Conditional blastocyst complementation of a defective Foxa2 lineage efficiently promotes the generation of the whole lung

Millions suffer from incurable lung diseases, and the donor lung shortage hampers organ transplants. Generating the whole organ in conjunction with the thymus is a significant milestone for organ transplantation because the thymus is the central organ to educate immune cells. Using lineage-tracing mice and human pluripotent stem cell (PSC)-derived lung-directed differentiation, we revealed that gastrulating Foxa2 lineage contributed to both lung mesenchyme and epithelium formation. Interestingly, Foxa2 lineage-derived cells in the lung mesenchyme progressively increased and occupied more than half of the mesenchyme niche, including endothelial cells, during lung development. Foxa2 promoter-driven, conditional Fgfr2 gene depletion caused the lung and thymus agenesis phenotype in mice. Wild-type donor mouse PSCs injected into their blastocysts rescued this phenotype by complementing the Fgfr2-defective niche in the lung epithelium and mesenchyme and thymic epithelium. Donor cell is shown to replace the entire lung epithelial and robust mesenchymal niche during lung development, efficiently complementing the nearly entire lung niche. Importantly, those mice survived until adulthood with normal lung function. These results suggest that our Foxa2 lineage-based model is unique for the progressive mobilization of donor cells into both epithelial and mesenchymal lung niches and thymus generation, which can provide critical insights into studying lung transplantation post-transplantation shortly.


Introduction
Whole organ generation to treat various intractable diseases has long been challenging (Hackett et al., 2010;Kemter et al., 2020;Ott et al., 2010;Petersen et al., 2010;Wang, 2019).Organ bioengineering strategy based on recellularizing tissue-specific progenitors into the decellularized scaffolds, induced pluripotent stem cell (iPSC)-derived organoids, or 3D bioprinters are the nextgeneration tissue transplant therapies (Guyette et al., 2014;Kotton and Morrisey, 2014;Petersen et al., 2010;Tian et al., 2021).Even with these techniques, however, the mammalian lung is one of the most challenging organs to replicate because of its anatomical complexity and cellular diversity.It contains hundreds of airway branches and a thin micron-sized alveolar layer of inflated and wellvascularized alveoli composed of billions of cells from more than 50 different cell types (Crapo et al., 1982;Kotton and Morrisey, 2014;Stone et al., 1992;Travaglini et al., 2020).Donor organs for lung transplantation are in short supply worldwide, but the technology does not exist to generate whole lungs composed of tissue-specific epithelial and mesenchymal cells, including endothelial cells.Lungs grow fully only through natural lung development.
Mesendoderm is a bipotent transitional state between the PS and nascent mesoderm labeled by Mixl1, Pdgfrα, and Brachyury (T) during gastrulation that can give rise to both DE and mesoderm (Hart et al., 2002;Tada et al., 2005).Although it was speculated that mesendoderm might form LPM and DE, there have been no conclusive genetic studies on whether mesendoderm gives rise to both lung epithelium and mesenchyme.Pdgfrα is expressed in the epiblast-derived mesendoderm, the primitive endoderm (PrE), and its extra-embryonic endoderm derivatives, such as parietal and visceral endoderm, around E5.5-E7.5.Foxa2 plays a pivotal role in alveolarization and airway goblet cell expansion (Wan et al., 2004), while there was a significant knowledge gap regarding Foxa2 lineage during lung development.
Blastocyst complementation (BC) has been proposed as a promising option for tissue-specific niche complementation (Chen et al., 1993).This unique technology has been further developed into intraand interspecies organ generations such as kidney, pancreas, thymus, and blood vessels (Hamanaka et al., 2018;Kobayashi et al., 2010;Usui et al., 2012;Yamaguchi et al., 2017).However, co-generation of targeted organs and thymus has never been reported, while educating host immune cells for organ transplantation is critical to avoid graft-versus-host disease (GvHD) post-transplantation (Bos et al., 2022).The production of entire organs, including tissue-specific epithelium and mesenchyme, including endothelium, was also tricky because the origin of pulmonary endothelium seems similar to the other organs.Unfortunately, even with BC, the lungs produced were non-functional and very inefficient, and in addition, the chimeric lungs contained a substantial amount of host-derived tissue (Kitahara et al., 2020;Wen et al., 2021).Previously, we established the conditional blastocyst complementation (CBC) approach, which targets specific lineages complemented by donor PSCs (Mori et al., 2019).Using lineage-specific drivers of lung endoderm in CBC avoids the effects of genetic manipulation in non-target organs for the generation of empty organ niches that lead to functional chimeric lung generation (Mori et al., 2019).However, most of the lung mesenchyme and endothelium were still derived mainly from the host cells, which was the severe limitation of CBC (Mori et al., 2019).Given that the CBC approach targeted endodermal lungs, we speculated that this limitation was due to a significant gap in our knowledge of the origin of all lung cell types, especially pulmonary mesenchyme, including endothelium.In particular, the complementation of endothelium is a critical issue for overcoming hyperacute rejection after lung transplantation.To overcome this critical issue, we explored the origin and the program of whole lung epithelium and mesenchyme, the major components of the lung.
We hypothesized that targeting a single lung precursor lineage may facilitate the designing of the entire lung generation.Our Foxa2-based CBC methodology diverges from our prior Shh-based CBC model (Mori et al., 2019), which targeted the earliest lung precursor lineage, mesendoderm, allowing us to label entire lung epithelium and about 20% of mesenchyme, enough for leading to generating the whole lungs efficiently.Accidentally, we also found that Fgfr2 deficiency in the Foxa2 lineage caused a thymus agenesis phenotype, which was simultaneously rescued by injecting pluripotent stem cells (PSCs) into the blastocysts.Our Foxa2 lineage-based CBC approach results in the highly efficient co-generation of functional lungs and thymus, offering new avenues for exploring the future of autologous lung transplantation.

