30 and P60 (Fig. 1E-H). Notably, the activities of the transgenic Nkx2.1-Cre can be detected from E10.5 onwards in the embryonic lungs (Tiozzo et al., 2009). Previous studies show that deletion of Wnt5a leads to the thickening of the epithelium and mesenchyme in the embryonic lungs (Li et al., 2002). Moreover, Wnt7b is enriched in the lung epithelium (Shu et al., 2002), and loss of Wnt7b causes severe lung hypoplasia with reduced proliferation of the epithelium and mesenchyme (Rajagopal et al., 2008). We did not observe lung hypoplasia in Sftpc-Cre; Wls^loxp/loxp and Nkx2.1-Cre; Wls^loxp/loxp embryos, suggesting that Wnt7b likely exerts its function earlier than E10.5. In light of the distinctive phenotypes observed in the lungs of Shh-Cre; Wls^loxp/loxp and Sftpc-Cre/Nkx2.1-Cre; Wls^loxp/loxp mutants, Wls likely plays a different role before and after E10.5 during lung development.

Deletion of Wls with Shh-Cre alters the differentiation of lung progenitor cells (Cornett et al., 2013; Jiang et al., 2013). However, differentiation of airway progenitor cells seems unaffected in Sftpc-Cre; Wls^loxp/loxp mutants (Fig. S2). Similar to the controls, Sox2 is expressed in the airway epithelium (Fig. S2A), and the ratio of ciliated cells (Foxj1⁺), club cells (Scgb1a1⁺) and neuroendocrine cells (Syp⁺) is comparable with controls (Fig. S2A-D). However, the numbers of AT2 cells per alveolar saccule are increased in Sftpc-Cre; Wls^loxp/loxp mutants at P30 (Fig. S2E), although the ratio of AT2 cells to alveolar Nkx2.1⁺ cells is unchanged (81.49±1.49% versus 84.95±0.60%, P=0.0605) (Fig. S2F). The transcript levels of Sftpc are also similar between mutants and littermate controls (Fig. S2G). In addition, AT1 cells (Pdpn⁺) are normally distributed in the alveolar sacs of Sftpc-Cre; Wls^loxp/loxp mutants (Fig. S2C). These data indicate that Wls is required for epithelial differentiation prior to E10.5, and deletion of epithelial Wls after E10.5 has a limited impact on epithelial differentiation. Considering these distinct phenotypes, E9.5-E10.5 represents a unique stage in mouse lung development.

The transient presence of myofibroblasts is required for the formation of the secondary septae during neonatal alveologenesis (Branchfield et al., 2016). α-SMA⁺ myofibroblasts are rarely detected in the alveoli at P0. However, at P10 a significant number of myofibroblasts are present in the peripheral lungs, and these α-SMA⁺ cells are reduced to a minimal level at P30 (Fig. S3A). Remarkably, numerous α-SMA⁺ cells remain to form thick air sac walls in the alveoli of Sftpc-Cre; Wls^loxp/loxp and Nkx2.1-Cre; Wls^loxp/loxp mutants at P30 (Fig. 2A-D and Fig. S3A-D). Similar to the heterogeneous lung pathology observed in individuals with BPD (Thebaud et al., 2019), extensive numbers of Nkx2.1⁺ and SPC⁺ cells mixed with a few myofibroblasts are also found in the high cell-density areas (Fig. S3D). The increased numbers of α-SMA⁺ cells in Sftpc-Cre; Wls^loxp/loxp mutants were confirmed by flow cytometry (FACS) analysis (8.76%±0.69% versus 26.17%±1.78%, P<0.001) (Fig. 2E,F and Fig. S4A). Consistently, the transcript and protein levels of α-SMA are increased in Sftpc-Cre; Wls^loxp/loxp mutants at P30 (Fig. 2G,H). Persistent α-SMA⁺ myofibroblasts produce and deposit ECM proteins, including collagen and fibronectin in the alveoli, as evidenced by Masson's trichrome staining (Fig. 2I). The transcript levels of Col1a1 and Col3a1 are also increased in the mutant lungs (Fig. 2J). Taken together, these data support that deletion of epithelial Wls results in the persistent presence of α-SMA⁺ mesenchymal cells with increased collagen deposition within the alveoli.

