Developmental dynamics of bronchioalveolar stem cells revealed by dual-reporter lineage tracing

Abstract

Bronchioalveolar stem cells (BASCs), located at the bronchioalveolar duct junction (BADJ), contribute to adult lung regeneration after injury, yet their role during postnatal lung morphogenesis remains poorly characterized. In this study, we showed that BASCs first emerged at birth and persisted throughout the postnatal saccular stage with stable localization and abundance. To trace their fate, we generated a Sftpc^DreER knock-in mouse and combined it with Scgb1a1^CreER and dual-reporter H11^{LSL − ZsG_RSR−tdT} mice to establish a dual recombinase-mediated lineage tracing system (BASC-Tracer). Using this system, we observed that BASCs expanded postnatally and migrated into both alveolar and airway compartments. In the alveoli, they lost SCGB1A1 expression and differentiated into AT2 and AT1 cells to drive alveolar maturation, whereas in the airways, they lost SFTPC expression and gave rise to club cells. Furthermore, we employed a hyperoxia-induced bronchopulmonary dysplasia model to assess the injury response of BASCs during lung development. We found that hyperoxic exposure markedly suppresses BASC regenerative capacity. Finally, spatial transcriptomic analysis revealed that BASCs reside within a distinct BADJ-associated niche, characterized by a unique repertoire of signaling ligands that distinguishes this niche from those of AT2 and club cells. Together, these findings delineate the dynamic behavior of BASCs during lung development and provide spatial characterization of their niche.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-026-03498-5.

Introduction

Lung development proceeds through a series of well-characterized morphogenetic stages, including the embryonic (embryonic day [E] 9.5–E12.5), pseudoglandular (E12.5–E16.5), canalicular (E16.5–E17.5), saccular (E17.5–postnatal day [P] 5), and alveologenesis (P5–P39) stages [1–3]. During the saccular stage in mice, terminal saccules form and the distal lung undergoes structural expansion in preparation for postnatal gas exchange [3–5]. At this stage, the bronchioalveolar duct junction (BADJ), a specialized location at the interface between alveoli and airways, is established. Bronchioalveolar stem cells (BASCs) emerge at this location [6–9]. BASCs are a distinct population of lung epithelial progenitors, defined by co-expression of surfactant protein C (SFTPC) and secretoglobin family 1 A member 1 (SCGB1A1) and have been reported to first emerge during alveologenesis [6]. However, whether BASCs emerge concomitantly with the establishment of the BADJ during the saccular stage remains unknown.

Owing to BASCs unique anatomical location and pivotal role in epithelial repair, BASCs have garnered considerable interest. In the adult lung, following alveolar injury, BASCs migrate distally from the BADJ and differentiate into AT2 cells, which function as progenitors with self-renewal capacity and the ability to generate AT1 cells, thereby restoring the gas exchange surface [7, 10]. In response to airway injury, BASCs migrate proximally and give rise to club cells, the resident bronchiolar progenitors responsible for airway epithelial regeneration [6, 7, 10]. These findings establish that BASCs possess multilineage differentiation potential in the context of adult lung injury; however, their contribution to epithelial lineage specification during lung development remains to be determined.

In this study, we observed the presence of BASCs as early as P0. To further investigate their dynamic behavior during lung development, we generated Sftpc^DreER knock-in mice and crossed them with Scgb1a1^CreER and H11^{LSL − ZsG_RSR−tdT} dual-reporter mice to establish a BASC-specific dual-recombinase lineage-tracing system. This system enables high-resolution, in vivo fate mapping of BASCs throughout lung development. Our results reveal their active contribution to both alveolar and airway epithelial lineages. Furthermore, we analyzed the proliferative responses of BASC-lin cells under hyperoxic injury. We also characterized their stromal niche by analyzing spatial transcriptomic data. Together, this study establishes a developmental framework for understanding BASC dynamics during postnatal lung morphogenesis under both physiological and perturbed conditions.

Materials and methods

Animal studies

Scgb1a1^CreER (Cat# 016225) and R26R^tdT (Cat# 007909) mice were obtained from the Jackson Laboratory, and H11^{LSL − ZsG_RSR−tdT} (Cat# NM-KI-200319) mice were obtained from Shanghai Model Organisms Center, Inc.

