Wnt5a/β-catenin axis is involved in the downregulation of AT2 lineage by PAI-1

Abstract

Failure to regenerate injured alveoli functionally and promptly causes a high incidence of fatality in coronavirus disease 2019 (COVID-19). How elevated plasminogen activator inhibitor-1 (PAI-1) regulates the lineage of alveolar type 2 (AT2) cells for re-alveolarization has not been studied. This study aimed to examine the role of PAI-1-Wnt5a-β catenin cascades in AT2 fate. Dramatic reduction in AT2 yield was observed in Serpine1^Tg mice. Elevated PAI-1 level suppressed organoid number, development efficiency, and total surface area in vitro. Anti-PAI-1 neutralizing antibody restored organoid number, proliferation and differentiation of AT2 cells, and β-catenin level in organoids. Both Wnt family member 5A (Wnt5a) and Wnt5a-derived N-butyloxycarbonyl hexapeptide (Box5) altered the lineage of AT2 cells. This study demonstrates that elevated PAI-1 regulates AT2 proliferation and differentiation via the Wnt5a/β catenin cascades. PAI-1 could serve as autocrine signaling for lung injury repair.

INTRODUCTION

Elevated plasminogen activator inhibitor-1 (PAI-1) in the circulatory system is an independent biomarker predicting the severity and mortality of coronavirus disease 2019 (COVID-19), acute respiratory distress syndrome (ARDS) in adulthood, and preterm newborns (1–4). Proteomics analysis of COVID-19 lungs and airway fluid biopsies shows an increment in PAI-1 expression (5, 6). This is consistent with the association between elevated PAI-1 and other lung injury diseases (7–9). Increased PAI-1 may play a critical role in the aging (10). PAI-1 level could be increased several folds depending on diseases and tissue types. As a specific inhibitor of plasminogen activators, urokinase-type plasminogen activator (uPA) and tissue-type plasminogen activator (tPA), increased PAI-1 results in a decline in plasmin and fibrinolytic activity. In addition to the fibrinolytic role, PAI-1 regulates the senescence of cells. The expression of PAI-1 in stem cells is dramatically increased (11). Therefore, PAI-1 will serve as both paracrine and autocrine signaling for priming lung stem cells.

Alveolar type 2 (AT2) epithelial cells function as facultative progenitors to maintain homeostasis and lung injury repair through self-renew (proliferation) and differentiation to alveolar type 1 (AT1) cells (12). This AT2-AT1 differentiative lineage is regulated by the Notch, Hedgehog, and Wnt signaling pathways (13, 14). The role of the fibrinolytic niches in regulating AT2 lineage is emerging. Either suppression of endogenous uPA or delivery of PAI-1 adversely affects the wound healing assay of alveolar epithelial cells (15). Our study recently showed that gene encoding urokinase-type plasminogen activator (Plau; encoding uPA) deficiency led to the suppressed lineage of mouse AT2 cells via a uPA-A6-CD44-epithelial sodium channels (ENaC) cascade (16). Moreover, Plau mediates yes-associated protein (YAP) regulation of proliferation of epidermal stem cells (17). In contrast to uPA knockout, PAI-1 deficiency enhances the proliferation of endothelial cells in vitro (18). Although the elevation in PAI-1 is the most common feature of inflamed lungs, the role of elevated PAI-1 in the AT2 lineage has not been reported.

This study aimed to test the effects of increased PAI-1 on the proliferation and differentiation of primary AT2 cells in three-dimensional (3-D) organoids. Both feeder-free and fibroblast-AT2 cocultures were utilized (19). To mimic elevated PAI-1 in injured lungs, primary AT2 cells were isolated from gene encoding plasminogen activator inhibitor-1 (Serpine1) transgenic mice overexpressing the human PAI-1 gene. Our results suggest that increased expression of PAI-1 declined the lineage of AT2. The Wnt5 signaling pathway could be a key player in regulating AT2 lineage by PAI-1.

MATERIALS AND METHODS

Animal Husbandry

All mice purchased from Jackson Laboratory were kept in a pathogen-free facility. A 12-h light/dark cycle and ad libitum supply for food and water were provided. Age, sex, and weight-matched (4–12 mo) wild type (wt) and Serpine1^Tg mice were euthanized for experiments, as approved by the Institutional Animal Care and Use Committee of the University of Texas at Tyler Health Science Center.