Pdgfrα + lineage during gastrulation gives rise to the entire lung mesenchyme
To determine the origin of LPM and pulmonary endothelium that leads to the whole lung generation, we performed lung mesenchyme precursor lineage-tracing analysis using Pdgfra CreERT2/+ ; Rosa tdTomato/+ mice.At E5.5 tamoxifen (Tm) administration, E14.5 lung immunostaining analysis showed that the entire lung mesenchyme was labeled with tdTomato but not in the epithelium (Figure 1A).E12.5 morphometric analyses showed an extremely high proportion of lung mesenchyme labeling with tdTomato (Figure 1B, C).Among three embryos at E12.5, two showed 100% tdTomato labeling for lung mesenchyme markers, while one showed about 60% labeling in the mesenchyme (Figure 1C).Sma + airway smooth muscle cells (86.9% ± 22.7), Pdgfrα + lung mesenchyme (78.7% ± 36.9), and VE-cadherin + endothelium (82.5% ± 30.4) were labeled with tdTomato (Figure 1C).These results indicate that the whole lung mesenchyme originates from the Pdgfrα lineage around early-to-mid-streak-stage embryos before LPM formation.In addition, the Pdgfrα lineage labels 20.5% ± 18.3 the lung epithelium (Figure 1B, arrows), suggesting that the contribution of the Pdgfrα lineage to the lung endoderm is limited.It indicates that the Pdgfrα lineage faces challenges in rendering the entire lung niches effectively defective across both the endoderm and mesoderm for CBC-mediated lung generation.