Generation of Sftpc^DreER knock-in mice

The Sftpc^DreER knock-in mouse line was established through CRISPR/Cas9-assisted homology-directed repair at the endogenous Sftpc locus in C57BL/6J mice. A guide RNA target site (5’ CTACTATATCTAGCATCCAC 3’), located at the stop codon of Sftpc and followed by a PAM sequence (AGG), was selected for genome editing (Ensembl ID: ENSMUSG00000022097; chromosome 14: 70758389–70761521). The corresponding crRNA (CTACTATATCTAGCATCCAC) was paired with tracrRNA to form the functional gRNA complex.

To facilitate precise integration, a donor plasmid was constructed containing a 3.0-kb 5’ homology arm, a 2 A-DreER coding sequence, and a 3.0-kb 3’ homology arm. The vector was assembled using a recombination-based cloning approach and linearized before embryo injection. The CRISPR/Cas9 components—Cas9 mRNA, gRNA, and the donor plasmid—were co-injected into fertilized one-cell C57BL/6J embryos, which were subsequently transferred into pseudopregnant recipients. Out of 168 embryos transferred, 23 pups were born. PCR and sequencing analysis identified two F0 founder mice carrying the targeted insertion. These founders were bred with wild-type C57BL/6J mice to produce the F1 generation. Genotyping of F1 offspring was performed using long-range PCR. Amplification at the 5’ end produced a 6.2-kb fragment in knock-in alleles (forward: 5’ CTTTAAAAGGGAATTGGAGGCATGT 3’; reverse: 5’ GGTGGCTCACCCCTATAATAATCTT 3’), whereas the wild-type allele yielded an 8.1-kb product. At the 3’ end, positive knock-in alleles showed a 3.5-kb band (forward: 5’ AAGAATTGGAGGACTGGAAAGATGA 3’; reverse: 5’ CCAAAAATCACTCCCATTCTCAACA 3’), and wild-type alleles gave a 7.5-kb band. Five F1 heterozygous mice were confirmed by sequencing to carry the correct insertion. One of these was used to establish a stable breeding colony.

Genomic DNA extraction and genotyping

Mouse tails were lysed in 1×TE buffer at 95 °C for 5 min. After cooling, proteinase K (200 µg/mL; BBI, A600451-0250) was added, and the samples were incubated overnight at 65 °C. The following day, samples were incubated again at 95 °C for 5 min to inactivate residual enzymes, completing the DNA extraction process. All mice were genotyped using specific primers designed based on the knock-in sequence, allowing clear distinction between mutant and wild-type alleles. Representative results of PCR genotyping, as well as details of primer design and sequences for all mouse models used in this study, are presented in Table S1.

Histology and immunofluorescence

The right ventricle of the mice was perfused with 1–3 mL PBS, and the lungs were inflated with 4% PFA, then fixed in 4% PFA for 2 h at 4 °C, washed four times with PBS for 30 min each at 4 °C, and embedded in OCT after 30% sucrose incubation.

For immunofluorescence staining, slides were fixed in 4% PFA at room temperature for 10 min, then washed three times with PBS. For OCT-embedded slides, antigen retrieval (Cat# BRR2004CLX; Biocare Medical) was performed for 30 min at 95 °C. Slides were washed with 0.1% Tween 20 in PBS (PBST) three times for 5 min. Slides were then incubated in blocking buffer (3% donkey serum in PBST) for at least 1 h, followed by overnight incubation in primary antibodies diluted in 1% donkey serum in PBST at 4 °C. The following primary antibodies were used: rabbit anti-mouse SFTPC (Cat# AB3786; Millipore Sigma; 1:2000), goat anti-mouse tdTomato (tdT) (Cat# AB8181-200; OriGene; 1:500), rat anti-mouse RAGE (Cat# MAB11794; R&D Systems; 1:200), mouse anti-mouse SCGB1A1 (Cat# sc-390313; Santa Cruz Biotechnology; 1:250), goat anti-mouse SCGB1A1 (Cat# ABS1673; Millipore Sigma; 1:200). Slides were washed with PBST and then incubated for 1 h at room temperature in secondary antibodies diluted in PBST. The following secondary antibodies were used: donkey anti-mouse IgG (H + L), Alexa Fluor 647 (Cat# A31571; Thermo Fisher Scientific; 1:500); donkey anti-rabbit IgG (H + L), Alexa Fluor 647 (Cat# A31573; Thermo Fisher Scientific; 1:500); donkey anti-goat IgG (H + L), Alexa Fluor 647 (Cat# A21447; Thermo Fisher Scientific; 1:500); donkey anti-rat IgG (H + L), Cy5 (Cat# 712-605-153; Jackson ImmunoResearch; 1:500). DAPI (0.2 µg/mL) (Cat# XW287189031; Sinopharm) was added for 5 min, and slides were then washed with PBS three times for 5 min and treated with FluorSave to maintain fluorescence.