Shh-Cre; Wls^loxp/loxp mutants show severe pulmonary hemorrhage due to abnormal vasculature (Cornett et al., 2013; Jiang et al., 2013). By contrast, pulmonary vasculature development seems normal in Sftpc-Cre; Wls^loxp/loxp mutants, as determined by the endothelial marker Endomucin at P10 (Fig. 3A). However, at P30 the ratio of lung endothelial cells among mesenchymal cells (CD45⁻EpCAM⁻) is significantly decreased in Sftpc-Cre; Wls^loxp/loxp mutants (51.53±0.32% versus 23.23±0.77%, P<0.0001) (Fig. 3B,C and Fig. S4B). Surprisingly, 9.44±0.27% endothelial cells co-express α-SMA in the alveoli, which were confirmed by immunostaining with the endothelial markers Pecam1 (CD31) and ERG (Fig. 3D,E). FACS analysis further reveals a dramatic increase of Pecam1⁺α-SMA⁺ population in mutant lungs (0.14±0.01% versus 1.34±0.09%, P<0.001) (Fig. 3F,G and Fig. S4C). In addition, the transcript levels of the myofibroblast markers Acta2 and Tagln are increased by about sevenfold in the isolated endothelial cells, which also show a fourfold increase of Col1a1 and Col3a1 (Fig. 3H). Reduced pulmonary microvessels have been reported in the enlarged air sac of BPD lungs (Bhatt et al., 2001). Sftpc-Cre; Wls^loxp/loxp mutant lungs also demonstrate enlarged air sacs with reduced numbers of endothelial cells. Interestingly, endothelial cells in the arteries of individuals with pulmonary hypertension exhibit endothelial to mesenchymal transition (EndMT), co-expressing α-SMA and endothelial markers Pecam1 and vWF (Good et al., 2015; Ranchoux et al., 2015). We also observed extensive Pecam1⁺α-SMA⁺ endothelial cells in the mutant lungs. Hyperoxia exposure has been shown to reduce Pecam1 expression while promoting α-SMA expression in cultured human pulmonary microvascular endothelial cells (Zhang et al., 2018). Although it remains unknown whether Wnt signaling is involved in hyperoxia-associated EndMT, our data support that Wnts from the epithelium are crucial for endothelial differentiation in the developing lung.

To study whether Wls continues to play important roles in alveologenesis during early postnatal stage, Sftpc-CreER; Wls^loxp/loxp and Ager-CreER; Wls^loxp/loxp mice were administrated with tamoxifen at P2 and P10 to delete Wls in AT2 and AT1 cells, respectively (Figs S5A and S6A). The Sftpc-CreER mouse line targets AT2 cells efficiently as ∼99% of AT2 cells are labeled by the Rosa26-tdtomato reporter allele after two doses of tamoxifen (Fig. S6B,C). Epithelial differentiation seems unperturbed in mutants with normal histology and differentiation of club, ciliated, AT1 and AT2 cells when examined at P30 (Fig. S5B,C and S6D,E). Although alveolar endothelial cells and myofibroblasts seem unaffected in Sftpc-CreER; Wls^loxp/loxp mutants (Fig. S6E), extensive α-SMA⁺ endothelial cells are present in the alveoli of Ager-CreER; Wls^loxp/loxp mutants (Fig. 3I). A recent study indicates that the Ager-CreER mouse line from the Jackson Laboratory targets the majority of AT1 and some AT2 cells (Penkala et al., 2021). Given normal endothelial development in Sftpc-CreER; Wls^loxp/loxp mutants, our results suggest that Wnts from AT1 cells are required for proper endothelial differentiation at the neonatal stage. Of note, deletion of Wls in AT1 cells at E15.5 has minimal effects on alveologenesis when examined at P5 (Zepp et al., 2021), suggesting that Wnts from AT1 cells are needed for mesenchymal differentiation at later stages of alveologenesis.

Neonatal hyperoxia causes permanent lung damage that renders the affected individuals more susceptible to environmental challenges (O'Reilly et al., 2008). Given that loss of epithelial Wls leads to persistent α-SMA⁺ mesenchymal cells, we asked whether Sftpc-Cre; Wls^loxp/loxp mutants display more severe pulmonary fibrosis following bleomycin challenge. Remarkably, fibrotic areas are significantly enlarged in mutants when compared with controls when examined at day14 after bleomycin treatment (Fig. 4A,B). Masson's trichrome staining reveals extensive collagen deposition in the interstitial space of mutant lungs (Fig. 4C). Consistent with this, α-SMA⁺ cells are enriched in the fibrotic foci of mutant lungs (Fig. 4D,E). These results suggest that Sftpc-Cre; Wls^loxp/loxp mutants are more susceptible to bleomycin challenge and that the persistent presence of alveolar α-SMA⁺ mesenchymal cells may serve as a source of fibrotic myofibroblasts during pulmonary fibrosis.