AT2 Cell Isolation with Fluorescence-Activated Cell Sorting

Mouse AT2 cells were isolated from wt (C57BL/6j) and serpine1^Tg strains, as described earlier (20). Briefly, lungs were injected with dispase (50 U/mL) and removed from euthanized mice. After incubating in dispase (50 U/mL) for 45 min, lungs were gently teased and incubated in medium [DMEM/F12(1:1) plus 0.01% DNase I] for 10 min. Cells were passed through a series of cell strainers (100, 40, and 10 μm; Corning) and centrifuged at 125 g for 10 min. AT2 cells were negatively selected by magnetic cell sorting (MACS) by labeling them with biotin-conjugated anti-CD16/32, CD45, and CD119 antibodies (BD PharMingen), followed by incubating with streptavidin-coated magnetic particles. The live epithelial cell adhesion molecule (EpCAM) positive AT2 cells were sorted with fluorescence-activated cell sorting (FACS) by labeling cells with Alexa Fluor conjugated EpCAM antibodies and 7-aminoactinomycin D (7AAD) dye to discriminate between live and dead cells. The viability and purity of FACS sorted AT2 cell preparations were ∼94% and ∼98%, respectively.

Cocultures of AT2 Organoids with Fibroblasts

Fibroblasts and AT2 cells were seeded in Matrigel to grow organoids, as we previously reported (16). Briefly, 2,000 AT2 cells were mixed with 1 × 10⁵ mouse lung fibroblasts (MLG2908) and suspended into 60 µL of growth factor reduced Matrigel (Corning) diluted (1:1) with organoid growth medium (Supplemental Table S1; all Supplemental material is available at https://doi.org/10.6084/m9.figshare.20134106). The mixed cell suspension (60 µL) was seeded on the apical chamber of clear polyester membrane transwell inserts (6.5 mm diameter, 0.4 µm pore size, Corning Costar) and incubated for 30 min to allow the matrix to solidify. Organoid growth medium (500 µL) was then added to the bottom well and incubated for 4 days. On the fourth day, the medium was switched to the organoid growth medium without anaplastic lymphoma kinase (ALK) inhibitor, and cultures were maintained for 10 days, with the medium changed every 48 h.

Feeder-Free 3-D Organoid Cultures

AT2 cells were seeded in Matrigel to culture organoids, as previously described with modification (19). Briefly, 2,000 cells were suspended into 60 µL of growth factor reduced Matrigel (Corning) diluted (1:1) with organoid growth medium (a maturation medium, AMM; Supplemental Table S2). The cell suspension (60 µL) was seeded on the apical chamber of clear polyester membrane transwell inserts (6.5 mm diameter, 0.4 µm pore size, Corning Costar) and incubated for 30 min to allow the matrix to solidify. Then 500 µL of AMM with IL-1b and Y-27632 was added to the bottom well and incubated for 4 days. After that, the medium was switched to AMM without IL-1b and Y-27632 (Supplemental Table S2) and changed every 3 days. On the tenth day, the medium was switched to a differentiation medium (ADM) (Supplemental Table S2) and changed every 3 days.

Treatment of Organoids with PAI-1 Antibody, Wnt5a, and Box5

Feeder-free AT2 organoids were treated with PAI-1 antibody, Wnt family member 5A (Wnt5a; 200 ng/mL, R&D, 645-WN-010), and Wnt5a-derived N-butyloxycarbonyl hexapeptide (BOX5; 200 µM, Sigma, No. 681673) from day 0 to day 10. Box5 is a Wnt-5a-derived hexapeptide that antagonizes Wnt-5a-mediated cellular activities in an isoform-specific manner. PAI-1 antibody was added at a series of concentrations (0, 0.01, 0.1, 1, and 10 µg/mL). Primary AT2 cells were seeded in transwell inserts to grow organoids in triplicate. Images were captured on day 10 for proliferating organoids and day 17 for differentiated organoids. Cultures treated with the same volume of vehicles served as controls.

Differential Interference Contrast Imaging

Following 10-day proliferation and 7-day differentiation, colonies were visualized with an Olympus IX 73 microscope combined with a microscope objective (×4, NA:0.16, FN:26.5, UPlanSAPo, Olympus, Tokyo, Japan). Differential interference contrast (DIC) images of whole transwell inserts were captured with a Hamamatsu photonics CMOS Camera (Orca Flash 4.0; 2,048 × 2,048 pixels). Colony number, colony-forming efficiency, diameter, and surface area of organoids were analyzed using ImageJ software (21).

