In Vivo Generation of Lung and Thyroid Tissues from Embryonic Stem Cells Using Blastocyst Complementation

Abstract

Rationale: The regeneration and replacement of lung cells or tissues from induced pluripotent stem cell– or embryonic stem cell–derived cells represent future therapies for life-threatening pulmonary disorders but are limited by technical challenges to produce highly differentiated cells able to maintain lung function. Functional lung tissue–containing airways, alveoli, vasculature, and stroma have never been produced via directed differentiation of embryonic stem cells (ESCs) or induced pluripotent stem cells. We sought to produce all tissue components of the lung from bronchi to alveoli by embryo complementation.

Objectives: To determine whether ESCs are capable of generating lung tissue in Nkx2-1^−/− mouse embryos with lung agenesis.

Methods: Blastocyst complementation was used to produce chimeras from normal mouse ESCs and Nkx2-1^−/− embryos, which lack pulmonary tissues. Nkx2-1^−/− chimeras were examined using immunostaining, transmission electronic microscopy, fluorescence-activated cell sorter analysis, and single-cell RNA sequencing.

Measurements and Main Results: Although peripheral pulmonary and thyroid tissues are entirely lacking in Nkx2-1 gene-deleted embryos, pulmonary and thyroid structures in Nkx2-1^−/− chimeras were restored after ESC complementation. Respiratory epithelial cell lineages in restored lungs of Nkx2-1^−/− chimeras were derived almost entirely from ESCs, whereas endothelial, immune, and stromal cells were mosaic. ESC-derived cells from multiple respiratory cell lineages were highly differentiated and indistinguishable from endogenous cells based on morphology, ultrastructure, gene expression signatures, and cell surface proteins used to identify cell types by fluorescence-activated cell sorter.

Conclusions: Lung and thyroid tissues were generated in vivo from ESCs by blastocyst complementation. Nkx2-1^−/− chimeras can be used as “bioreactors” for in vivo differentiation and functional studies of ESC-derived progenitor cells.

At a Glance Commentary

Scientific Knowledge on the Subject

The regeneration or replacement of lung cells and tissues using pluripotent stem cells, embryonic stem cells (ESCs), and induced pluripotent stem cells represent future therapies for life-threatening pulmonary disorders but are limited by technical challenges to produce highly differentiated cells able to maintain lung function. We sought to produce all tissue components of the lung from bronchi to alveoli by embryo complementation with donor ESCs.

What This Study Adds to the Field

Using blastocyst complementation, we produced ESC-derived lungs and thyroids in Nkx2-1^−/− mouse embryos lacking these tissues. ESC-derived respiratory cell lineages were indistinguishable from endogenous lung cells based on gene expression signatures, cell surface proteins, and ultrastructural features. Nkx2-1^−/− chimeras can be considered for use as “bioreactors” for in vivo differentiation of patient-derived induced pluripotent stem cells into respiratory cell lineages for future cell therapies.

The lung is a remarkably complex organ that consists of more than 25 major resident cell types and provides the entire body with the oxygen needed for terrestrial life (1, 2). Lung structure and morphogenesis are highly conserved across mammalian species, and lung formation requires specific sequences of signaling and transcription-mediating interactions among a diversity of cells derived from endodermal, mesodermal, and ectodermal progenitors. The formation of the trachea and lung buds occurs approximately at Embryonic Day 9.5 after conception in mice and 3–6 weeks after conception in humans. Lung morphogenesis requires the expression of TTF1 (homeobox transcription factor NKX2-1) (1, 3). NKX2-1 is the earliest known marker of the lung field, being expressed in primitive respiratory epithelial progenitors located in the primordial lung buds. Later in development, embryonic NKX2-1 progenitors give rise to all respiratory epithelial lineages as the lung field expands to form the main bronchi and conducting airways during the process of branching morphogenesis (1, 2). The formation of gas-exchange units, termed “alveoli,” occurs before and after birth, when NKX2-1 progenitors differentiate into alveolar type 1 (AT1) and type 2 (AT2) epithelial cells, which are closely apposed to endothelial cells of developing alveolar capillaries. Nkx2-1^−/− mice die at birth from respiratory failure caused by lung agenesis (4), demonstrating the importance of the Nkx2-1 gene in lung morphogenesis.