Tamoxifen administration

For lineage labeling studies, tamoxifen was dissolved in corn oil and administered intraperitoneally at a dosage of 100 mg/kg per day.

Hyperoxic injury

Newborn mouse pups from dams delivering on the same day were randomized at birth (P0) and assigned to equal-sized litters. Following randomization, mouse cages were maintained either in room air (normoxia, 21% O₂) or exposed to hyperoxia (80% O₂) from P0 to P2, after which all mice were returned to room air and maintained under normoxic conditions until the indicated time points for tissue harvest.

Spatial transcriptomics and cell–cell communication analysis

Stereo-seq spatial transcriptomics data were imported from h5ad files into R using zellkonverter and converted into SingleCellExperiment objects. Raw count matrices were extracted from the layers/raw_counts slot and used to construct Seurat objects (Seurat v4). Spatial coordinates were obtained from obsm/spatial and stored in the metadata as imagecol and imagerow, and data were log-normalized using NormalizeData. BASCs, AT2 cells, and club cells were defined based on marker gene expression, with BASCs identified by combined Scgb1a1 and Sftpc expression at the bronchioalveolar duct junction, AT2 cells defined by high Sftpc expression, and club cells defined by high Scgb1a1 expression. To define cell-type–specific spatial microenvironments, a k-nearest neighbor (kNN) approach was applied based on spatial coordinates, and the 20 nearest neighboring spots (k = 20) surrounding each BASC, AT2 cell, or club cell were selected, excluding the central spot to avoid self-inclusion. From these neighboring spots, mesenchymal microenvironmental cells were identified by the absence of Epcam, Ptprc, and Pecam1 expression and retained as niche cells. Cell–cell communication analysis was performed using NicheNet, modeling mesenchymal niche cells as sending cells and BASC, AT2, or club epithelial cells as receiving cells, while incorporating the spatial neighborhood constraint to restrict candidate interactions and prioritize ligand–target gene regulatory relationships supported by spatial proximity.

Statistical analysis

Statistical analysis was carried out using GraphPad Prism software. One-tailed Student’s t tests were used to generate P values. A P value less than 0.05 was considered significant. N.S., not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Cell counting and image analysis

At least three samples per genotype/condition were used, and at least 5 randomly selected sections were chosen for each sample. Cell counts for stained cells were performed in Fiji using the “Cell Counter” plug-in. For all analyses, the performers were blinded to the specimen genotype and condition. Results were averaged between each specimen and SDs were calculated per condition. Fluorescence signals were visualized under a fluorescence microscope (Zeiss). Statistical analysis was performed by evaluating fluorescence intensities of 100–120 cells per group.

Results

BASCs appear at birth and are consistently maintained at the BADJ

BASCs have been detected at the BADJ in mouse lungs, their characteristic anatomical niche, and are identifiable as early as postnatal day 6 (P6) during the alveologenesis phase [6]. Given that the BADJ forms at E17 during the saccular stage [11], we hypothesized that BASCs emerge during this developmental period. To test this, we performed immunofluorescence staining for SFTPC and SCGB1A1 on lung tissues collected from E17.5 to P4 (Fig. 1A). SFTPC⁺SCGB1A1⁺ BASCs were absent at E17.5, emerged at P0, and persisted throughout the postnatal saccular stage, with stable localization and abundance (Fig. 1B, C).

Fig. 1.