Wnt signaling plays a crucial role in adult lung regeneration (Frank et al., 2016; Hogan et al., 2014; Kim et al., 2019; Nabhan et al., 2018; Tanjore et al., 2013; Zacharias et al., 2018). We therefore asked whether deletion of Wls in AT2 cells affects lung regeneration in Sftpc-CreER; Wls^loxp/loxp mice. We challenged the adult mutants with bleomycin after Wls ablation (Fig. S6F). Deletion of Wls prior to bleomycin treatment has minimal effects on lung regeneration, and the lung fibrotic area is comparable with controls (Fig. S6G,H). Masson's trichrome staining reveals a similar collagen deposition in the interstitial spaces (Fig. S6I). In addition, the numbers of myofibroblasts in the peripheral lungs are similar in mutants and controls (Fig. S6J,K). These findings are in contrast with the phenotypes observed in Sftpc-Cre; Wls^loxp/loxp mutants, suggesting that defective lungs beginning at an early postnatal state underlies worsening regeneration and fibrosis.

In conclusion, we demonstrate the essential role of epithelial Wnts in the formation of alveoli during the neonatal stage. Loss of Wls leads to abnormal differentiation of mesenchymal cells, resulting in the persistent presence of myofibroblasts and ectopic expression of α-SMA⁺ in the endothelial cells lining the enlarged air-sac. Our studies further suggest that Wnts produced by AT1 but not AT2 cells regulate endothelial differentiation at the early postnatal stage. This is consistent with the close proximity of AT1 cells and endothelium in the alveoli (Vila Ellis et al., 2020), and growth factors secreted by AT1 cells likely have direct impacts on pulmonary vasculature. Finally, in light of distinct phenotypes seen in Shh-Cre;Wls^loxp/loxp and Sftpc-Cre; Wls^loxp/loxp mutants, E9.5-E10.5 may represent a unique stage in mouse lung development.

The Sftpc-Cre (Okubo and Hogan, 2004), Nkx2.1-Cre (Xu et al., 2008), Sftpc-CreER (Rock et al., 2011), Ager-CreER (Jackson Laboratory 032771), Axin2-lacZ (Lustig et al., 2002) and Wls^loxp/loxp (Fu et al., 2011) mouse strains have been reported previously. For postnatal alveologenesis analysis, Sftpc-Cre;Wls^loxp/loxp, Nkx2.1-Cre;Wls^loxp/loxp, Sftpc-CreER;Wls^loxp/loxp and Ager-CreER;Wls^loxp/loxp mice are on a C57BL/6 and 129/SvEv mixture background. For bleomycin treatment, Sftpc-Cre;Wls^loxp/loxp and Sftpc-CreER;Wls^loxp/loxp mice were backcrossed with C57BL/6 breeders for at least six generations. All mouse experiments and care were conducted in accordance with procedures approved by the Institutional Animal Care and Use Committee at Columbia University.

Lung tissues were isolated after the mice were euthanized with isoflurane and fixed in 4% paraformaldehyde (PFA) overnight. Thick lung sections were made as previously described (Branchfield et al., 2016). Briefly, 4% low-melt agarose was used for embedding whole lung lobes. 75 µm sections were cut using a Leica vibratome. For paraffin section preparation, the tissues were dehydrated with ethanol and cleared in a histoclear solution, and then embedded in paraffin wax. 7 µm sections were cut and mounted on the slides for further histology staining and immunostaining. Hematoxylin and Eosin (H&E) staining was performed as previously described (Fang et al., 2020; Jiang et al., 2021). Masson's trichrome staining was performed by using commercial staining kit (Sigma, H15-1KT) according to the manufacturer's instruction. Nikon SMZ1500 Inverted microscope was used for obtaining images.

Immunofluorescence staining was performed as previously described (Jiang et al., 2017). Briefly, the sections were deparaffinized and rehydrated. Following antigen retrieval with high-pressure heating in antigen unmasking solution (Vector Laboratory, H-3300) the sections were blocked with 5% normal donkey serum (Jackson ImmunoResearch Laboratories, 017-000-121) and then incubated with primary antibodies anti-Sox2 (Invitrogen, 14-9811-82, 1:200), anti-Foxj1 (Invitrogen, 14-9965-82, 1:100), anti-CC10 (Santa Cruz, sc-365992, 1:200), anti-SPC (Abcam, ab211326, 1:500), anti-Pdpn (DSHB, 8.1.1-c, 1:500), anti-αSMA (Sigma, A2547, 1:200), anti-SM22α (Abcam, ab14106, 1:200), anti-ERG (Abcam, ab92513, 1:200), anti-tdTomato (Biorbyt, orb182397, 1:1000), anti-Hopx (Santa Cruz, sc-398703, 1:100), anti-Endomucin (Santa Cruz, sc-65495, 1:200), anti-CD31 (BD, 550274, 1:100), anti-synaptophysin (Abcam, ab32127, 1:200), anti-Ki67 (Cell Signaling Technology, 9129S, 1:200) and anti-TTF1 (Nkx2.1, Abcam, ab76013, 1:200) at 4°C overnight. Secondary antibodies (Jackson ImmunoResearch Laboratories) were added to the sections followed by extensive washing with PBS and incubation for 2 h at room temperature. For TUNEL staining, In Situ Cell Death Detection Kit (Roche, 11684795910) was used according to the instructions. The nuclei were counterstained with DAPI, and images were obtained with Zeiss LSM T-PMT confocal laser-scanning microscope and analyzed by Zeiss software. For three-dimensional reconstructions, IMARIS software was used as previously described (Branchfield et al., 2016).