Given a high clinical need for new therapeutic approaches to treat pulmonary disorders, recent studies have focused on the generation of respiratory epithelial cell lineages through the directed differentiation of pluripotent embryonic stem cells (ESCs) or induced pluripotent stem cells (iPSCs). AT2 and club-like epithelial cells were generated from mouse and human ESCs and iPSCs using in vitro differentiation protocols (5–7). Likewise, ESCs and iPSCs have been used to produce intestinal epithelial cells, cardiomyocytes, hepatocytes, neurons, and many other cell types (8, 9). When implanted into laboratory animals, newly differentiated cells undergo morphogenesis and tissue remodeling to form three-dimensional organoids that often contain ESC-derived cell types as well as cells recruited from the host (8, 10). Although organoids resemble the histological structure of developing organs, gene expression and functional studies indicate significant differences between ESC-derived cells and endogenous cells that have undergone normal morphogenesis in the embryo (5–7, 10). To circumvent this problem, blastocyst complementation has been used for in vivo differentiation of ESCs, in which ESCs were injected into blastocysts of animals lacking critical morphogenetic genes to create chimeras. Pdx1^−/− embryos, which lack a pancreas, were complemented with ESCs that formed an entire pancreas in which both exocrine and endocrine cells were derived from ESCs, whereas stromal and vascular cells were mosaic (11, 12). Likewise, ESCs generated a functional pancreas in Pdx1 promoter-Hes1 apancreatic pigs (13). Endothelial cells were complemented with ESCs in Flk1-mutant mice (14), kidney in Sall1-mutant rats (15), lymphocytes in Rag2-deficient mice (16), and neuronal tissues in mice with forebrain-specific expression of diphtheria toxin (17). Recently, mouse ESCs were used to generate lung tissues in mouse embryos with Fgf10 deficiency (18) or embryos with epithelial-specific deletion of either Fgfr2 or β-catenin (19). Simultaneous generation of lung and thyroid tissues from ESCs has not yet been achieved.

Herein, we used blastocyst complementation to produce ESC-derived lungs and thyroids in Nkx2-1^−/− mouse embryos lacking these tissues. Although Nkx2-1^−/− chimeras died at birth because of tracheoesophageal fusion, the chimeras formed thyroids and peripheral lung lobes, the latter of which consisted of airways, alveolar sacs, stroma, and perfused vasculature. Newly formed lungs in Nkx2-1^−/− chimeras had normal histology and ultrastructure. Based on single-cell RNA sequencing (RNAseq) and fluorescence-activated cell sorter (FACS) analysis of chimeric lung tissue, ESCs contributed to the vast majority of respiratory epithelial cell types in Nkx2-1^−/− chimeras, whereas endothelial, immune, and stromal cells were mosaic. ESC-derived and chimeric respiratory cells were indistinguishable from endogenous cells based on gene expression signature, ultrastructure, and cell surface proteins. Using blastocyst complementation, ESC-derived lungs and thyroids were produced in Nkx2-1^−/− chimeric mice, providing an alternative to “bioengineered” tissues growing on artificial matrices or in complex organoid systems.

Our abstract was previously presented at the 2019 Grover Conference.

Methods

Mice and Blastocyst Complementation

The Nkx2-1^−/− mouse line was described previously (4). After the mating of Nkx2-1^+/− mice (CD1 background), mouse blastocysts were obtained at Embryonic Day 3.5 (E3.5), injected with GFP (green fluorescent protein)–labeled mouse ESCs (mESCs) (C57Bl/6 background) (20) and transferred into pseudopregnant CD1 females as described in the online supplement. Embryos were harvested at different time points for analysis. To generate Nkx2-1^−/−; tdTomato embryos, Nkx2-1^+/− males were bred with tdTomato^tg/tg females (Jackson Lab). All procedures were approved by the Institutional Animal Care and Use Committee of Cincinnati Children’s Research Foundation.

Single-Cell RNAseq Analysis

Read alignment, quality controls, false discovery rate, identification of cell clusters, and quantification of cluster-specific gene expression in mouse lung single-cell RNAseq datasets are described in References 21 and 22 and the online supplement. To assess the transcriptomic similarity of cells from control (Nkx2-1^+/−) and Nkx2-1^−/− chimeric mouse lungs, the transcriptomic profiles of control and Nkx2-1^−/− single-cell RNAseq datasets were merged directly without using batch correction methods. The FindVariableGenes function in Seurat (version 2.3.4) was used to identify highly variable genes in the merged data. For epithelial cell types in Nkx2-1^−/− single-cell RNAseq, cells with epithelial-specific gene signatures (1, 21) and z-scores <0 for both GFP and Nkx2-1 mRNAs were classified as endogenous epithelial cells; otherwise, they were classified as ESC-derived cells. For nonepithelial cell types, cells with lineage-specific gene signatures (1, 21) and z-scores <0 for GFP mRNA were classified as endogenous cells; otherwise, they were classified as ESC-derived cells.

Histology, Immunostaining, and Transmission Electron Microscopy

Tissue sections were stained with hematoxylin and eosin (23, 24) or used for immunostaining as previously described (25–28). Primary antibodies are described in Table E1 in the online supplement and References 29–32. Secondary antibodies conjugated with Alexa Fluor 488, 594, or 647 were used for colocalization experiments (33, 34). Fluorescence images were obtained as described (35–37). For transmission electron microscopy, lung tissue was fixed, embedded in epoxy resin, and imaged as described in the online supplement and References 38 and 39.

FACS Analysis

FACS analysis was performed using single-cell suspension obtained from enzyme-digested E17.5 lung tissue as described (40, 41). The identification of pulmonary cell types and FACS antibodies is described in the Table E2 and References 22 and 42–44. Stained cells were analyzed using a five-laser FACSAria II (BD Biosciences) (45, 46).