Postnatal emergence and stable localization of BASCs during the saccular and alveolar stages. A Schematic of tissue collection from E17.5 to P14. B Immunofluorescence staining for SFTPC (red) and SCGB1A1 (green) showing SFTPC⁺SCGB1A1⁺ BASCs (arrowheads) at the BADJ. C Quantification of BASC numbers at the BADJ across postnatal time points during the saccular alveolar stages

We next investigated whether BASCs undergo changes during the subsequent alveologenesis stage by analyzing lung tissues from P6 to P14 (Fig. 1A). BASCs remained stably localized at the BADJ, with their numbers remaining relatively constant over this period (Fig. 1B, C). However, this finding does not rule out the possibility that BASC progeny contribute to alveolar and airway epithelial lineages during lung development. A lineage-tracing system capable of labeling BASCs will therefore be required to define their developmental dynamics.

Scgb1a1-lin cells encompass BASCs and contribute to alveolar epithelial lineages

Given that BASCs are positive for SCGB1A1 expression, we employed the Scgb1a1^CreER driver to target them. To this end, we crossed Scgb1a1^CreER mice with H11^{LSL − ZsG_RSR−tdT} reporter mice, in which tamoxifen-induced Cre activation triggers ZsGreen (ZsG) expression and labels the Scgb1a1-lin cells (Fig. 2A). To assess whether this tool can also label BASCs, we administered tamoxifen at P0–P2 and collected lung tissues at P7 and P14 (Fig. 2B). Immunofluorescence staining for SFTPC confirmed that SFTPC⁺ZsG⁺ BASCs were successfully labeled (Fig. 2C). Moreover, we observed that ZsG⁺ cells had migrated into the alveoli by P14, which was not observed at P7 (Fig. 2D). Given the anatomical proximity of the BADJ to alveoli, we hypothesized that BASCs contribute to alveolar epithelial lineages during lung morphogenesis.

Fig. 2.

Scgb1a1-lineage labeling marks BASCs and reveals migration of Scgb1a1-lin cells into the alveoli. A Breeding strategy for generating Scgb1a1^CreER:H11^{LSL-ZsG_RSR-tdT} mice. B Tamoxifen induction strategy for labeling Scgb1a1-lin cells during early postnatal development. C Immunofluorescence staining for SFTPC (red) and Scgb1a1-lin cells (green) demonstrated the presence of BASCs at the BADJ at P7 and P14. D Distribution of Scgb1a1-lin cells in lung at P7 and P14

To exclude the possibility that ZsG⁺ alveolar cells arose from nonspecific Cre leakage, we performed parallel lineage-tracing experiments using Scgb1a1^CreER:R26R^tdT mice. Following tamoxifen administration at P0–P2, lung tissues were collected at P5 (Fig. S1A). No Scgb1a1-lin cells were detected in the alveoli (Fig. S1B), and all Scgb1a1-lin cells were positive for SCGB1A1 protein staining (Fig. S1C), indicating that Scgb1a1^CreER does not ectopically label alveolar epithelial cells. In addition, a potential contribution from SCGB1A1⁺ club cells cannot be excluded. To address this question, a more precise lineage-tracing system is needed. Therefore, we propose generating Sftpc^DreER mice and crossing them with Scgb1a1^CreER:H11^{LSL − ZsG_RSR−tdT} mice. In this dual-recombinase system, Dre activation induces tdT expression, independent of Cre-mediated ZsG labeling. Consequently, BASCs would be uniquely labeled as tdT⁺ZsG⁺ cells, enabling specific fate tracking and analysis of their dynamic behavior during lung development.

Generation of the Sftpc^DreER Mouse for Labeling Sftpc-lin

To generate the Sftpc^DreER knock-in strain, we used CRISPR/Cas9-mediated homologous recombination to insert a cassette immediately upstream of the endogenous Sftpc stop codon (Fig. 3A). The targeting cassette consisted of a 2 A self-cleaving peptide sequence followed by Dre recombinase fused to a mutated estrogen receptor (DreER), and terminated with a woodchuck hepatitis virus posttranscriptional regulatory element (WPRE) and polyadenylation (polyA) signal. The donor construct was assembled via In‑Fusion cloning and flanked by 3 kb homology arms to ensure precise targeting.

Fig. 3.

