Single-cell Wnt signaling niches maintain stemness of alveolar type 2 cells

Abstract

Alveoli, the lung's respiratory units, are tiny sacs where oxygen enters the bloodstream. They are lined by flat alveolar type 1 (AT1) cells, which mediate gas exchange, and AT2 cells, which secrete surfactant. Rare AT2s also function as alveolar stem cells. We show that AT2 lung stem cells display active Wnt signaling, and many of them are near single, Wnt-expressing fibroblasts. Blocking Wnt secretion depletes these stem cells. Daughter cells leaving the Wnt niche transdifferentiate into AT1s: Maintaining Wnt signaling prevents transdifferentiation, whereas abrogating Wnt signaling promotes it. Injury induces AT2 autocrine Wnts, recruiting "bulk" AT2s as progenitors. Thus, individual AT2 stem cells reside in single-cell fibroblast niches providing juxtacrine Wnts that maintain them, whereas injury induces autocrine Wnts that transiently expand the progenitor pool. This simple niche maintains the gas exchange surface and is coopted in cancer.

Figure 1. Wnt pathway gene Axin2 marks a rare population of AT2 cells with stem cell activity

(a–c) Section through alveoli of adult (2 month old) Axin2-CreERT2;Rosa26-mTmG mouse lung immunostained for Cre recombinase reporter mGFP, AT2 cell marker pro-surfactant protein C (SftpC), and DAPI 5 days after 3 daily injections of 3 mg tamoxifen (3d "pulse") to induce CreERT2 and lineage label Axin2-expressing (Axin2⁺) cells (Axin2>GFP, green). Most AT2 cells ("bulk" AT2 cells) are Axin2^- (box b in a, close up in b), but rare ones are Axin2⁺ (box c in a, close up in c) indicating activation by a Wnt signal ("Wnt-active"). (d) FACS analysis of AT2 cells from lungs as above. One percent of AT2 cells (1.0 ±0.5%, n=3 biological replicates) express Axin2>GFP. (e) Alveolar section from lungs lineage labeled as above but harvested one year after labeling ("1 yr chase"). Note expansion of Axin2-lineage AT2 cells (arrowheads) as well as AT1 cells and fibroblasts (arrows), the latter arising from a different Axin2⁺ cell lineage (Nabhan et al, unpublished). (f–h) Close-ups of lungs as in e showing Axin2⁺ lineage-labeled AT2 cells in mitosis (f) and after generating another AT2 cell (g, arrowheads) or AT1 cells (h, arrows). AT2 marker in f is apical surface protein Muc1 (red). AT1 marker in h is membrane protein RAGE (white). (i) Quantification of percent AT2 cells that are Axin2-lineage labeled immediately following a 3d pulse as above or a 3 week (3wk) pulse (3d pulse each week for 3 weeks) at age 2 months, a 3d pulse at 4 months, or after a 3d pulse at 2 months followed by 1 year chase ("1yr chase"). Values shown are mean ±SD (n=3500 AT2 cells scored in 2–4 biological replicates for each condition). (j,k) Alveolar sections of Axin2-CreERT2; Rosa26-Rainbow mice given limiting dose (2 mg) of tamoxifen at age 2 months to label isolated Axin2+ cells (j) with one of several Rainbow colors, and immunostained for SftpC (red) 1 week (j) or 6 months (k) later. (j) Isolated Axin2+ AT2 cell expressing mOrange clone marker (yellow) 1 week after labeling. (k) Four-cell clone 6 months later (k). (l) Quantification of AT2 clone sizes 1 week and 6 months after labeling. ***, p<0.001 (t test). Scale bars: 25 µm (a,e), 5 µm (b,c,fh), 20 µm (k).