Statistical Analysis

Statistical significance was determined using a nonparametric Mann-Whitney U test or one-way ANOVA and Student’s t test. The log transformation was applied for skewed data to confirm with normality. P ≤ 0.05 was considered statistically significant. Data were presented as mean ± SD.

Results

Complementation with Mouse ESCs Induces Lung Morphogenesis in Nkx2-1^−/− Mice

Lung structure was examined in Nkx2-1^−/− and wild-type (WT) embryos at E17.5. Consistent with NKX2-1 requirement for lung morphogenesis (4), peripheral lung lobes were entirely absent in Nkx2-1^−/− embryos (Figure E1A). Instead of lung lobes, simplified epithelial cysts were found in the thoracic cavity of Nkx2-1^−/− mice (Figure E1B). Although NKX2-1 was detected in the bronchial and alveolar epithelium of WT lungs, NKX2-1 staining was absent in Nkx2-1^−/− embryos (Figure E1C). Bronchial and cystic epithelium in Nkx2-1^−/− embryos consisted of basal-like cells that expressed p63 and cytokeratin 5 (Figure E1D).

To determine whether ESCs can rescue lung morphogenesis in Nkx2-1^−/− embryos, blastocyst complementation was performed by injecting GFP-labeled mESCs into blastocysts obtained from crosses between Nkx2-1^+/− mice (Figure 1A). Embryos were transferred into surrogate females, and chimeric mice collected. WT and Nkx2-1^+/− chimeras were viable, whereas all Nkx2-1^−/− chimeras died at birth (Table 1). Visual and histological evaluation of E17.5 and newborn Nkx2-1^−/− chimeras showed the presence of lung lobes in the thoracic cavity (Figures 1B and 1C). GFP fluorescence was readily detected in the lung lobes of the chimeras (Figure 1B), consistent with the presence of mESC-derived cells. The mESC contribution to the lung of Nkx2-1^−/− chimeras was higher compared with the lungs of control WT mice and Nkx2-1^+/− chimeras (Figure 1B and Table 1). Nkx2-1^−/− chimeras retained endogenous epithelial cysts typical of Nkx2-1^−/− mice (Figure 1C). The prevalence of endogenous cysts inversely correlated with the abundance of mESC-derived lung tissue in Nkx2-1^−/− chimeras (Figure E2). Similar to control lungs, newly formed lung lobes in Nkx2-1^−/− chimeras contained bronchioles, alveolar saccules, and blood vessels containing erythrocytes within the lumen (Figure 1C). Thus, complementation with mESCs promotes and enabled lung morphogenesis in Nkx2-1^−/− embryos.

Figure 1.

Mouse embryonic stem cells (mESCs) generate lung (Lu) tissue in Nkx2-1^−/− embryos after blastocyst complementation. (A) Schematic representation of blastocyst complementation in Nkx2-1^−/− embryos. Green fluorescent protein–labeled mESCs (C57Bl/6 background) were injected into blastocysts obtained by crossing of Nkx2-1^+/− males with Nkx2-1^+/− females (CD-1 background). Embryos were implanted into surrogate females and harvested at different time points. (B) Bright-field and green fluorescent protein fluorescence images show Embryonic Day 17.5 (E17.5) Lus obtained from Nkx2-1^−/− and wild-type chimeras. (C) Hematoxylin and eosin staining shows the presence of peripheral Lu lobes in the thoracic cavity of newborn Nkx2-1^−/− chimeras. Remnants of endogenous Nkx2-1^−/− Lu cysts are visible in mESC-complemented Nkx2-1^−/− chimeras. High-magnification images show bronchioles, blood vessels, and cysts in Nkx2-1^−/− chimeric Lus. Scale bars, 200 μm (top) and 20 μm (bottom). (D) Transmission electron microscopy of Nkx2-1^−/− and control (wild-type) E17.5 chimeric Lus shows typical morphology of alveolar type 2 cells, containing lamellar bodies, multivesicular bodies, and glycogen. *Secreted surfactant. Right panel shows high-magnification image of areas within the designated box. (E) Ultrastructural features of ciliated and nonciliated epithelial cells (dashed line) are similar in Nkx2-1^−/− chimeras and control animals at E17.5. Scale bars in D and E, 5 μm (left) and 1 μm (right). AT2 = alveolar type 2; Br = bronchioles; Cy = cysts; GFP = green fluorescent protein; GLY = glycogen; He = heart; LA = left atrium; LB = lamellar bodies; LV = left ventricle; MVB = multivesicular bodies; NUC = nucleus; RV = right ventricle; V = blood vessels; WT = wild-type.

Table 1.