Generation of Sftpc^DreER knock-in mice. A Schematic diagram of the CRISPR/Cas9 strategy for knock-in of DreER into the endogenous Sftpc locus. B PCR genotyping strategy showing primer locations across 5′ and 3′ junctions and internal control. C Representative PCR gels confirming correct targeting of the Sftpc^DreER allele in mutant but not wild-type mice. D Summary of genotyping results confirming successful knock-in

Targeted insertion was confirmed by polymerase chain reaction (PCR) genotyping using primers spanning the 5’ and 3’ junctional regions, along with an internal control (Fig. 3B). In Sftpc^DreER mice, expected fragments of 6.2 kb (5’ homology arm), 3.5 kb (3’ homology arm), and 474 bp (internal) were amplified, whereas these fragments were absent in wild-type littermates (Fig. 3C, D). This new strain was completely viable and fertile without overt phenotypic or growth abnormalities. These data confirm that the Sftpc^DreER line develops and reproduces normally, supporting its utility for in vivo lineage tracing of Sftpc-lin cells dynamics.

Specific labeling of AT2 cells and BASCs by Sftpc^DreER mice

We crossed Sftpc^DreER mice with the H11^{LSL − ZsG_RSR−tdT} reporter mice to label Sftpc-lin cells (Fig. 4A). Adult Sftpc^DreER:H11^{LSL − ZsG_RSR−tdT} mice received intraperitoneal injections of tamoxifen or corn oil (Fig. 4B). No tdT⁺ cells were detected in corn oil-treated mice, confirming the absence of background recombination (Fig. 4C). In addition, tamoxifen administration specifically labeled cells in the alveoli and at the BADJ, with no tdT⁺ signals detected in the airways (Fig. 4C). To evaluate labeling specificity and efficiency, we examined the colocalization of Sftpc-lin cells (tdT⁺) with SFTPC protein expression. Approximately 100% of tdT⁺ cells co-expressed SFTPC protein, and around 93% of SFTPC⁺ cells were labeled by tdT (Fig. 4D-F). These results validated the Sftpc^DreER model as a precise and specific tool for targeting SFTPC⁺ epithelial cells. We next assessed BASCs labeling by this model. In tamoxifen-treated adult Sftpc^DreER:H11^{LSL − ZsG_RSR−tdT} mice, co-staining for SCGB1A1 protein revealed that tdT⁺ cells at BADJ co-expressed SCGB1A1, confirming efficient BASCs labeling (Fig. 4G). Collectively, these findings demonstrate that the Sftpc^DreER mouse enables specific labeling of SFTPC⁺ cells, including both AT2 cells and BASCs in vivo, providing a valuable tool for lineage tracing and functional studies.

Fig. 4.

Specific labeling of AT2 cells and BASCs using Sftpc^DreER mice. A Breeding strategy for generating Sftpc^DreER:H11^{LSL-ZsG_RSR-tdT} mice. B Tamoxifen induction strategy for labeling Sftpc-lin cells in adult mouse. C Immunofluorescence showing the distribution of tdT⁺ cells (red) following tamoxifen induction. D-F Immunofluorescence staining and quantification of the reciprocal overlap between tdT⁺ cells (red) and SFTPC⁺ cells (green). G Immunofluorescence staining showing co-expression of tdT⁺ (red) and SCGB1A1⁺ (green) in BASCs at the BADJ

BASCs migrate into airway epithelium and commit to club cell fate

Using the above tools, Scgb1a1^CreER:Sftpc^DreER:H11^{LSL − ZsG_RSR−tdT} mice (BASC-Tracer) were generated for precisely labeling BASCs (Fig. 5A). In this dual recombinase system, SFTPC+ cells express DreER, which excises a Rox-flanked stop cassette to activate tdT, while SCGB1A1⁺ cells express CreER, which excises a LoxP-flanked stop cassette to activate ZsG (Fig. 5B). Using this system, we were able to specifically label BASCs and their progeny (tdT⁺ZsG⁺). To investigate the dynamic behavior of BASCs during lung development, tamoxifen was administered to BASC‑Tracer at P0-P2 and the lungs were analyzed at P7, P14, P21, P30, and P60 (Fig. 5C). At P7, dual-positive cells (tdT⁺ZsG⁺) were exclusively confined to the BADJ and were not detected within the airways or alveolar parenchyma (Fig. 5D, Fig. S2A). In addition, immunofluorescence staining analysis confirmed that these cells were BASCs (Fig. S2B). Moreover, the progeny of lineage-labeled BASCs (tdT⁺ZsG⁺) started to migrate into the airways by P14, and their numbers increased progressively from P30 to P60 (Fig. 5D).