EdU Labeling of AT2 Cells

Proliferating AT2 cells with active DNA synthesis in organoids were detected using the Click-iT 5-ethynyl-2-deoxyuridine (EdU) assay kit (Thermo, C10499). Organoids on day 10 were incubated with 10 mM EdU for 3 h in a culture medium and then further processed for staining with the Click-iT EdU Alexa Fluor 488, following the manufacturer’s instructions. Images were captured and analyzed for the percentage of EdU⁺ cells. The z sections of an entire organoid were stacked for counting total (Hoechst-stained nucleus) and EdU⁺ cells using a cell counter plug-in for the ImageJ. The percentage of EdU⁺ cells was calculated and compared.

Immunofluorescence Staining and Confocal Imaging

Organoids were fixed in 4% paraformaldehyde for 45 min at room temperature. The fixed organoids were permeabilized with 0.2% Triton X-100 for 16 h. Nonspecific antigen binding was blocked by 3% BSA with 5% goat serum. Anti-receptor for advanced glycation end-products (RAGE; 1:200, R&D Systems) and anti-prosurfactant protein C (SPC; 1:300, Millipore) antibodies were used to detect AT1 and AT2 cells, respectively. Organoids were incubated with β-catenin antibody (1:100, Santa Cruz sc-7963) to detect whole organoid content. Then organoids were washed five times with PBS, 15 min each, to remove primary antibodies. Secondary antibodies were applied [Alexa Fluor 647 and Alexa Fluor 488 were used at 1:500 concentration, and Hoechst (1:1,000) was used to counterstain nuclei]. Organoids were then washed five times again with PBS, 15 min each to remove access secondary antibodies. Transwell membrane was removed and mounted with a water-based self-hard-set fluoro gel mounting medium (Electron Microscopy Sciences, Cat. No. 17985-10). Images were captured using a Zeiss LSM 510 confocal microscope. At least 10 organoids were imaged per transwell. Z-stacks were captured with a 1 μm step size. All images were subsequently processed and analyzed using ImageJ software.

RT-PCR Assay

The following primer pairs synthesized by Sigma-Aldrich were used for real-time PCR. Pro-SPC: 5′- ACAATCACCACCACAACGAG-3′ (forward) and 5′- AGCAAAGAGGTCCTGATGGA-3′ (reverse); GAPDH: 5′- TTGAGGTCAATGAAGGGGTC-3′ (forward) and 5′- TCGTCCCGTAGACAAAATGG-3′ (reverse). qPCR assays were performed with a SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, Cat. No. 1725271). Reaction for pro-SPC and GAPDH was performed using a single cycle of 95°C for 1 min, followed by 40 cycles of 95°C for 10 s and 58°C for 30 s. Relative expression of pro-SPC was calculated using the 2^Δct comparative method, with normalization of each sample against the expression of endogenous reference gene GAPDH. Further, the expression level of pro-SPC in PAI-1 deficient cells was normalized to that of wt preparations.

RESULTS

PAI-1 Downregulates Mouse AT2 Hemostasis in Mouse Lungs and Renewal in Organoids

An elevated PAI-1 level is a hallmark of lung injury (1, 2, 22–25). Transgenic mice overexpressing human Serpine1 gene mimicked the disrupted fibrinolytic niche in patients with ARDS (26, 27). To compare the AT2 yield between wt and Serpine1 transgenic (Serpine1^Tg) mice, lung cells freshly isolated were labeled with EpCAM antibody and sorted using FACS. There was a significant reduction in AT2 cells in Serpine1^Tg mice (mean 1.0 ± 0.05 × 10⁶ cells vs. 1.5 ± 0.07 × 10⁶ cells for wt, P = 0.001, Fig. 1A). This observation was confirmed by quantitating pro-SPC (a biomarker of AT2 cells) transcripts (0.30 ± 0.12 vs. 1.00 for wt, P = 0.004, Fig. 1B). When equal AT2 cells were mixed with feeder cells (fibroblasts) and grown as spheroids (Fig. 1C), lesser organoids were developed by Serpine1^Tg cells than those of wt controls (41.0 ± 4 vs. 72.33 ± 3.38 for wt, P = 0.004, Fig. 1D). The schematic illustration shows feeder and AT2 cells seeded in Matrigel in a transwell chamber. Over a period (of 4–12 days), AT2 cells proliferate and differentiate to develop 3-D organoids. Feeder cells provide niche support to AT2 cells for proliferation and differentiation (Fig. 1C). Correspondingly, the total surface area, an index for forming in vivo inner surface area of the alveoli, decreased (1.36 ± 0.06 vs. 2.36 ± 0.12 for wt, P = 0.002, Fig. 1E), suggesting a potential of suppressed differentiation from AT2 into AT1 cells. Indeed, the AT1/AT2 ratio was markedly different between the two groups (1.53 ± 0.11 vs. 2.36 ± 0.29 for wt, P = 0.02, Fig. 1, F and G). These results suggest that elevated PAI-1 may downregulate the lineage of AT2 cells.