Generation of Chimeras Using mESCs and Mouse Blastocysts

Developmental Time	Total Mice (n)	Nkx2-1+/+		Nkx2-1+/−		Nkx2-1−/−
		Mice (n)	High-Contribution Chimeras [n (%)]	Mice (n)	High-Contribution Chimeras [n (%)]	Mice (n)	High-Contribution Chimeras [n (%)]
E13.5	33	7	3 (42.86)	17	6 (35.29)	9	8 (88.89)
E17.5	53	13	6 (46.15)	26	13 (50.00)	14	13 (92.86)
P0/P1	21	4	2 (50.00)	7	3 (42.86)	10*	9 (90.00)

Definition of abbreviations: E13.5 = Embryonic Day 13.5; E17.5 = Embryonic Day 17.5; mESC = mouse embryonic stem cell; P0/P1 = 0–24 hours after birth.

Blastocysts were obtained from crosses between Nkx2-1^+/− mice. High-contribution chimeras exhibited >10% of mESC-derived cells in the lung tissue, whereas low-contribution chimeras exhibited <10% of mESC-derived pulmonary cells.

^*All Nkx2-1^−/− chimeras died at birth.

mESCs Differentiate into Multiple Respiratory Epithelial Cell Types in Nkx2-1^−/− Chimeras

Columnar AT2 cells containing lamellar bodies and multivesicular bodies (consistent with the ultrastructural features of mature AT2 cells) were found in the lungs of Nkx2-1^−/− chimeras by transmission electronic microscopy (Figures 1D and E3A). Although AT2 cells were entirely absent in Nkx2-1^−/− mice, AT2 cells in Nkx2-1^−/− chimeras were abundant and expressed pro-SP-C (Figure 2A). Secreted surfactant and mature SP-C and SP-B proteins were detected in the lungs of Nkx2-1^−/− chimeras (Figures 1D and E4), consistent with alveolar epithelial cell differentiation. Cystic epithelium was undifferentiated in noncomplemented Nkx2-1^−/− embryos and Nkx2-1^−/− chimeras (Figure E3C and E3D). Similar to WT lungs, peripheral lung regions of Nkx2-1^−/− chimeras contained squamous AT1 cells expressing T1α (Figure 2A). T1α was also detected in cystic epithelium (Figure 2A). The vast majority of AT2 and AT1 cells in lungs of Nkx2-1^−/− chimeras expressed GFP (Figure 2D), indicating their mESC origin.

Figure 2.

Mouse embryonic stem cells (mESCs) differentiate into alveolar and airway epithelial cell types in Nkx2-1^−/− chimeras. (A) Immunostaining shows the lack of pro-SPC staining in cystic epithelium of Nkx2-1^−/− newborn lungs. After mESC complementation, pro-SPC–positive alveolar type 2 (AT2) (white) and T1α-positive AT1 cells (red) are present in peripheral lung regions of Nkx2-1^−/− chimeras. mESC-derived cells are detected with GFP (green fluorescent protein) (green). Nuclei were counterstained with DAPI (blue). High-magnification images of AT2 (white arrowheads) and AT1 cells (red arrowheads) are shown in inserts. Basal-like cells expressing T1α are seen in cystic epithelium of Nkx2-1^−/− mice (yellow arrowheads). Inserts show high-magnification images of epithelial cells. Scale bars, 100 μm and 5 μm (insert). (B–C) mESCs differentiate into ciliated and club cells in bronchiolar epithelium of Nkx2-1^−/− chimeras. Club cells were stained for CCSP (B), whereas ciliated cells were stained for β-tubulin (C). Ciliated and club cells are shown with white arrowheads. Scale bars in B, 10 μm and 4 μm (insert); scale bars in C, 20 μm and 8 μm (insert). (D) Percentages of GFP+ and GFP− cells were counted using high-magnification images (n = 4–20 for each group). *P < 0.05 and **P < 0.01. Br = bronchiolar; N.S. = not significant; WT = wild-type.

Bronchiolar epithelium in Nkx2-1^−/− chimeras had ultrastructural features of normal columnar epithelium, consisting of primarily ciliated and club cells (Figures 1E and E3B). Club cells expressed CCSP (Figure 2B), whereas apical surfaces of ciliated cells stained for β-tubulin (Figure 2C). The majority of ciliated and club cells in conducting airways of Nkx2-1^−/− chimeras expressed GFP (Figure 2D), indicating mESC origin. GFP-positive goblet and basal cells were also found in proximal airways and trachea of Nkx2-1^−/− chimeras and stained for MUC5AC or cytokeratin 5, respectively (Figure E5). Nkx2-1^−/− chimeras had extensive pulmonary vascular netwcorks consisting of endothelium and smooth muscle that were derived from either mESCs or endogenous progenitor cells (Figures 2D and E6). Altogether, mESCs differentiated into multiple respiratory cell types in complemented lungs of Nkx2-1^−/− chimeras.

mESC Complementation Induces Thyroid Morphogenesis but Does Not Correct the Tracheoesophageal Fusion in Nkx2-1^−/− Embryos

Consistent with perinatal lethality, the trachea and esophagus in Nkx2-1^−/− newborn chimeras were fused above the tracheal bifurcation, and findings were similar in Nkx2-1^−/− newborn mice without mESC complementation (Figures E7, E8A, and E8C). Although tracheal cartilage and smooth muscle were present in Nkx2-1^−/− chimeras, they were mispatterned, as shown by immunostaining for αSMA and SOX9 (Figures E8B and E9A). In the area of tracheoesophageal fusion, mESC-derived cells were mostly detected in tracheal epithelium based on the location of cartilage and immunostaining for basal cell markers cytokeratin 5 and p63 (Figures E9B and E10). A subset of luminal cells in esophageal epithelium of WT mice and Nkx2-1^−/− chimeras was also derived from donor mESCs (Figure E9C). The mESC-derived and endogenous cells had similar proliferation rates in the tracheal and esophageal epithelium, as shown by immunostaining for Ki-67 (Figure E11).