Fig. 5.

Generation of BASC-Tracer and tracking BASC migration into the airway. A Generation of Scgb1a1^CreER:Sftpc^DreER:H11^{LSL-ZsG_RSR-tdT} triple-transgenic mice (BASC-Tracer). B Schematic of specific labeling of club cells, AT2 cells, and BASCs in the BASC-Tracer system. C Tamoxifen administration timeline at P0–P2 and lung collection at multiple postnatal time points. D Distribution of BASC-tracer⁺ (tdT⁺ZsG⁺) cells from P7 to P60 (yellow arrows indicate BASCs migrating into the airway; white arrows indicate BASCs at the BADJ). E–G Immunofluorescence staining and quantification of SFTPC⁺ or SCGB1A1⁺ in BASC-tracer⁺ (tdT⁺ZsG⁺) airway cells at P60

To determine whether these migrated cells still retain BASC identity (SFTPC⁺SCGB1A1⁺), we performed immunofluorescence staining for SFTPC and SCGB1A1 on P60 lung tissues and found that the tdT⁺ZsG⁺ cells migrated into the airways expressed SCGB1A1 but no longer retained SFTPC expression (Fig. 5E-G). These results suggest that BASCs migrate into the airway during postnatal development and contribute to the formation of airway epithelium.

BASCs migrate to alveoli and differentiate into alveolar cells

We next examined whether BASCs also contribute to alveolar epithelium during lung development. Examination of alveolar compartments from P7 to P60 demonstrated that lineage-labeled BASC progeny were detected in alveoli by P14, and their numbers increased thereafter (Fig. 6A). BASC-lin cells (tdT⁺ZsG⁺) migrating into the alveoli retained SFTPC expression but lost SCGB1A1 expression (Fig. 6B-D). Notably, a subset of alveolar tdT⁺ cells (including both ZsG⁻ and ZsG⁺ cells) exhibited a flattened, elongated morphology resembling AT1 cells (Fig. 6E). Given that previous studies have shown AT2 and BASCs can give rise to AT1 cells following lung injury in adult mice [10, 12, 13], we speculate that they may possess the same potential to differentiate into AT1 cells during normal development. To test this, we performed immunofluorescence staining for RAGE, an AT1 marker, on tissue sections from P60 lungs. Co-localization analysis showed that these cells were positive for RAGE protein staining, confirming their identity as AT1 cells (Fig. 6E). Taken together, these results establish BASCs as progenitors in lung development with the capacity to generate alveolar AT1/AT2 cells, as well as airway club cells.

Fig. 6.

Tracking BASC migration into the alveoli during lung development. A Distribution of BASC-tracer⁺ (tdT⁺ZsG⁺) cells in alveoli from P7 to P60 (white arrows indicate BASCs migrating into the alveoli). B-D Immunofluorescence staining and quantification of SFTPC⁺ or SCGB1A1⁺ in BASC-tracer⁺ (tdT⁺ZsG⁺) cells at P60 (white arrows indicate BASCs migrating into the alveoli). E Immunofluorescence staining showing a subset of tdT⁺ cells (both ZsG⁺ and ZsG⁻) co-expressing the AT1 marker RAGE

Neonatal hyperoxia impairs regenerative capacity of BASC-lin cells

We next determined how BASCs respond to injury during lung morphogenesis. Using the hyperoxia-induced bronchopulmonary dysplasia (BPD) model, neonatal BASC-Tracer mice were exposed to hyperoxia from P0 to P2 and subsequently returned to room air until P14, whereas controls remained in room air throughout (Fig. 7A). Lung tissues were then harvested to assess the impact of hyperoxic exposure on BASCs. The number of BASC-labeled cells was significantly reduced following hyperoxic injury compared with normoxic controls (Fig. 7B, C). Consistently, immunofluorescence analysis of the proliferative marker Ki67 revealed a marked suppression of proliferative activity in BASC-lin cells (Fig. 7B, D). Similarly, hyperoxic exposure significantly inhibited the proliferative capacity of AT2 and club cells (Fig. 7E-H). Together, these findings suggest that impaired proliferation of epithelial progenitor populations may represent a key pathogenic mechanism underlying BPD phenotypes, including alveolar simplification.