Figure 1.

Mouse AT2 yield and downregulation of AT2 fate by Serpine1 gene in 3-D cocultured organoids. The purity and viability of AT2 cells of wild-type (wt) and Serpine1^Tg (Tg) mice were confirmed, as we described previously (16, 20). A: AT2 yield. EpCAM-positive cells were counted as AT2 subpopulations by FACS. n = 10 pairs. B: Pro-SPC mRNA level in AT2 cells. n = 3/group. C: representative differential interference contrast (DIC) images of 3-D cultures. The images were captured on day 10. AT2 cells were cocultured with feeder cells in Matrigel to grow 3-D organoids in transwell inserts. Scale bar, 1 mm. D: number of organoids per dish. n = 3/group. E: total surface area of organoids per dish. n = 3/group. F: immunofluorescent images of organoids. Organoids were incubated with RAGE (1:200) and pro-SPC (1:200) antibodies for recognizing AT1 and AT2 cells, respectively. Nuclei were stained with Hoechst dye. Organoids were visualized by a Zeiss confocal microscope. Images were analyzed using ImageJ software. Scale bar, 100 mm. G: quantitative analysis of AT1 to AT2 ratio. n = 21 organoids/group. Unpaired two-tailed Student t test for A, B, D, E, and G to compare wt and Tg groups. Data expressed as means ± SE. AT1, alveolar type 1; AT2, alveolar type 2; EpCAM, epithelial cell adhesion molecule; FACS, fluorescence-activated cell sorting; SPC, surfactant protein C; 3-D, three-dimensional.

Fibroblasts Are Not Required for the Downregulation of AT2 Lineage

Fibroblasts may regulate the fate of AT2 cells through the release of nutritious and signal molecules. Another concern is that the normal fibrinolytic system in feeder cells may interfere with the over-expressed human PAI-1 in Serpine1^Tg AT2 cells. We, therefore, confirmed the results from the feeder-cell and AT2 cocultured organoids in feeder-free 3-D cultures. The schematic illustration shows AT2 cells seeded in Matrigel in a transwell chamber and growth factor enriched media in the bottom well to compensate for niche support without feeder cells (Fig. 2A). This two-stage protocol allows us to track the proliferation and differentiation of AT2 cells (Fig. 2A, inset). Intriguingly, Serpine1^Tg cells formed lesser proliferative organoids (48 ± 3 organoids vs. 61 ± 3 organoids for wt controls, P = 0.002, n = 18 transwell inserts) and surface area (3.59 ± 0.35 vs. 5.82 ± 0.48 for wt controls, P = 0.002, Fig. 2, B and C). Schematic illustration shows feeder-free proliferating AT2 organoids lined with only AT2 cells (Day 0–10) in the presence of AMM medium and differentiating AT2 organoids consisting of both AT1 and AT2 cells (Day 11–17) in the presence of ADM medium (Fig. 2D). When the organoids were subsequently differentiated, this reduction in the organoid number and surface area in Serpine1^Tg cells remained (60.22 ± 3.51 vs. 75.67 ± 1.68 for wt controls, P = 0.001, Fig. 2, D–F).

Figure 2.

Suppression of the lineage of mAT2 cells by Serpine1 in feeder-free organoids. Primary AT2 cells from wt and Serpine1^Tg mice were grown as fibroblast-free organoids. A: DIC images of proliferation cultures. Organoids were cultured with AT2 cells for feeder-free organoids and analyzed as shown in the inset of A for two phases: proliferation and differentiation. Images of the organoids were captured on day 10. Scale bar, 1 mm. B: organoids count. n = 9 transwell inserts per group. C: total surface area of organoids. n = 9 transwell inserts per group from at least three mice. D: DIC images of differentiated cultures. Both AT1 and AT2 cells were visualized. E: organoid count. n = 9 transwell inserts per group. F: surface area of total organoids. n = 9 transwell inserts per group. G: EdU assay. Cells were stained with EdU (green) and Hoechst (blue) dyes. Dotted lines depict the boundary of organoids for cell counting. Scale bar, 100 mm. H: percentage of EdU positive (EdU⁺) proliferating AT2 cells. n = 12 organoids/group. Data expressed as means ± SE. Unpaired two-tailed Student t test for B, C, E, F, and H. AT1, alveolar type 1; AT2, alveolar type 2; DIC, differential interference contrast; EdU, 5-ethynyl-2-deoxyuridine; wt, wild type.