Foxa2 lineage labels Pdgfrα + mesendoderm niche during mouse development
Single-cell RNA-seq (scRNA-seq) analysis using Foxa2-Venus fusion protein reporter mice indicated that the Foxa2 lineage might give rise to LPM and DE (Scheibner et al., 2021).
The online version of this article includes the following source data for figure 1: Source data 1.Quantification of Pdgfra CreERT2/+ ; Rosa26 tdTomato/+ lineage labeling by immunofluorescence (IF) morphometric analysis.
Fgfr2 knockout in lung mesenchyme suppressed their proliferation and showed higher chimerism of the entire lung by CBC Previous reports described that Fgfr2 is critical for lung epithelial cell development but also important for lung mesenchyme proliferation (Yin et al., 2008).Although tdTomato + Fgfr2 knockout mesenchymal cells remained in early lung development at E14.5 (Figure 4I), the percentage of Ki67 + proliferating cells was significantly higher in GFP + donor cells compared to tdTomato + host cells (Figure 4J,  K).These results suggest that generating Fgfr2-deficient niche in the host Foxa2-lineage-derived lung mesenchyme is efficient for donor iPSCs recruitment in the defective Foxa2 lineage mesenchyme niches during development.
These results indicate that Foxa2 Cre/+ ; Fgfr2 cKO provides defective niches in both lung epithelium and mesenchyme and thymic epithelium to be complemented by donor PSCs, enabling the entire lungs and thymus co-generation.
The online version of this article includes the following source data and figure supplement(s) for figure 4: Source data 1.Morphometric analysis: % GFP in Nkx2-1 + epithelial cells and Sma + airway smooth muscle cells in Foxa2 Cre ; Fgfr2 cKO; Rosa tdTomato and Shh Cre ; Fgfr2 cKO; Rosa tdTomato .Source data 2. Morphometric analysis: % Ki67 labeling in mesenchymal cells of E14.5 Foxa2 Cre/+ ; Fgfr2 cKO chimeric lungs.Lung epithelial cell precursors were well known to be Shh + DE in the lung development field (Cardoso and Kotton, 2008;Christodoulou et al., 2011;Harris et al., 2006;Kadzik and Morrisey, 2012;Tian et al., 2011;Weaver et al., 1999;Xing et al., 2008), and indeed, the Shh-lineage traces putative DE-derived epithelial lineage but little lung mesenchyme (Figure 4-figure supplement 3).Targeting the endodermal lung lineage driven by Shh was sufficient for lung epithelial complementation but insufficient to generate whole lungs, and host-derived cells remained substantially in the mesodermal lung component (Mori et al., 2019).In contrast to lung epithelial precursors, the orderly commitment and the origin of the entire pulmonary mesenchyme needed to be better defined.We showed that gastrulating Pdgfrα lineage is the origin of the whole lung mesoderm, including endothelium (Figure 1A, B).Furthermore, Foxa2 or Pdgfrα lineage labels a population primarily comprised of the earliest specified precursors in the distal compartment of mesendoderm, similar to human iPSC cell-derived directed differentiation protocol.Our findings, summarized in Figure 5-figure supplement 1A, pinpoint the lineage hierarchy of specified lung precursors in gastrulating mesendoderm, further supported by the scRNA-seq analysis in early embryonic development (Figure 5-figure supplement 1; Pijuan-Sala et al., 2019).
As previously indicated that nascent mesoderm differentiation into a CPP fate (Bardot et al., 2017;Devine et al., 2014;Ng et al., 2022;Peng et al., 2013), we clarified the orderly mesendoderm progression of gene expression, Mixl1, Pdgfrα, and Foxa2, and lung progenitor-related markers that parallels the commitment of Foxa2 + Pdgfrα + mesendoderm to an LPM and DE fate in the lung-directed differentiation protocol using human iPSC (Figure 3), not clarified in the previous pioneering works (Chen et al., 2017;Hawkins et al., 2017;Huang et al., 2014).We uncovered that PDGFRα + mesendoderm primes the cell fate of Foxa2 + DE after the MIXL1 + mesendoderm specification in the directed significance at ns: non-significant.(C) Gross morphology of 4 weeks rescued lung and thymus from Foxa2 Cre/+ ; Fgfr2 cKO; Rosa tdTomato (Fgfr2 cKO) or Foxa2 Cre/+ ; Fgfr2 hetero; Rosa tdTomato (Fgfr2 hetero) injected with PSC CAG-GFP .Scale bar = 5 mm.differentiation.How the progenitors acquire the lung epithelial cell fate of NKX2-1 with the distal lung tip marker, SOX9, and TBX4 + mesenchyme marker is also fundamental for understanding human lung development.
