Multilineage murine stem cells generate complex organoids to model distal lung development and disease

Abstract

Organoids derived from mouse and human stem cells have recently emerged as a powerful tool to study organ development and disease. We here established a three-dimensional (3D) murine bronchioalveolar lung organoid (BALO) model that allows clonal expansion and self-organization of FACS-sorted bronchioalveolar stem cells (BASCs) upon co-culture with lung-resident mesenchymal cells. BALOs yield a highly branched 3D structure within 21 days of culture, mimicking the cellular composition of the bronchioalveolar compartment as defined by single-cell RNA sequencing and fluorescence as well as electron microscopic phenotyping. Additionally, BALOs support engraftment and maintenance of the cellular phenotype of injected tissue-resident macrophages. We also demonstrate that BALOs recapitulate lung developmental defects after knockdown of a critical regulatory gene, and permit modeling of viral infection. We conclude that the BALO model enables reconstruction of the epithelial-mesenchymal-myeloid unit of the distal lung, thereby opening numerous new avenues to study lung development, infection, and regenerative processes in vitro.

Keywords: BALO; BASC; lung organoids; stem cells.

Figure 1. 3D co‐culture of BASC and rMC results in the formation of BALO

1. A

   Gating strategy for sorting of EpCAM^highCD24^lowSca‐1⁺ cells and rMC from lung homogenate of adult mice.

2. B

   Time course of epithelial stem/progenitor cell proliferation and differentiation within BALO at days 8, 11, and 21 of co‐culture with rMC.

3. C

   Clonal expansion of EpCAM^highCD24^lowSca‐1⁺ cells derived from either GFP‐expressing mice or tdTomato‐expressing mice at day 21 of co‐culture.

4. D

   Percentages of SCGB1A1⁺SFTPC⁺ (BALOs), SCGB1A1⁺SFTPC⁻ (bronchiolospheres), and SCGB1A1⁻SFTPC⁺ (alveolospheres) organoids per well at day 21 of culture derived from EpCAM^highCD24^lowSca‐1⁺ cells isolated from Scgb1a1 ^mCherry Sftpc ^YFP reporter mice (n = 4 biological replicates).

5. E

   Representative confocal images of days 8–21 of culture showing endogenous SCGB1A1 and SFTPC expression during BALO formation derived from EpCAM^highCD24^lowSca‐1⁺SCGB1A1⁺SFTPC⁺ cells isolated from Scgb1a1 ^mCherry Sftpc ^YFP reporter mice.

6. F–H

   Heat map (left) and tSNE plot (right) (F) of digested day 21 BALO cultures depicting four distinct clusters (C1, airway, blue; C2, alveolar, purple; C3, MYO, orange; and C4, LIF, green). Violin plots of selected genes representing airway‐associated genes (G) (Itgb4, Trp63, Krt7, and Sox2) and alveoli‐associated genes (H) (Cxcl15, Lyz2, Sftpc, and Hopx). Each violin plot shows the frequency distribution of the mean transcript level (log₂).

Data information: Scale bars = 100 μm (B, C right, E) or 500 μm (C left). Bar charts presented as the mean ± SEM.

Figure EV1. 3D reconstruction of BALO depicts proximo‐distal cell specification

1. A, B

   Representative confocal picture (A) and 3D reconstruction (B) of a day 21 culture showing endogenous SCGB1A1 and SFTPC expression within BALO derived from EpCAM^highCD24^lowSca‐1⁺SCGB1A1⁺SFTPC⁺ cells isolated from Scgb1a1 ^mCherry Sftpc ^YFP reporter mice. Scale bars represent 50 μm.

Figure 2. The BALO model mimics the 3D morphology and cellular composition of the bronchioalveolar compartment with proximo‐distal patterning and pulmonary surfactant secretion

1. A

   mRNA expression analysis of epithelial cell differentiation markers Sftpc (AEC II), Hopx (AEC I), Foxj1 (ciliated cells), Muc5ac (secretory cells), and p63 (basal cells) in BALO at days 0, 10, and 21 of culture (n = 3 biological replicates with pooled cells from 4 cultures per replicate).

2. B

   Electron microscopy of BALO alveoli showing two cuboidal epithelial cells (AEC II) connected by tight junctions (red square) with numerous mitochondria (M) and lamellar bodies (LB). A LIF with numerous lipid droplets (*) and a MYO with cisterns of rough ER (red arrow) are located at the basal side (left). The alveolar lumen is filled with lamellar surfactant including tubular myelin (white arrow) (lower right). Exocytosis of a lamellar body (*) from an AEC II (upper right). Scale bars indicate 500 nm.

3. C

   Electron microscopy of bronchiolar‐like airway depicting columnar ciliated cells (red arrowhead) with basal bodies (black arrowhead) at the left side of the longitudinally sectioned lumen. The boxed area indicates an apical junctional complex between two ciliated cells: 1 = tight junction and 2 = adherens junction. Scale bar indicates 500 nm (in insert: 100 nm).

4. D

   Staining of lamellar bodies in BALO with RFP LysoTracker. Scale bars represent 100 μm.