Figure 2. Single, Wnt-secreting fibroblasts comprise the AT2 stem cell niche

(a,b) Alveolar sections of 2 month old adult Sftpc-CreER; Rosa26-mTmG lungs 1 week after induction with tamoxifen to label AT2 cells (green), and immunostained for Wnt secretion protein Porcupine (PORCN, white). Note rare stromal cells (non-AT2 cells, TdTomato+, red, arrow), but noAT2 cells (GFP+, green, arrowheads), expressing PORCN, some contacting AT2 cells (b). (c) Porcn mRNA levels determined by single cell RNAseq of adult lung fibroblasts and AT2 cells. A subpopulation of Pdgfrα-expressing fibroblasts express Porcn. Blue, DAPI. (d) Effect of 5 daily doses of PORC inhibitor C59 (+) or vehicle control (-) on Axin2-expresssion in AT2 cells at age 2 months, measured by multiplex single molecule proximity ligation in situ hybridization (PLISH) for Axin2 and Sftpc mRNA. Values shown are percentage (mean ±SD; of Sftpc-expressing AT2 cells that co- express Axin2 (n=900 AT2 cells scored in 3 biological replicates for each condition). ***, p=0.001 (t test). (e, f) Effect on Axin2-expresssion in AT2 cells by inhibiting fibroblast secretion of Wnts by conditional deletion of Wntless using lung mesenchyme driver Tbx4^LME-Cre (e) or Pdgfrα-CreERT2 induced with 3 mg tamoxifen 3 days before analysis (f). *, p=0.02; ***, p=0.003 (t test). The Axin2⁺ AT2 cells remaining in the Wntless deletion conditions could be due to incomplete deletion (or perdurance) of Wntless^fl from fibroblasts, a Wnt signal provided by another cellular source (see below), or a Wntless-independent signal. (g) Expression of the 19 Wnt genes, fibroblast markers Pdgfrα and Col1a2, Axin2, and three ubiquitously-expressed control genes (Ubc, Ppla, Actb) (rows) in 47 alveolar fibroblasts (columns) isolated from wild type B6 adult lungs and analyzed by single cell RNA sequencing. Note subpopulation with robust expression of Wnt5a and lower expression of several other Wnt genes, and sporadic cells expressing other Wnt genes. Many Wnt5a-expressing fibroblasts co-express Pdgfrα. (h) Expression of Wnt5a, Axin2, and AT2 marker SftpC detected by PLISH in alveolar section of adult (2 month old) lung. Blue, DAPI. Note rare Wnt5a-expressing cell (box j). (i, j) Close up of boxed regions in h showing AT2 cell far from (i) or near (j) a Wnt5a-expressing cell. AT2 cell near Wnt5a source expresses Axin2 (green), indicating it is activated by the secreted signal; insets in i and j show Axin2 channel of the AT2 cell. (k) Quantification of data as in d–f showing percentage (mean ±SD) of AT2 cells located far from a Wnt5a source (>15 um, n=132 cells from 3 biological replicates) or near a Wnt5a source (<15 um, n=150 cells) that express Axin2. ***, p<0.0001 (t test). (l) Axin2 expression in AT2 cells isolated from adult Axin2-lacZ mice and cultured for 5 days with indicated Wnt proteins (Wnt5a, Wnt3) or antagonist Dickkopf3 (Dkk3) at 1ug/mL, then assayed with fluorogenic LacZ (beta-galactosidase) substrate Spider-gal. ***, p<0.0001 (t test). Scale bars: 10 µm (a, b, h) and 5 µm (j).

Figure 3. Wnt signaling maintains AT2 stem cells by preventing reprogramming to AT1 fate