Our lineage tracing analysis also highlighted the unanticipated Foxa2 lineage program, the progressive increase of lineage labeling by spontaneous expression of Foxa2 mRNA that occupies more than half of the lung mesenchyme during lung development.Foxa2 lineage + lung mesenchyme is a part of Pdgfrα lineage + cells, potentially providing a unique competitive developmental niche during lung development.Further analyses using Foxa2CreERT2-lineage-tracing mice are required to clarify it.Intriguingly, the Foxa2 lineage + mesenchyme did not show any discrete fates that would predict anatomical localization-depleting Fgfr2 in the Foxa2 lineage results in the loss of the Fgfr2 mitogen-mediated function of Foxa2 lineage + lung mesenchyme, which leads to the loss of proliferative ability in most host lung mesenchyme.The Foxa2-lineage labeling in lung mesenchyme is around 20% at E14.5 (Figure 2D, E).Moreover, the donor cell complementation in lung mesenchyme at E14.5 still holds host-derived lung mesenchyme but decreases at E17.5 and adulthood (Figures 4I, C,  5E, F).These results indicate that the defective Foxa2 lineage is critically essential for efficient lung complementation.
We also found that the combination of Fgfr2 depletion in Foxa2 lineage preserved other major organs, such as kidney, glandular stomach, hair follicles, tooth bud, and limbs, distinct from the agenesis or dysgenesis phenotype caused by the systemic Fgfr2 depletion (Arman et al., 1999;De Langhe et al., 2006;De Moerlooze et al., 2000;Sekine et al., 1999).This is because the timing and the requirement of Fgfr2 are different from the lineage in each organ (Bates, 2011;Revest et al., 2001).We accidentally found a thymus agenesis phenotype in the Foxa2 Cre/+ ; Fgfr2 cKO mice but not in Shh Cre/+ ; Fgfr2 cKO mice and its complementation.The thymus is raised in the third pharyngeal pouch with the requirement of the Forkhead Box N1 (Foxn1) transcription factor in epithelial cells (Chhatta et al., 2021), while the redundant role of Foxa2 and Foxa1 in T cell maturation was reported (Ramachandran et al., 2022).The previous report generated thymus using BC by the Foxn1 depletion in host animals, followed by the PSC injections (Yamazaki et al., 2022).Our reports showed novel evidence of the requirement of Fgfr2 signaling for thymus formation, specifically in the Foxa2 lineage, associated with the previous statements regarding the critical role of Fgfr2 in thymus formation and thymic epithelial cell differentiation (Dooley et al., 2007;Revest et al., 2001).This study showed evidence that thymus formation is essentially formed in the Foxa2 lineage but not Shh, supported by the previous report that the dynamic Shh expression regulates larynx-esophageal separation and esophageal constriction on the fourth pharyngeal pouches above the lungs during development (Ramachandran et al., 2022).Corresponding to the previous report, the mice with the rescued thymus showed no apparent signs of immune deficiency and survived until adulthood (Yamazaki et al., 2022).In these complemented mice, immune cells were most likely educated in self-recognition by donor cells' derived thymus epithelium and induced central tolerance.How thymic epithelium is distinctively lineage labeled by Foxa2 and Shh, and the tolerance difference should be investigated further.It is also important to note that GvHD is a crucial risk factor for long-term survival following human lung transplantation (Bos et al., 2022).If we transplant the complemented lungs as a graft containing the conditioned host-and donor-derived immune cells to the recipient that provided donor PSC, it may not cause GvHD.This unique research direction of autologous lung transplantation must be extensively investigated using recipient-derived iPSC with our CBC lung generation models.
An unambiguous proof of the developmental origins of patterned organs is critical for developing regenerative strategies and a better understanding of the genes responsible for congenital malformations.The human-derived developmental program modification to match the host animals is vital for the future development of whole lung generation via CBC using human iPSCs.More broadly, our studies offer a new paradigm that can be applied to modeling various congenital lung diseases of both lung mesenchyme and epithelium and future autologous transplantation therapies using iPSC in the near future.