Immunostaining of E10.5 chimeric embryos for NKX2-1 and SOX2 demonstrated the lack of normal dorsoventral patterning in the foregut of Nkx2-1^−/− chimeras (Figure E12A). NKX2-1⁺ cells expressed GFP and formed cell clusters in different epithelial domains of the foregut (Figures E12A and E12B), findings that were consistent with the lack of tracheal–esophageal separation in newborn Nkx2-1^−/− chimeras (Figure E8). There were no differences in proliferation rates between endogenous and mESC-derived epithelial cells in the foregut of E10.5 chimeras (Figures E12B and E12D). Furthermore, mESC complementation did not change cellular proliferation and apoptosis in respiratory epithelium at E13.5, as demonstrated by immunostaining for Ki-67⁺ (Figure E13) and activated (cleaved) caspase 3 (Figure E14).

Histological assessment of E17.5 Nkx2-1^−/− chimeras showed the presence of thyroid tissue, which was lacking in Nkx2-1^−/− embryos without mESC complementation (Figure 3A). Donor mESCs efficiently contributed to thyrocyte progenitor cells that expressed NKX2-1 and PAX8 (Figures 3A and 3B). mESC-derived cells were also observed in forebrain structures of E13.5 chimeric embryos, including in diencephalon and corpus striatum (Figures E15A and E15B). Compared with WT mice and Nkx2-1^+/− chimeras, the mESC contribution to NKX2-1⁺ neuronal progenitors in Nkx2-1^−/− chimeras was inefficient (Figure E15B). Altogether, mESC complementation promotes thyroid morphogenesis but does not correct the tracheoesophageal fusion or increase NKX2-1⁺ neuronal progenitors in the forebrain of Nkx2-1^−/− chimeric embryos.

Figure 3.

Mouse embryonic stem cell (mESC) complementation enables thyroid morphogenesis in Nkx2-1^−/− embryos. (A) Immunostaining of frozen sections for NKX2-1 shows thyroid tissue in Embryonic Day 17.5 Nkx2-1^−/− chimeras, which was lacking in Nkx2-1^−/− embryos without mESC complementation. NKX2-1⁺ thyrocytes in Nkx2-1^−/− chimeras express green fluorescent protein, indicating mESC origin. Right panels are high-magnification images of the thyroid (yellow box). (B) Immunostaining for PAX8 shows that a majority of PAX8⁺ thyrocytes in Nkx2-1^−/− chimeras express green fluorescent protein. Scale bars, 200 μm (left panels) and 50 μm (right panels). Es = esophagus; GFP = green fluorescent protein; Th = thyroid; Tr = trachea; WT = wild-type.

mESCs Differentiate into Respiratory Epithelial Progenitor Cells and Mesenchyme during Lung Morphogenesis in Nkx2-1^−/− Chimeras

During lung development, mature epithelial cell lineages differentiate from epithelial progenitors in the embryonic lung tubules that are surrounded by mesenchyme, which differentiates into stromal, smooth muscle, and endothelial cells (1, 2). We tested whether mESCs contribute to epithelial and mesenchymal progenitors in E13.5 embryos. To visualize endogenous cells, mESC complementation was performed in Nkx2-1^−/− blastocysts that expressed a tdTomato reporter transgene (Nkx2-1^−/−; tdTomato^tg/−) (Figure 4A). Lungs were present in mESC-complemented Nkx2-1^−/−; tdTomato^tg/− embryos at E13.5 and exhibited tubular structures typical of the pseudoglandular stage of lung development (Figures 4B and 4C). Visually compared with controls, GFP expression was higher in lung tissue of Nkx2-1^−/−; tdTomato^tg/− chimeras (Figures 4B and 4C), whereas GFP concentrations in the heart and liver were similar (Figure 4D). GFP was detected in a majority of distal epithelial cells in Nkx2-1^−/−; tdTomato^tg/− chimeras, as shown by the colocalization of GFP with SOX9 (Figures 4E and 4F). The numbers of GFP-expressing mesenchymal cells were similar in control and Nkx2-1^−/−; tdTomato^tg/− chimeric lungs (Figure 4E).

Figure 4.