Moreover, overexpression of the Serpine1 gene diminished proliferating AT2 cells (43.36 ± 3.16 vs. 59.90 ± 1.5 for wt controls, P = 0.0001, Fig. 2, G and H). Surprisingly, very few cells in differentiated organoids were detected as EdU⁺ cells (Supplemental Fig. S1). We demonstrate that the downregulation of AT2 fate is independent of the interaction with fibroblasts but possibly due to the altered proliferation of AT2 cells.

Serpine1^Tg Organoids Morphologically Differ from Wt Controls

Immunofluorescent images of the organoids also revealed the importance of fibrinolytic activity in maintaining the morphology and size of organoids. Differences in the arrangement pattern of AT1 and AT2 cells in organoids were compared. Serpine1 altered the structure and thickness of organoids. Typically, wt organoids were larger and consisted of a single large lumen, whereas Serpine1^Tg organoids were smaller with multiple small lumens. In Serpine1^Tg organoids, more alveolus-like cavities were observed compared with the wt group, which has a single large cavity (Fig. 3, A and B). Flatted AT1 cells lined both the inner and outer surface of wt organoids. In contrast, only the outer surface of Serpine1^Tg organoids was partially covered by squamous transdifferentiating AT2 cells. AT2 cells and AT1 cells were embedded in the outer spherical and inner septal wells. Total AT1 and AT2 cells were counted in overall organoids from both groups. The number of AT1 cells in wt organoids was greater than Serpine1^Tg organoids (100.4 ± 10.29 vs. 164.6 ± 25.9 for wt controls, P = 0.03, Fig. 3C), whereas there was no difference in AT2 cells (124.9 ± 14.32 vs. 111.1 ± 18.93 for wt controls, P = 0.56, Fig. 3D).

Figure 3.

Serpine1 gene regulates the structure of organoids. Feeder-free differentiated organoids were cultured up to day 17. A: representative confocal images of the small, medium, and large wt organoids. Organoid size (diameter): small 50–150 µm, medium 151–500 µm, and large >501 µm. The white arrowhead points to a large single lumen marked by white dotted lines. Orange arrowheads show a thin double-layer wall of organoids surrounding the large single lumen. The schematic inset (right end) represents a typical large wt organoid consisting of a single large lumen. B: representative confocal images of the small, medium, and large organoids from Serpine1^Tg organoids. Scale bar, 100 mm. Orange arrowheads point to a thick single-layer wall surrounding multiple small lumens. White arrowheads point to multiple small lumens marked with dotted lines. Serpine1^Tg organoids were illustrated, showing multiple small lumens. C: AT1 cells per organoid. D: AT2 cell count. E: AT1/AT2 ratio of total organoids. n = 10 organoids/group. F: AT1/AT2 ratio sorted by the size of organoids. n = 10 organoids/group. Unpaired two-tailed Student t test for C and D. Data expressed as means ± SE. AT1, alveolar type 1; AT2, alveolar type 2; Serpine1, gene encoding plasminogen activator inhibitor-1; wt, wild type.

Consequently, AT1/AT2 ratio was significantly reduced in Serpine1^Tg cultures (0.81 ± 0.04 vs. 1.52 ± 0.06 for wt controls, P < 0.0001, Fig. 3E). Further, this difference in AT1/AT2 ratio was found in different sizes (small, medium, and large) of organoids (Fig. 3F). These observations suggest that during the maturation of organoids from small to large size, more alveolus-like septa were developed in Serpine1^Tg organoids, which could be related to inconsistent AT1/AT2 ratio with the wt group.