(a–d) Wnt regulation of reprogramming to AT1 fate. Alveolar sections of 8 month old adult Lyz2-Cre; Rosa26-mTmG (a, control), Lyz2-Cre; Rosa26-mTmG; β-catenin^fl/fl (b, AT2 cell loss of Wnt activity), and Lyz2-Cre; Rosa26-mTmG; β-catenin^Ex3/+ (c, AT2 cell constitutive Wnt activity) immunostained for AT2 marker SftpC (red) and Lyz2-Cre (AT2 marker) lineage trace (mGFP, green). Lyz2-Cre turns on in mature AT2 cells (16). Dashed circles, alveolar renewal foci identified by squamous AT1 cells that arise from AT2 stem cells and express AT2 lineage trace. Note increased reprogramming to AT1 fate when β-catenin is deleted to eliminate Wnt signaling (b), and decreased reprogramming when β-catenin exon3 (Ex3) is deleted to constitutively activate Wnt signaling (c). Scale bar, 50 µm. Quantification (d) shows percent (mean ±SD) of alveoli with AT2 lineage-labeled AT1 cells (n=25 100µm z-stacks scored in 2 or 3 biological replicates for each condition). ***, p=0.002 (Kruskal-Wallis test). (e,f) Deletion of β-catenin results in loss of AT2 stem cells during reprogramming to AT1 fate. Close up of renewal foci (e) as above in wild type β-catenin control (Lyz2-Cre; β-catenin^+/+, upper panels) and β-catenin conditional deletion (Lyz2-Cre; β-catenin^fl/fl, bottom panels) lungs. Note AT1 cell (arrow) and its AT2 parent cell (open arrowhead) in control focus, and absence of AT2 parent cell (*) in β-catenin-deleted focus, implying loss of stem cell by direct reprogramming (without self renewal) to AT1 fate. Quantification (f) shows percentage (mean ±SD) of AT1 cells derived from AT2 lineage (GFP⁺) that lack a parent AT2 cell. ***, p=0.0004 (t test). (g–i) Wnt signaling prevents reprogramming to AT1 fate in culture. AT2 cells isolated from wild type (B6) adult mouse lungs were cultured under AT2-maintaining conditions (Matrigel, minimal media, 1% FBS, FGF-7; g,h) or AT1 promoting conditions (poly-lysine coated glass, minimal media, 10% FBS; i) without (control) or with the indicated Wnt pathway antagonists (0.15 ug/ml Dkk3) or agonists (0.1 ug/ml Wnt5a or 10 nM CHIR99201, a canonical Wnt pathway activator). After 4 days of culture, cells were immunostained for SftpC (red) and Podoplanin (green) (g) and the percentage (mean ±SD) of AT2 cells (small, cuboidal SftpC⁺; arrowheads) and AT1 cells (large, squamous, Podoplanin⁺; arrow) quantified (n=500 cells from 3 biological replicates for each culture condition) (h,i) **, p=0.002; ***, p<0.001 (t test). (j, k) AT2 cells isolated from 2 month old Sftpc-CreER; Rosa26-mTmG adults were cultured as above (AT2-maintaining conditions, g) without (control) or with indicated Wnt proteins (at 100 ng/mL) and/or EGF (at 50 ng/mL), then proliferation measured by incorporation of synthetic nucleotide EdU. Quantification in k (n= 400 cells scored per condition, 4 biological replicates). * p=0.007, ***; p<0.001 (t test). n.s., not significant. Scale bars: 50 µm (c, g), 10 µm (e), 5 µm (j).

Figure 4. Epithelial injury induces ancillary Axin2-expressing AT2 cells in alveolar repair

(a–d) System for genetically-targeted acute epithelial injury. Shh-Cre drives expression of Diptheria toxin receptor (DTR) and mGFP transgenes throughout epithelium; sporadic epithelial cell death ("ablation") is induced by injecting limiting dose (150 ng) of Diptheria toxin (DT). mGFP (upper panels) and mTomato (lower panels) micrographs of the same dual-labeled alveolar section of adult Shh-Cre; Rosa26-LSL-DTR/Rosa26-mTmG lungs18 hours after vehicle (a, control) or DT injection (b), or 5d after DT injection without (c) or with (d) Porcn inhibitor C59 to block Wnt secretion. Disruption of alveolar epithelium (mGFP, upper panels) is apparent at 18 hrs after DT injection (b, asterisks), and epithelial repair nearly complete at 5 days (c). Blocking Wnt secretion inhibits repair (d, asterisks). Other alveolar tissues (mTomato, lower panels) remain intact. (e–g) Effect of epithelial injury on proliferation and Wnt signaling in AT2 cells. Alveolar sections of Shh-Cre; Rosa26-LSL-DTR animals injected with vehicle (control, upper panels) or DT (ablation, lower panels) as above and immunostained for AT2 marker SftpC (red) and proliferation marker Ki67 (white) (e), or probed by PLISH for SftpC (red) and Axin2 (green) expression (f) five days after injury. DAPI, blue. Quantification (g) (n=250 cells in 4 animals for each condition) of percentage of AT2 cells that express Ki67 (mean ±S.D, upper plot) or Axin2 (mean ±S.D, lower plot). Epithelial ablation induces proliferation and Wnt pathway activation in most AT2 cells, and both effects are abrogated by Wnt secretion inhibitor C59. ***, p<0.001 for each comparison indicated (t test). Scale bars: 25 µm (d), 10 µm (e) 5 µm (f).