Peripheral lung epithelial tubules in Embryonic Day 13.5 (E13.5) Nkx2-1^−/− chimeras are formed by mouse embryonic stem cell (mESC)–derived cells. (A) Schematic representation of mESC complementation in Nkx2-1^−/−; tdTomato reporter embryos. (B–D) Analysis of GFP (green florescent protein) and tdTomato fluorescence shows high contribution of mESCs to respiratory epithelial tubules of Nkx2-1^−/−; tdTomato E13.5 embryos. mESC progenies are detected by GFP (green). Endogenous cells are detected by tdTomato (red). Yellow in C and D shows a fusion between mESC-derived and endogenous cells in the myocardium and smooth muscle layers adjacent to epithelium. mESC contribution to the heart and liver is similar in Nkx2-1^−/−; tdTomato and control embryos (Nkx2-1^+/−; tdTomato). Scale bar in B, 200 μm; scale bars in C and D, 100 μm. (E) Percentages of GFP+ cells show that a majority of distal lung epithelial cells in Nkx2-1^−/−; tdTomato E13.5 embryos is derived from mESCs. mESC contribution to lung mesenchyme is unaltered in Nkx2-1^−/−; tdTomato chimeras compared with control animals (n = 14–16 for each group). **P < 0.01. (F) GFP colocalizes with SOX9 in distal lung epithelial tubules of E13.5 Nkx2-1^−/− chimeras. Scale bar, 50 μm. Epi = epithelium; He = heart; Li = liver; Lu = lung; Me = mesenchyme; N.S. = not significant.

mESC-derived and Endogenous Cells Have Similar Cell Surface Marker Proteins

To assess the contribution of mESCs to respiratory cell types, FACS analysis was performed using enzymatically digested E17.5 chimeric lungs, identifying epithelial cells (CD326⁺CD31⁻CD45⁻), endothelial cells (CD31⁺CD45⁻), hematopoietic (immune) cells (CD45⁺CD31⁻), fibroblasts (CD140a⁺CD31⁻CD45⁻), and pericytes (NG2⁺CD326⁻CD140a⁻CD31⁻CD45⁻) (Figures 5A and E16). Consistent with the immunostaining of lung sections (Figures 2 and 4), GFP was detected in the vast majority of epithelial cells in Nkx2-1^−/− chimeras, whereas other respiratory cell types were mosaic (Figures 5B and 5C). The expression of the cell marker proteins, including CD31, CD45, CD326, CD140a, and NG2, was restricted to corresponding cell lineages in Nkx2-1^−/− chimeras (Figure 5D). Interestingly, the mean GFP fluorescence intensity for each marker was similar in mESC-derived and endogenous cells regardless of the chimeras’ genotypes (Figure 5D). Thus, mECS-derived and endogenous cells are indistinguishable based on cell surface expression of marker proteins.

Figure 5.

Mouse embryonic stem cells (mESCs) differentiate into epithelial (Epi), endothelial (Endo), and mesenchymal cell types in Nkx2-1^−/− chimeras. (A) Fluorescence-activated cell sorter gating strategy for Endo (a), hematopoietic (b), and Epi (c) cells; fibroblasts (d); and pericytes (e). (B) The percentage of GFP (green fluorescent protein)-positive cells was determined in each cell type based on selective cell surface markers. Single-cell suspensions were obtained from lungs of Embryonic Day 17.5 chimeras (Nkx2-1^+/+ [n = 7], Nkx2-1^+/− [n = 6], and Nkx2-1^−/− [n = 4]). mESC contribution is significantly higher in pulmonary Epi cells (CD326⁺CD31⁻CD45⁻) of Nkx2-1^−/− chimeras. **P < 0.01. mESC contributions to Endo cells (CD31⁺CD45^–), hematopoietic (immune) cells (CD45⁺), fibroblasts (CD140a⁺CD31⁻CD45⁻), and pericytes (NG2⁺CD326⁻CD140a⁻CD31⁻CD45⁻) are similar to control animals. (C) Histograms show GFP concentrations in respiratory cells of chimeras (green) compared with GFP concentrations in age-matched WT lungs without mESC complementation (blue). (D) Fluorescence-activated cell sorter analysis shows that mESC-derived (GFP⁺) and endogenous cells (GFP⁻) in Nkx2-1^−/− and Nkx2-1^+/− (control) chimeras have similar cell surface expression of lineage-selective marker proteins. Mean fluorescence intensity was used to compare cell surface expression of CD31, CD45, CD326, CD140a, and NG2 in different respiratory cell subtypes (n = 4–6 embryos in each group). E17.5 = Embryonic Day 17.5; FACS = fluorescence-activated cell sorter; Fibr = fibroblasts; MFI = mean fluorescence activity; N.S. = not significant; Peri = pericytes; WT = wild-type.