Fig. 7.

Hyperoxic injury suppresses BASC proliferation during lung development. A Experimental scheme for hyperoxic exposure in newborn mice. B–D Immunofluorescence staining and quantification of BASCs-Tracer⁺ cell numbers per BADJ and the proportion of Ki67⁺BASCs-Tracer⁺ cells(orange arrows indicate Ki67⁺BASCs-Tracer⁺ cells; yellow arrows indicate Ki67⁻ BASCs-Tracer⁺ cells). E and F Immunofluorescence staining and quantification of the proportion of Ki67⁺ AT2 cells (orange arrows indicate Ki67⁺ AT2 cells). G and H Immunofluorescence staining and quantification of the proportion of Ki67⁺ club cells (orange arrows indicate Ki67⁺ club cells)

Spatial transcriptomics reveals distinct stromal niches at the BADJ, alveoli, and airways

To systematically compare the microenvironmental features surrounding BASCs, AT2 cells, and club cells, we performed Stereo-seq–based spatial transcriptomics analysis on lung tissues [14]. Using Scgb1a1 expression as a spatial reference, we delineated the BADJ region within lung sections, and spatially mapped BASCs as a discrete epithelial population enriched in close proximity to the BADJ. We operationally defined the BASC niche as fibroblasts neighboring BASC-associated spots, terming them “BASC niche cells” (Fig. 8A). Using the same strategy, we defined club cells and their neighboring airway fibroblasts, as well as AT2 cells and their neighboring alveolar fibroblasts (Fig. 8B, C).

Fig. 8.

Spatial transcriptomic comparison of BASCs, club cells, AT2 cells, and their respective niches. A–C Spatial transcriptomic maps showing the distribution of BASCs, club cells, AT2 cells (red), and their corresponding niche cells (yellow), and dot plots displaying representative marker gene expression for each epithelial population and its respective niche. D–F NicheNet analysis predicting ligand–receptor interactions from respective niche cells to BASCs, club cells, and AT2 cells. G Dot plot showing the expression intensity of ligands in niche cells corresponding to BASCs, club cells, and AT2 cells

To investigate niche-derived signaling to epithelial progenitor cells, we applied NicheNet to model the microenvironment and predict ligand–receptor interactions that could target these progenitors. NicheNet analysis predicted highly similar ligand activities across these three epithelial progenitor cells (Fig. 8D-F). However, the actual, spatially-resolved expression profiles of these ligands differed substantially among these their respective niches. Notably, FGF signaling was significantly enriched in the BASC microenvironment, in line with prior studies demonstrating its activation during lung morphogenesis [3, 15, 16] (Fig. 8G). These findings suggest that the BADJ constitutes a microenvironment with distinct signaling features, potentially supporting BASC maintenance and/or fate transitions during postnatal lung development.

Discussion

In this study, immunofluorescence analysis revealed the presence of BASCs as early as P0. To further delineate their developmental dynamics, we employed a dual-recombinase BASC-Tracer system to trace BASCs from P0 through adulthood. BASCs remained restricted to the BADJ at P7. From P14 onward, their progeny was observed to depart from BADJ and migrated to airway and alveolar parenchyma. Previous studies using split-Cre lineage tracing reported that BASCs are absent in embryonic lungs at E18.5 and only become detectable by P6 [6]. However, differences in lineage-tracing sensitivity may account for the delayed detection of BASCs in these models compared with our direct immunofluorescence approach.

Given the fundamental link between development and regeneration, we demonstrate that BASCs depart from their niche at the BADJ to airway and alveolar parenchyma during postnatal development, suggesting a contribution of BASCs to airway and alveolar maturation. This developmental migration behavior mirrors their injury-activated response observed in lung regeneration models [17–19]. Previous studies have demonstrated that hyperoxic exposure in early life significantly suppresses the proliferative activity of AT2 and club cells, leading to BPD [20, 21]. Here, we report a similar suppression of BASC proliferation under the same injury condition. This suggests that the impairment of BASCs may therefore represent a previously underappreciated mechanism impairing normal lung morphogenesis as well as the defective lung repair and alveolar simplification seen in BPD.