Figure 5. Epithelial injury induces autocrine Wnt signaling in bulk AT2 cells

(a) Alveolar sections immunostained as indicated from 2 month old Shh-Cre; Rosa26-LSL-DTR/Rosa26-mTmG mouse 3 days after DT injection to induce sporadic epithelial ablation. Note AT2 cells (cuboidal GFP+ cells), but not other cells, express Porcupine (PORCN, white). (b,c) Quantifications showing percentage of PORCN⁺ cells that are AT2 cells, before and 3d after ablation (b) (n=200 scored PORCN+ cells in 3 mice per condition), and percentage (mean ±SD) of AT2 cells expressing PORCN at indicated time after ablation (c) (n=500 scored AT2 cells in 4 mice per time point). (d) Alveolar sections of Shh-Cre; Rosa26LSL-DTR animals probed by PLISH for Wnt7b (green) and SftpC (red) 1 day after injection with vehicle (Control) or DT (Ablation). DAPI, blue. Ablation induces Wnt7b expression by AT2 cells. (e) Quantification showing time course of induction of Wnt7b expression by AT2 cells following ablation (n=300 AT2 cells scored per animal, 4 biological replicates per time point, mean ±S.D). ***, p<0.001 (t test) for both comparisons indicated. (f) mRNA induction of indicated Wnt genes and Wnt pathway target genes (Axin2, Lef1) measured by qRT-PCR of lineage-labeled AT2 cells sorted from mice before (control) and 2d after hyperoxic alveolar injury (5d at 75% O₂) (see scheme in Fig. S8b). Values shown are fold mRNA induction relative to control (mean ±SD, n=3 mice per condition). A suite of Wnt genes is induced in AT2 cells, most of them different from the Wnt genes expressed by the fibroblast niche cells under control conditions (highlighted red; see Fig. 2g). Dashed line, control value. *, p< 0.05, ***, p<0.001 (g,h) Function of autocrine Wnt signaling in AT2 cell proliferative response following hyperoxic alveolar injury. Alveolar section immunostained as indicated from adult (2 month old) Wntless^fl/fl mouse with alveolar epithelium infected with AAV9-GFP-Cre virus (Fig. S8d–f) to mosaically delete Wntless from AT2 cells (to prevent autocrine Wnt secretion) 1 week before hyperoxic injury. Note proliferation (Ki67 staining, red) of AT2 cells (Muc1 apical surface marker, white) induced by hyperoxic injury (see Fig. S8a), but not AT2 cell infected with AAV-Cre-GFP (green) to delete Wntless. (h) Quantification of g, showing percentage of AT2 cells that were Ki67⁺ following hyperoxic alveolar injury to AAV9-Cre-GFP-infected control (Wntless^+/+) and Wntless^fl/fl mice, scored for all or just GFP+ (AAV9-Cre-GFP-infected) AT2 cells. Values shown are mean ±SD (n=500 AT2 cells scored per mouse, 3 mice per condition, except only 52 GFP+ cells were scored in Wntless^fl/fl because they were rare). Note almost complete elimination of proliferative response in GFP+ AT2 cells, implying key role for autocrine Wnt signaling in their proliferative response to injury. (j) Model of alveolar stem cells and their niche. During homeostasis (top), the niche is composed of a single fibroblast constitutively expressing Wnt5a and/or other Wnt genes (Fig. 2g) that provide a "juxtacrine" signal (arrow) to the neighboring AT2 cell (green cytoplasm indicating lineage trace label, black nucleus indicating Wnt signal reception and Axin2 induction), selecting and maintaining it as a stem cell (left panel). Upon receiving a mitogenic signal from a dying AT1 cell (middle panel), the activated stem cell proliferates and the two daughter cells (green) compete for the niche. One remains in the niche as the stem cell (stem cell renewal). The other leaves niche, losing the Wnt signal and disinhibiting the AT1 differentiation program (stem cell reprogramming) to form a new AT1 cell (right panel). Although bulk AT2 cells are normally quiescent (bottom left panel), following acute epithelial injury bulk AT2 cells are recruited as "ancillary stem cells" by induction of autocrine Wnt signaling by a different suite of Wnts including Wnt7b (middle panel), which allows unlimited proliferation in response to mitogenic signaling from dying cells. Autocrine Wnt signaling diminishes as the epithelial injury resolves (right panel). Scale bars: 10um (a, g), 5um (d).