5. E

   Western blot analysis of the surfactant proteins: pro‐SPC, mature SPC, pro‐SPB, mature SPB, and SPA in lung homogenate (LH), AEC, and day 21 BALO (n = 3 biological replicates).

6. F, G

   Alveolar diameter (F) and number of alveoli (G) in tdTomato⁺ BALO at days 15, 21, 30, and 40 of culture were measured from n = 5 BALOs in n = 3 biological replicates BALOs.

7. H, I

   Representative scheme and images of day 40 BALO alveoli (H) and airway (I). BALO alveolar‐like structures (H) are shown (h′) (red arrowheads) in semi‐thin section (0.5 μm) stained with Toluidine blue. Scale bar indicates 100 μm (left). Electron microscopy showing AEC I (h″) ultrastructure within BALO. Scale bars indicate 2,500 nm (center) and 1,000 nm (right). An airway‐like structure (i′) is shown with secretory and ciliated cells (red arrowheads) in semi‐thin sections (0.5 μm) longitudinally cut and stained with Toluidine blue. Scale bar indicates 50 μm (far left). Electron microscopy of a bronchiolar‐like airway (I) depicting pseudostratified epithelium (i″) with a basal‐like cell (BC), not reaching the lumen in which cilia (C) are seen, located between a secretory (SC) and a ciliated cell (Ci). Scale bars indicate 1,000 nm (left and right). Mature cilia (i‴) in BALO at higher magnification depicting the 9 × 2 + 2 structure with central doubled microtubules (insert, red arrowhead), a characteristic for motile cilia. Scale bar indicates 100 nm (far right).

Data information: Bar and dot charts presented as the mean ± SEM and probability determined using one‐way ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001).

Figure EV2. BALOs express markers of differentiated airways and alveoli

1. A, B

   Representative scheme and confocal images of the BALO alveolar‐like (A) and airway‐like (B) structures. (A) Representative image of AEC II SFTPC staining (white asterisks) and AEC I endogenous PDPN expression in day 28 BALO isolated from the lungs of Pdpn ^GFP reporter mice. Scale bars represent 25 μm. Dotted lines indicate the alveoli. (B) Representative fluorescence images of BALO's airway‐like structures stained for β‐IV tubulin⁺ ciliated cells (red arrowheads) and SCGB1A1⁺ club cells. Scale bars represent 25 μm (left and middle) and 5 μm (right).

Figure 3. Distinct subsets of EpCAM⁻Sca‐1⁺ rMC‐derived fibroblasts are indispensable for BALO growth, cell differentiation, and branching morphogenesis, and model the mesenchymal niche of the lung

1. Representative images of day 21 BALO and tdTomato⁺ rMC stained with LipidTOX (green). Scale bar represents 100 μm.

2. Fluorescence images of αSMA (green) and neutral lipids (LipidTOX red) staining in WT‐derived rMC and BALO (dotted lines indicate single BALO or insert) at day 21 of culture. Scale bars represent 50 μm (left) and 25 μm (right).

3. Representative flow cytometric dot plots and histograms of PDGFRα expression in rMC (EpCAM⁻CD45⁻CD31⁻Sca‐1⁺) isolated from the lung homogenate of Pdgfra ^GFP reporter mice.

4. Representative images of BALO formation at day 8, 15, and 21 of co‐culture. WT BASCs were co‐cultivated with sorted rMC (total PDGFRα+ population) or rMC expressing either low or high levels of PDGFRα‐GFP. Scale bars represent 100 μm.

5. tSNE plots and violin plots depicting selected genes representing MYO (Pdgfrα, Tagln, Acta2, Eln, and Axin2) (top panels) and LIF‐associated genes (Fgf10, Apoe, Serpina3n, Gsn, and Gas6) (bottom panels). Each violin plot shows the frequency distribution of the mean transcript level (log₂). C3 (MYO, orange) and C4 (LIF, green) refer to the scRNA‐Seq experiment in Fig 1G.

6. Fluorescence image of αSMA⁺PDGFRα^high MYO (yellow arrows) and LipidTOX⁺PDGFRα^low LIF (red arrows) from sorted PDGFRα‐GFP rMC after 21 days of BALO culture. Scale bars represent 50 μm (left) and 25 μm (right).

7. Fluorescence image containing PDGFRα^high MYO from sorted PDGFRα‐GFP rMC within BALO derived from tdTomato⁺ mice at day 21 of culture. Scale bars represent 100 μm (left) and 50 μm (right).

Figure 4. TR‐Mac engraft into BALO alveolar‐like regions, maintain their identity, and drive epithelial cell differentiation

1. Gating strategy to define TR‐Mac from BAL of adult tdTomato⁺ mice.

2. Representative images of BALO after microinjection of tdTomato⁺ TR‐Mac at day 14. TR‐Mac are preferentially found in alveoli (right). Scale bars represent 100 μm (left) and 50 μm (right).

3. Fluorescence confocal images of CD206, Siglec‐F, and isotype control staining 14 days after tdTomato⁺ TR‐Mac microinjection in a day 28 BALO. Scale bars represent 100 μm (left) and 25 μm (right).