Our previous work, along with studies by others, indicate that the alveolar and airway regions of the lung possess distinct stromal niches, which play a critical role in regulating the fate of epithelial progenitor cells during injury and repair [22–25]. More direct evidences from Lee et al.. demonstrated that club cells co-cultured with Lgr6⁺ airway fibroblasts preferentially form airway-like organoids, whereas co-culture with Lgr5⁺ alveolar fibroblasts promotes the formation of alveolar-like structures [26]. These findings from regeneration contexts support a model proposing that physiological BASC differentiation during development is similarly governed by spatially distinct niche-derived cues. Specifically, signals within the alveolar microenvironment promote AT2 maturation [27, 28], while signals within the airway compartment promote club cell differentiation [29, 30]. Therefore, a unique microenvironment at the BADJ itself may be essential for establishing and maintaining BASC identity. Consistent with this hypothesis, spatial transcriptomic analysis revealed that distinct niches exhibit marked differences in the relative expression levels of signaling ligands, suggesting that these differences may orchestrate BASC fate decisions.

This study has certain limitations that also point to future directions. First, lineage tracing based on static imaging and marker co-expression cannot fully resolve dynamic lineage relationships; future studies employing live imaging and single-cell trajectory analysis will be valuable to validate our findings. Second, while our work provides new insights into BASC dynamics, key questions remain regarding their precise functional contribution and regulatory mechanisms. Addressing these questions will require two approaches: developing genetic tools to selectively manipulate BASCs in vivo to test their necessity, and applying single-cell spatial omics across developmental stages to define the niche signals controlling their fate. Targeted co-culture assays or direct signaling perturbations will be essential to strengthen mechanistic claims.

Conclusions

In summary, this study defines the developmental dynamics of BASCs by establishing their early postnatal presence (detected at P0) and bipotent differentiation into alveolar and airway lineages. Moreover, we identify BASCs as a target of neonatal hyperoxic injury, implicating their dysfunction in impaired lung development. The distinct molecular niche we delineated provides a foundational framework for understanding the signals that regulate this key progenitor population.

Abbreviations

AT1

Alveolar type 1 cell

AT2

Alveolar type 2 cell

BADJ

Bronchioalveolar duct junction

BASC

Bronchioalveolar stem cell

BPD

Bronchopulmonary dysplasia

CreER

Tamoxifen-inducible Cre recombinase

DAPI

4′,6-diamidino-2-phenylindole

DreER

Tamoxifen-inducible Dre recombinase

ECM

Extracellular matrix

E

Embryonic day

Ki67

Proliferation marker protein Ki67

LSL

LoxP-stop-LoxP cassette

P

Postnatal day

PFA

Paraformaldehyde

polyA

Polyadenylation signal

RAGE

Receptor for advanced glycation end products

RSR

Rox-stop-Rox cassette

SCGB1A1

Secretoglobin family 1 A member 1

SFTPC

Surfactant protein C

Stereo-seq

Spatial Enhanced resolution omics-sequencing

WPRE

Woodchuck hepatitis virus posttranscriptional regulatory element

Authors’ contributions

Conceptualization: C.W. and Y.L.Methodology: Y.L., J.L., S.C. H.W., and R.L.Investigation: Y.L., S.C., J.L., H.W., R.L. and L.J.Supervision: C.W and Y.X.Writing—original draft: Y.LWriting—review & editing: R.L., Y.L, Y.X. and C.W.

Funding

This work was supported by grants from the National Key R&D Program of China 2024YFA1107703, the Major Project of Guangzhou National Laboratory GZNL2023A01003, the Strategic Priority Research Program of the Chinese Academy of Sciences XDB0830000, the Zhongshan Science and Technology Bureau CXTD2023009, and the State Key Laboratory of Drug Research SKLDR-2022-LH-11 to C.W.

Data availability

Stereo-seq data were deposited into the STOmics Datasets under accession number STDS0000062 and are available at the following URL: https://db.cngb.org/stomics/datasets/STDS0000062.

Declarations

Ethics approval and consent to participate

All experiments were conducted in accordance with institutional ethical guidelines for animal care and were approved by the Institutional Animal Care and Use Committees at Shanghai Institute of Materia Medica and Zhongshan Institute for Drug Discovery (2025-03-WCQ-01).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.