PAI-1 Neutralizing Antibody Restores the Fate of Serpine1^Tg AT2 Cells

To validate that the decrease in colony number in vitro is attributed to excessive PAI-1 proteins, we treated organoids with PAI-1 neutralizing antibodies during the periods of proliferation and differentiation cultures (Fig. 4A). Based on the dose-response study (Supplemental Fig. S1), the optimized concentration (0.1 mg/mL) was included in the culture medium and replaced, as shown in the inset of Fig. 4A. PAI-1 antibody restored the organoid numbers of proliferating (33.71 ± 1.32 for Tg group vs. 51 ± 1.1 for Tg+ab group, P < 0.0001, Fig. 4B) and differentiated cultures (35 ± 1.7 for Tg group vs. 48.67 ± 1.2 for Tg+ab group, P < 0.0001, Fig. 4C). Further, the reduction in EdU⁺ AT2 cells in Serpine1^Tg organoids was eliminated (66 ± 2.7 for wt vs. 70.95 ± 3.8 for the Tg+ab group, P = ns, Fig. 4, D and E). In parallel, the PAI-1 neutralizing antibody increased the AT1/AT2 ratio, which was significantly lesser than that of wt organoids (1.28 ± 0.07 for wt vs. 0.87 ± 0.02 for Tg group, P = 0.0003 and 1.28 ± 0.07 for wt vs. 1.18 ± 0.07 for Tg+ab group, P = 0.33, Fig. 4, F and G).

Figure 4.

PAI-1 neutralizing antibody restores spheroid formation. A: DIC images of proliferating and differentiated organoids. Feeder-free organoids were cultured under proliferating (left three) and differentiating conditions (right three DIC images) and treated with PAI-1 Ab (0.1 mg/mL), as shown in the inset. Images were taken on day 10 (left three images) and day 17 (right three images) for proliferation and differentiation, respectively. Scale bar, 1 mm. B: quantitation of proliferating organoids. n = 9 transwell inserts per group from at least three mice. C: quantitation of differentiated organoids. n = 7 wells/group from three mice. D: images of EdU labeled AT2 cells. Cells within the dashed area were counted. Scale bar, 100 mm. E: EdU positive AT2 cells. n = 6 organoids/group. F: representative confocal images of organoids for tracking AT2 lineage. Organoids were fixed on day 17 and labeled with RAGE antibody (AT1 marker, red), SFTPC antibody (AT2 marker, green), and Hoechst dye (blue). Scale bar 100 mm. G: AT1/AT2 ratio. n = 9 organoids/group. H: representative confocal images (z stack of three planes) of organoids labeled with β-catenin antibody. I: mean fluorescent intensity (MFI) of β-catenin. n = 6 organoids/group. One-way ANOVA test for B, C, E, G, and I. Data expressed as means ± SE. AT2, alveolar type 2; DIC, differential interference contrast; EdU, 5-ethynyl-2-deoxyuridine; PAI-1, plasminogen activator inhibitor-1.

Wnt5a Regulates AT2 Organoid Formation in Serpine1^Tg Mice

PAI-1 regulates cell migration via the β-catenin (28). As a classic signaling pathway, the Wnt5a/β-catenin cascade regulates the hemostasis and regeneration of the alveolar epithelium (29–31). We postulated that the suppression of AT2 lineage in vivo and Serpine1^Tg organoids could be regulated by the Wnt5a/β-catenin pathway. Indeed, the β-catenin expression in AT2 organoids was moderately increased in Serpine1^Tg organoids. PAI-1 neutralizing antibody prevented the elevation in the expression of β-catenin in differentiated organoids (5.35 ± 0.08 for Tg group vs. 3.76 ± 0.31 for Tg+ab group, P = 0.0005, Fig. 4, H and I). Our results suggest that Wnt5a/β-catenin axis activated by PAI-1 may be involved in the AT2 fate.