4. Electron microscopy depicting filopodium of TR‐Mac (white dashed lines) with a characteristic actin filament bundle (yellow background) (left panel) in contact with AEC I within BALO alveolar‐like structures. Uptake of surfactant by TR‐Mac is depicted by a (*). Scale bar indicates 1,000 nm (left, in insert: 500 nm) and 500 nm (right, in insert: 250 nm).

5. Representative confocal images of Cx43 staining in tdTomato⁺ TR‐Mac monoculture in Matrigel and tdTomato⁺ TR‐Mac microinjected at day 14 and analyzed in a mature day 21 BALO. Scale bars represent 50 μm (overview) and 5 μm (close up).

6. Heat map (left) and tSNE plot (right) of the comparative analysis of digested day 23 BALO cultures with (+) and without (−) microinjected TR‐Mac depicting six distinct clusters (C1, club/secretory cells, red; C2, basal cells, yellow; C3, rMC, green; C4, AEC II, blue; C5, AEC I, purple; and C6, ciliated cells, pink).

7. Expression data dot plots of genes found differentially regulated in cluster C1 between day 23 BALO with (C1+) and without (C1−) microinjected TR‐Mac. The circle size illustrates the number of cells expressing a specific gene.

8. Percentage of terminally differentiated cells (AEC I and ciliated cells) in day 23 BALO cultures with (+) and without (−) microinjected TR‐Mac.

Data information: Bar and dot charts presented as the mean ± SEM and probability determined using t‐test (**P < 0.01).

Figure 5. Manipulation of WNT signaling pathway during BALO morphogenesis by knockdown of miR‐142‐3p gene expression recapitulates developmental defects observed in embryonic lung explants

1. A

   Representative images of E11.5 lung explants after treatment for 48 h with 100 μM Scra or mo142‐3p reveal reduced size and branching morphogenesis after miR‐142-3p knockdown.

2. B

   Organoid diameter (n = 30–50 per group) in μm before addition of 4 μM Scra or mo142‐3p at day 6 (control) and 5 days after treatment (at day 11 BALO culture) in n = 3 biological replicates.

3. C, D

   Representative images of β‐galactosidase staining in TOPGAL epithelium (C) and TOPGAL rMC (D) BALO cultures before (day 6, control) or 5 days after treatment with either 4 μM Scra or mo142‐3p (day 11 of co‐culture). BASCs and rMC were isolated from the lung homogenate of TOPGAL mice. β‐galactosidase⁺ rMC at day 11 of culture are indicated with arrows.

4. E

   Representative transmission and confocal images after LysoTracker staining indicating branching and number of branching points in n = 4 BALOs 15 days after treatment with 4 μM Scra or mo142‐3p (day 21 of co‐culture) in n = 3 biological replicates.

5. F

   mRNA levels of epithelial and mesenchymal miR‐142-3p expression in Scra and mo142‐3p-treated organoids 5 days after treatment (n = 3–4 biological replicates with pooled cells from 4 cultures per replicate).

Data information: Scale bars = 100 μm. Bar charts presented as the mean ± SEM and probability determined using t‐test (*P < 0.05, **P < 0.01).

Figure 6. BALOs support influenza virus infection and allow modeling of lung infection and injury

1. Representative images of day 21 BALO after microinjection of SC35M‐GFP IAV (green) into the central airway‐like (*) after 0 and 12 h pi visualize IVA spread to the alveolar‐like regions. Dotted lines illustrate BALO borders. Scale bars represent 100 μm.

2. Quantification of plaque‐forming units in supernatants from mock or SC35M IAV‐infected BALO 48 h after infection (n = 5 biological replicates). N.D.: not detectable.

3. Relative NP expression in PR8 IAV‐infected (HA⁺) or non‐infected (HA⁻) epithelial cells isolated from BALOs 48 h pi, FACS‐sorted according to EpCAM and HA expression (n = 3 biological replicates with pooled cells from 4 cultures per replicate).

4. Representative flow cytometric data showing the percentage of EpCAM⁺NP⁺ cells in mock‐ and SC35M IAV‐infected BALOs at 48 h pi in n = 3 biological replicates with pooled cells from 4 cultures per replicate.

5. Representative fluorescence images of a day 21 distal region in BALO generated from mTmG reporter mouse and infected by SC35M‐Cre IAV after 0, 10, 14, and 25 h. Arrows indicate cell death. Scale bars represent 25 and 10 μm in the insert.

6. mRNA expression of Ifnb in mock and PR8 IAV‐infected BALOs at 48 h pi (n = 3 biological replicates with pooled cells from 4 cultures per replicate).

7. Release of TNF‐α (6 h pi), IL‐6 (48 h pi), and IL‐1β (48 h pi) detected by Bio‐Plex^® Multiplex immunoassay in the supernatant of mock and IAV‐infected BALO cultures with and without TR‐Mac (n = 3 biological replicates).

Data information: Bar charts presented as the mean ± SEM and probability determined using t‐test (F) and one‐way ANOVA (G) (*P < 0.05, **P < 0.01).