Single-Cell RNAseq Identifies Close Similarities in Gene Expression Signatures between Nkx2-1^−/− Chimeras and Normal Lungs

To compare gene signatures in Nkx2-1^−/− chimeras and control lungs, single-cell RNAseq was performed using the 10× platform. Based on known gene expression signatures (1, 21), 5,555 cells from nine major cell subtypes were identified in control lungs at E17.5 (Figure 6A). These include airway epithelial cells, AT1 cells, AT2 cells, AT1/AT2 cells, matrix fibroblasts, αSMA⁺ fibroblasts, pericytes, endothelial cells, and myeloid cells (Figure 6A). Using the same gene signatures, 5,905 cells from Nkx2-1^−/− chimeras subclustered in the identical cell subtypes (Figure 6B). A combined analysis of Nkx2-1^−/− chimeric and control cells demonstrated similar distributions of epithelial, endothelial, immune, and mesenchymal lineages (Figures 6C and 6D) and among major cell subtypes (Figure 6E), indicating identical cell types in the Nkx2-1^−/− chimera and control lungs. Using GFP mRNA to identify mESC-derived cells, we found that the contribution of mESC-derived cells was higher in the Nkx2-1^−/− chimera compared with the control mouse (Figures E17A and E17B). Epithelial cell subtypes in Nkx2-1^−/− chimeras, including AT2, AT1, AT1/AT2, and airway epithelial cells, exhibited the highest contribution from mESCs (Figure E17B and Tables E3 and E4). Only 23 of 2,335 epithelial cells (0.01%) were endogenous in Nkx2-1^−/− chimeras based on epithelial gene signature and z-scores for GFP and Nkx2-1 mRNAs (Figure E18). The contribution of mESCs to mesenchymal, endothelial, and immune cell subsets in Nkx2-1^−/− chimeras ranged between 20% and 50% (Table E3), findings consistent with FACS analysis (Figure 5B) and immunostaining (Figure 2D). Thus, nearly all respiratory epithelial cells in Nkx2-1^−/− chimeras were mESC derived, whereas other cell types were mosaic.

Figure 6.

Single-cell RNA sequencing demonstrates close identity of all respiratory cell types in Nkx2-1^−/− and control chimeras. (A) Uniform manifold approximation and projection (UMAP) plot of single cells (n = 5,555) from control Embryonic Day 17.5 mouse lung (Nkx2-1^+/−) complemented with mouse embryonic stem cells (mESCs). Cells were colored by cell subtype information, including airway epithelial cells, alveolar type 1 (AT1), AT2, AT1/AT2 epithelial cells, matrix fibroblasts, myeloid cells, αSMA+ fibroblasts, pericytes, and vascular endothelial cells. (B) UMAP plot of single cells (n = 5,905) from Embryonic Day 17.5 Nkx2-1^−/− lung complemented with mESCs. Cells were colored by cell subtype information. (C) Cells from control and Nkx2-1^−/− chimeras were integrated, indicating transcriptomic similarity of the two mouse lungs. Expression profiles of the two datasets were integrated, projected onto a lower dimensional space using principal component analysis, and visualized using UMAP. No batch correction methods were used for integrating the two datasets. (D) UMAP plot of integrated data with cells colored by major cell types identified by independent analysis of each data. (E) UMAP plot of integrated data with cells colored by cell subtypes identified by independent analysis of each data. (F) Gene expression signatures in Nkx2-1^−/− and Nkx2-1^+/− chimeric lungs are similar. Cell type signature genes were obtained from the Lung Gene Expression Analysis web portal (https://research.cchmc.org/pbge/lunggens/mainportal.html). For each cell type, 50 signature genes with the lowest P values were used for cell identification. The heatmap shows gene expression signatures in endogenous (GFP⁻) cells from control Nkx2-1^+/− chimeric lung (orange), endogenous (GFP⁻) cells from Nkx2-1^−/− chimeric lung (blue), and mESC-derived (GFP⁺) cells from Nkx2-1^−/− chimeric lung (yellow). Differential expression was tested using a negative binomial-based test. Genes with a false discovery rate of P value < 0.1, log₂ fold change >1, expression frequency >20% in a cell subtype of control or Nkx2-1^−/− chimeras were considered unique for the cell subtype. Selective cell subtype markers are shown in the right column. AirwayEpi = airway epithelial cells; GFP = green fluorescent protein; MatrixFB = matrix fibroblasts; MyoFB = αSMA+ fibroblasts; VasEndo = vascular endothelial cells.

Gene expression signatures revealed significant similarities among all respiratory epithelial cell subtypes (Figure 6F). AT2 cells isolated from Nkx2-1^−/− and control (Nkx2-1^+/−) chimeras expressed AT2-specific genes, including Sftpc, Abca3, Slc34a2, and Lpcat1 (Figure 6F). AT1 cells expressed Ager, Pdpn, and Emp2 mRNAs, whereas Foxj1, Agr3, Scgb1a1, and Scgb3a2 were restricted to airway epithelial cells (Figure 6F). Gene expression signatures of AT1/AT2 progenitors were conserved in both chimeras. Gene expression signatures in matrix fibroblasts, αSMA⁺ fibroblasts, pericytes, endothelial cells, and myeloid cells were similar in all experimental groups (Figure 6F). For selected genes, we used violin plots to confirm similarities in cell specificity and expression levels of Nkx2-1, Cdh1, Sftpb, Pdpn, Sox2, Pecam1, Cdh5, Tcf21, Pdgfra, Pdgfrb, Ptprc (Cd45), and Cd68 in control and Nkx2-1^−/− chimeras (Figure E19). Altogether, the cellular compositions of the lung tissue and gene expression signatures of all cell types were similar in control and Nkx2-1^−/− chimeras.