Single cell suspension was obtained from lung tissues as previously described (Dong et al., 2015), and the cells were incubated with fluorophore-conjugated antibodies CD45 (Biolegend, 103114, 1:100), EpCAM (Biolegend, 118212, 1:100) and CD31 (Biolegend, 102449, 1:100) in FACS buffer (5% FBS with 0.5 mM EDTA in PBS) for 1 h at 4°C, followed by incubation with the live/dead kit (Invitrogen, L34955) for 10 min at room temperature to exclude dead cells. To obtain lung endothelial cells, live CD45⁻EpCAM⁻CD31⁺ cells are gated and sorted. For FACs analysis, a Foxp3/Transcription Factor Staining Buffer kit (eBioscience, 00-5523-00) was used for intracellular staining according to the instructions. The cells were then incubated with fluorophore-conjugated antibody against α-SMA (Invitrogen, 53-9760-82, 1:100 dilution) for 1 h at room temperature. Data were obtained by using BD LSRII and analyzed with FlowJo software. Live CD45⁻EpCAM⁻ mesenchymal cells were gated for further analysis.

X-gal staining was performed as previously described (Rodriguez et al., 2010). Briefly, lung tissues were dissociated and fixed with 4% PFA for 30 min at room temperature, and then the lungs were inflated with X-gal staining solution, including 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 50 mM MgCl₂, 0.5% NP-40 and 1 mg/ml X-gal, and incubated at 37°C overnight. Lung tissues were post-fixed with 4% PFA for 2 h at room temperature, and dehydrated and embedded in paraffin wax for further sectioning, as previously described (Que et al., 2007).

Western blotting analysis was performed as previously described (Jiang et al., 2013). Briefly, lung tissues were homogenized in Tris buffer with the presence of the proteinase inhibitors cocktail (TissueLyser II). SDS-PAGE was used for separating mixed proteins that were then transferred onto a polyvinylidene fluoride (PVDF) transfer membrane. Following blocking with 5% fat-free milk, the membrane was incubated in the primary antibodies anti-α-SMA (Sigma, 1:1000) and anti-GAPDH (ProteinTech, 1:5000) overnight at 4°C. The HRP-conjugated secondary antibody was used for assessing the expression of proteins.

Lung tissues were homogenized in Trizol buffer using TissueLyser II. Supernatants were obtained for total RNA extraction following centrifuging at 13,800 g for 10 min at 4°C. Total RNA was extracted according to instruction of commercial RNA extraction kit. The first-strand cDNA was synthesized from RNA by using SuperScript III First-Strand SuperMix kit. SYBR Green Supermix was used to quantify the cDNA by StepOnePlus Real-Time PCR System. The gene expression level was calculated relative to mouse β-actin by using the 2^−Δ(ΔCT) method. The sequences of qRT-PCR primers used in this study are described in Table S1.

Tamoxifen was purchased from Sigma-Aldrich and dissolved in sunflower seed oil to 200 mg/ml for stock solution. The mice were intraperitoneally injected with tamoxifen at a dose of 200 mg/kg every other day for total three doses. A phase period of 10 days was used to wash out the residual tamoxifen before any further treatments. For bleomycin treatment, the 6- to 8-week-old C57BL/6 background mice were anaesthetized by intraperitoneally injecting with tribromoethanol (Avertin) at a dose of 250 mg/kg. Under anesthesia, the neck skin was opened to expose the trachea by blunt dissection, and bleomycin (Fresenius Kabi, USP) was delivered at a dose of 1.75 unit/kg body weight through intratracheal injection with a 30-Gauge needle. Samples were harvested for analysis at 14 days post-bleomycin treatment.

Measurement of mean linear intercept (MLI) was performed as previously described (Liu et al., 2017). For quantification of different types of epithelial cells and α-SMA⁺ cells, at least five random fields (20×magnification) were captured, and positive cells were counted using ImageJ software. At least three replicates were included for each group. All data are presented as the means±s.e.m. using GraphPad Prism 8. Unpaired two-tailed Student's t-test was used to determine statistical significance. P<0.05 or less were considered statistically significant.

We thank the colleagues in the Que laboratory for critical reading of the manuscript. Flow cytometry was performed in the Columbia Center for Translational Immunology (CCTI) Flow Cytometry Core at Columbia University Medical Center, supported in part by the Office of the Director, National Institutes of Health under the awards S10RR027050 and S10OD020056.

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.199505.