To characterize the regulation of AT2 fate by the Wnt5a/β-catenin axis further, we treated both proliferating and differentiated cultures with Wnt5a and a specific inhibitor (Fig. 5A) based on the optimized dose for Box5 (Supplemental Fig. S2). Wnt5a but not Box5 resulted in a mild reduction in organoid number in proliferating organoids (54.33 ± 3.88 for Wnt5A group vs. 62 ± 1.67 for Box5 group, P = ns, Fig. 5B). Opposite effects on organoid size were observed between Wnt5a and Box5 (192.7 ± 9.55 for the Wnt5A group vs. 525.5 ± 25.37 for those treated with Box5, P < 0.0001 Fig. 5C). Similar to the proliferating cultures, Wnt5a and Box5 did not alter the colony number of differentiated organoids significantly (53.56 ± 3.63 for Wnt5A vs. 62.11 ± 1.91 for Box5, P < 0.0001, Fig. 5D); instead, the apparent size was down and upregulated by Wnt5a and Box5, respectively (216.1 ± 9.8 for Wnt5A vs. 333.5 ± 15.83 for Box5, P = ns Fig. 5E). This was accompanied by a consistent change in AT1/AT2 ratio (0.84 ± 0.02 for Wnt5A vs. 1.1 ± 0.05 for PBS group, P = 0.002 and 1.57 ± 0.15 for Box5 vs. 1.1 ± 0.05 for PBS, P = 0.01 Fig. 5G) and EdU⁺ AT2 cells (40 ± 2.51 for Wnt5A vs. 49.58 ± 2.88 for PBS, P = 0.03 and 60.68 ± 2.68 for Box5 vs. 49.58 ± 2.88 for PBS, P = 0.01 Fig. 5, H and I) in proliferating organoids. Surprisingly, very few EdU⁺ cells were seen in differentiated organoids (Supplemental Fig. S3). In addition, inhibition of the Wnt5/β-catenin signaling pathway by Box5 significantly suppressed the expression of β-catenin in differentiated organoids. Intriguingly, Wnt5 reduced the β-catenin expression moderately (4.85 ± 0.22 for Wnt5A vs. 5.59 ± 0.10 for PBS, P = 0.03 and 4.18 ± 0.21 for Box5 vs. 5.59 ± 0.10 for PBS, P = 0.0002, Fig. 5, J and K). These data show that the lineage of Serpine1^Tg AT2 cells is most likely regulated by the activated Wnt5a/β-catenin axis.

Figure 5.

Wnt5a/β-catenin axis regulates the lineage of AT2 cells. Feeder-free Serpine1^Tg organoids were treated with Wnt5a (200 ng/mL) and its inhibitor Box5 (200 µM) for up to 17 days, as shown in the inset of A. A: DIC images of proliferating (left three images) and differentiated cultures (right three images). The images of proliferative organoids and differentiated cultures were captured on day 10 and day 17, respectively. Scale bar, 1 mm. B: number of proliferative organoids. n = 9 transwell inserts per group. C: organoid size (diameter in µm). n = 9 transwell inserts per group. D: number of differentiated organoids. n = 9 transwell inserts per group. E: organoid size (diameter in µm). n = 9 transwell inserts per group of differentiated organoids. F: representative confocal images of organoids. Organoids were fixed after 17 days of culture and labeled with RAGE antibody (AT1 marker, red), SFTPC antibody (AT2 marker, green), and nuclei were stained with Hoechst dye (blue). Scale bar, 100 µm. G: quantitative analysis of AT1/AT2 ratio. n = 7 organoids/group. H: representative confocal images of EdU-labeled (green) AT2 cells in organoids and nuclei were stained with Hoechst dye (blue). All planes were Z stacked. Scale bar, 100 µm. I: quantification of EdU positive proliferating AT2 cells. n = 9 organoids/group. J: representative images of β-catenin antibody-incubated organoids (Z stacks of three planes). K: quantitative analysis of mean fluorescence intensity (MFI) of β-catenin. n = 6 organoids/group. One-way ANOVA for B–E, G, I, and K. AT2, alveolar type 2; Box5, Wnt5a-derived N-butyloxycarbonyl hexapeptide; DIC, differential interference contrast; EdU, 5-ethynyl-2-deoxyuridine; Wnt5a, Wnt family member 5A.

DISCUSSION

The objective of this study is to track the lineage of AT2 cells overexpressing human PAI-1 in 3-D cultures. The results show that elevated PAI-1 reduces AT2 cells pronouncedly in vivo and in vitro, in agreement with decreased EdU⁺ cells in organoids. The suppression of AT2 self-renewal may not depend on the presence of mesenchymal cells (fibroblasts). A significant reduction in AT1/AT2 ratio is associated with Serpine1^Tg organoids, which is due to maturation-dependent AT2-AT1 differentiation. The specific inhibition of PAI-1 on the lineage of AT2 cells is confirmed using a neutralizing antibody. The reduced AT1/AT2 ratio is accompanied by a dramatic morphological change, including more irregular septa with thicker walls within the Serpine1^Tg organoids. Mechanistic studies suggest that the increase in PAI-1 level could increase the expression of β-catenin signaling pathway (Fig. 6). Our study demonstrates that elevated PAI-1 can serve as autocrine signaling to regulate the lineage of AT2 cells in injured lungs.

Figure 6.