Discussion

Inactivation of the Nxk2-1 gene in mice causes severe defects in the formation of the lung, the thyroid, and forebrain-derived structures (4). NKX2-1 is required for the development of respiratory epithelial cell lineages (1, 2). In the present study, the vast majority of epithelial cells in Nkx2-1^−/− chimeric lungs were derived from mESCs after blastocyst complementation, indicating the competitive advantage of NKX2-1–expressing progenitors, which differentiate into diverse respiratory epithelial lineages in the Nkx2-1^−/− embryonic environment. Consistent with our studies, ESCs contributed to most of pancreatic epithelial cells in PDX1-deficient mice, rats, and pigs (11–13). The present study indicates that endothelial, stromal, and hematopoietic cells in lungs of both Nkx2-1^−/− and control chimeras were derived from either mESCs or endogenous progenitor cells. Thus, mESC-derived progenitors differentiating into nonepithelial cell lineages do not have an advantage over endogenous Nkx2-1^−/− cells to generate pulmonary vasculature and stroma in Nkx2-1^−/− chimeras. Despite mosaicism in the pulmonary vasculature and stroma, mESC-derived cells from multiple respiratory lineages were highly differentiated and indistinguishable from the endogenous cells based on gene expression signatures and cell surface proteins. Nkx2-1 expression in epithelial cells rescues lung morphogenesis and differentiation of mesenchymal cell types and enables Nkx2-1^−/− chimeras to generate entire lung lobes with a high contribution from mESCs. Our results are consistent with recent studies demonstrating that mouse ESCs can efficiently contribute to lung tissues in mouse embryos deficient for Fgf10 (18), Fgfr2, or β-catenin (19), all of which are characterized by severe disruptions in lung morphogenesis.

Despite the findings that peripheral lung and thyroid tissues were present in Nkx2-1^−/− chimeras, all died at birth. The lack of tracheoesophageal separation, the mispatterning of the proximal trachea, and the lack of NKX2-1–expressing neuronal progenitors in the forebrain-derived tissues likely cause perinatal lethality of the Nkx2-1^−/− chimeras. The mESCs efficiently contributed to tracheal but not esophageal epithelium, and the trachea remained fused with the esophagus in Nkx2-1^−/− chimeras. Thus, mESC complementation is insufficient to correct the tracheoesophageal fusion in Nkx2-1^−/− embryos. Although proliferation rates between endogenous and mESC-derived epithelial cells were unchanged, Nkx2-1^−/− chimeras showed abnormal dorsoventral patterning in the foregut at E10.5. The tracheoesophageal fusion in newborn Nkx2-1^−/− chimeras is likely a result of abnormal dorsoventral patterning at the early stages of tracheoesophageal morphogenesis. It is also possible that diminished or delayed mesenchymal–epithelial signaling mediated by FGF10 and SHH (both required for mesenchyme differentiation and formation of cartilage rings) (47, 48) contributes to defects in tracheoesophageal morphogenesis in Nkx2-1^−/− chimeras. Interestingly, Nkx2-1^−/− chimeras had thoracic cysts lined by non–NKX2-1–expressing basal-like cells that are typical of Nkx2-1^−/− embryos. The Nkx2-1^−/− cystic epithelial cells did not differentiate into other respiratory epithelial lineages in the chimeras, consistent with the intrinsic requirement for NKX2-1 in epithelial cell differentiation. CRISPR/Cas9-mediated deletion of genes critical for tracheoesophageal morphogenesis and basal cell differentiation may be needed in Nkx2-1^−/− chimeras to generate an entire respiratory system from donor mESCs in the future.

Blastocyst complementation between species to generate intraspecies chimeras (49), creates the opportunity to use human iPSCs to generate patient-specific organs using large animals as “biological reactors.” However, ESCs and iPSCs are capable of integrating into many embryonic tissues, including the brain, testes, and sensory organs (49, 50), raising important ethical concerns for the use of human–animal chimeras in regenerative medicine. Use of fetal tissues instead of adult chimeras and the introduction of additional genetic modifications to improve the selectivity of ESC integration, including those needed to obviate immunologic rejection, will be needed to advance the technology. Intraspecies Nkx2-1^−/− chimeras made in large animals (for example, pigs or sheep) may be suitable for the generation of chimeric lungs from patient-specific iPSCs, making chimeric lungs a potential source for lung transplantation or for the production of patient-specific respiratory progenitors for lung regenerative medicine.

Conclusions

In summary, blastocyst complementation in Nkx2-1^−/− mouse embryos was used to generate peripheral lung tissues with near complete contributions from mESCs to the respiratory epithelium. The mESC-derived pulmonary cells were indistinguishable from endogenous lung cells based on gene expression signatures, cell surface proteins, and ultrastructural features. Nkx2-1^−/− chimeras can be considered for in vivo differentiation of patient-derived iPSCs into respiratory cell lineages for future cell therapies.