Schematic summary of the mechanisms for the regulation of AT2 lineage by PAI-1. Fewer alveolar AT2 cells were isolated from Serpine1^Tg mice compared with wt controls. Serpine1^Tg organoids were characterized by smaller size, multisepta, and reduced AT1/AT2 ratio. The Wnt5/β-catenin axis is involved in the regulation of AT2 lineage by PAI-1. AT2, alveolar type 2; PAI-1, plasminogen activator inhibitor-1; Wnt5a, Wnt family member 5A.

The downregulation of AT2 proliferation by PAI-1 is supported by our previous report in Plau^−/− mice (16). In injured lungs, elevated PAI-1 is associated with a reduced uPA level and fibrinolytic activity. Elimination of soluble PAI-1 by neutralizing antibodies indicates PAI-1 could regulate the AT2 lineage extracellularly. Restoring the increased PAI-1 level in transgenic mice will lead to a normal/physiological PAI-1 level that was observed in wild-type controls. Thus, the lineage of AT2 cells in vivo could be identical to that in wt lungs. This reason is further supported by the AT2 yield in vivo. This is similar to the mechanism for uPA to regulate the proliferation and differentiation of AT2 cells (16). PAI-1-mediated regulation of AT2 fate is not altered in the absence of fibroblasts. This is consistent with a previous wound healing assay of AT2 cells (15).

We identify the role of the Wnt5a/β-catenin signal pathway. Our results cannot exclude the senescence and apoptosis of AT1 and AT2 cells. As a biomarker of senescence, PAI-1 favors apoptotic challenge in aged AT2 cells by enhancing the p53, p21, and bcl-2 associated X-protein (BAX) protein expression (11). This possibility may explain the difference between sftpc⁺ and EdU⁺ cells in Serpine1^Tg organoids; PAI-1 may lead to a significant increase in senescent AT2 cells (sftpc⁺EdU⁻ cells).

AT1 cells could program into AT2 cells in acute injured neonatal lungs, characterized by an increment in PAI-1 level (32). This reverse lineage in developing lungs is regulated by the Hippo signaling pathway. In adult lungs, this AT1-AT2 lineage has been described in the compensatory growth postpneumonectomy (33). The loss of YAP/TAZ is able to reprogram AT1 into AT2 cells either in normal lungs or those challenged by hypoxia (34). The question raised is whether AT1-AT2 lineage contributes to reduced AT1/AT2 ratio in Serpine1^Tg organoids. Further study is required to address this issue.

Epithelial-mesenchymal transition (EMT) is well described for human and mouse AT2 cells in bleomycin and smoke inhalation injured lungs (35, 36). This process is accompanied by a pronounced increase in PAI-1 and a reduced uPA expression (35, 36). Theoretically, cells in feeder-free organoids should express rage and sftpc markers. EMT and AT2-basal cell programming will lead to loss of epithelial markers, as seen in rage⁻sftpc⁻ cells in organoids. In addition, AT2 cells can be driven by aberrant mesenchyme into a basal cell lineage in severely injured human lungs (37). These possibilities could partially explain the reduction in the organoid size of Serpine1^Tg organoids.

Elevated PAI-1 is associated with mortality and severity of acute lung injury, including ARDS and COVID-19. Our study provides novel mechanisms for PAI-1 to suppress the AT2 lineage for re-alveolarization. As demonstrated by the PAI-1 neutralizing antibody, targeting PAI-1 may be a promising pharmaceutic strategy for re-epithelialization in injured lungs and other disorders accompanied by hyperactive PAI-1. In a nutshell, PAI-1 downregulates AT2 lineage via multifaceted mechanisms.

DATA AVAILABILITY

All data generated or analyzed during this study are included in this published article and its supplemental material.

SUPPLEMENTAL DATA

Supplemental Tables S1 and S2 and Supplemental Figs. S1–S3: https://doi.org/10.6084/m9.figshare.20134106.

GRANTS

This work was funded by NIH Grants HL134828, AI150550, and LM013460.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

H.-L.J. conceived and designed research; K.G.J. and R.Z. performed experiments; K.G.J. and R.Z. analyzed data; K.G.J., R.Z., and G.Y. interpreted results of experiments; K.G.J. and Y.L. prepared figures; K.G.J., X.G., and H.-L.J. drafted manuscript; R.Z., X.G., G.Y., and H.-L.J. edited and revised manuscript; K.G.J., R.Z., Y.L., X.G., G.Y., and H.-L.J. approved final version of manuscript.