Culture of human PSCs
All PSC lines were maintained in feeder-free conditions on laminin iMatrix-511 silk E8 (Amsbio, AMS.892021) in StemFit 04 complete Medium (Amsbio, SFB-504), supplemented with Primocin (Invivogen, ant-pm-1), and passaged with TrypLE Select (Gibco, A1285901).All human iPSC lines used were characterized for pluripotency and were found to be karyotypically normal.The BU3NGST cell line was kindly gifted by Dr. Finn Hawkins and Dr. Darrell Kotton at Boston University, Boston, MA.Dr. Jennifer Davis, the University of Washington School of Medicine, Seattle, WA, kindly gifted the Rainbow cell line.PD2 and TD1 hiPSC were generated from deidentified commercially available human peripheral blood mononuclear cell and tracheal epithelial cell lines via the manufacturing protocol of Sendai virus-mediated reprogramming (CytoTune2.0)(Thermo Fisher, A16517) under the CUIMC ESCRO guidelines.Every other month, all iPSC lines screened negative for mycoplasma contamination using a MycoAlert PLUS detection kit (Lonza, LT07-710).

Immunofluorescence
Before the immunostaining, antigen retrieval was performed using Unmasking Solution (Vector Laboratories, H-3300) for 10 min at around 100°C by microwave.4-7 µm tissue sections were incubated with primary antibodies in the staining buffer containing 0.025% Triton X-100, and 1% bovine serum albumin (BSA) at 4°C.Mouse primary antibody staining was done using M.O.M kit (Vector Labs, BMK-2202).Then washed in PBS and incubated with secondary antibodies conjugated with Alexa488, 567, or 647 (Thermo Scientific, 1:400) with NucBlue Fixed Cell Ready Probes Reagent (Hoechst) (Thermo Scientific, R37605) for 1.5 hr, and mounted with ProLong Gold antifade reagent (Invitrogen, P36930).The images were captured by a Zeiss confocal 710 microscope or Leica Stellaris 8 confocal microscopy or DMi8 Leica widefield microscope.The antibodies are listed in Supplementary file 1.

Immunocytochemistry
Cells on culture dishes were fixed with 4% paraformaldehyde (PFA) for 30 min at room temperature (RT), permeabilized, and blocked with staining buffer for 1 hr at RT.Primary antibodies were incubated overnight at 4°C in the staining buffer.After three washes in PBS, secondary antibodies and NucBlue Fixed Cell Ready Probes Reagent (Hoechst) were incubated for 1 hr.The samples were imaged using DMi8 Leica widefield microscope.The antibodies are listed in Supplementary file 1.

RNAScope in situ hybridization
RNA at E18.5 lung sections were stained by RNAScope probes: Mm-Foxa2-T8 (Advanced Cell Diagnostics, #409111-T8) or Negative control (NC) (Advanced Cell Diagnostics, #324341) using the RNAScope HiPlex12 Reagent kit v2 (Advanced Cell Diagnostics, #324419) according to manufactureprovided protocol.Subsequently, sections were incubated with tdTomato antibody for 2 hr at RT, washed in PBS, incubated with secondary antibody conjugated with Alexa488 with NucBlue Fixed Cell Ready Probes Reagent (Hoechst) for 1.5 hr, and mounted with ProLong Gold antifade reagent.The images of Foxa2 and NC in situ hybridization were captured with the same setting by a Zeiss confocal 710 microscope.

Real-time quantitative RT-PCR
Total RNA was extracted using a Direct-zol RNA MiniPrep Plus kit (Zymo Research, R2072), and cDNA was synthesized using Primescript RT Master Mix (Takara, RR036B).The cDNAs were then used as templates for quantitative RT-PCR analysis with gene-specific primers.Reactions (10 µl) were performed Luna Universal qPCR Master Mix (New England Biolabs, M3003X).mRNA abundance for each gene was determined relative to GAPDH mRNA using the 2 −ΔΔCt method.The primers are listed in Supplementary file 2. Data were represented as mean ± standard deviation of measurements.
The number of animals cells per group is provided in the legends.The undetected values in each biological experiment in Figure 3D were removed from the graphs.

EdU assay
100 µg of EdU (Click iT EDU cell proliferation kit, C10337) was injected to pregnant females by intraperitoneal injection 4 hr prior to analysis.Lungs from lineage-tracing mice at E14.5 and E18.5 were harvested and prepared for the FCM.Then cells were fixed by 4% PFA in PBS and stained by the anufacture's protocol.
nGFP + iPSC establishment E14.5 lung tissues of Rosa26 nTnG/nTnG mice were harvested in a dissociation buffer described above.The dissociated cells were seeded on a 10-cm dish, and only lung fibroblast survived after 1 week in Mouse Embryonic Fibroblast (MEF) medium (Mori et al., 2019).The fibroblasts were passaged using Accutase (Innovative Cell Technologies, AT104) and seeded on gelatin (Millipore-Sigma, ES006B)coated 6-well plates with a density of 0.1 million cells per well.Upon cell attachment, Yamanaka reprogramming factors were induced to iPSCs via Sendai virus using CytoTune2.0(Thermo Fisher, A16517).To establish nGFP + iPSCs, the Cre plasmid was transfected using Fugene HD transfection reagent (Promega, E2311), then GFP + tdTomato − live cells were sorted out by SONY MA900, and single clones were expanded.

Morphometric analysis
To determine the relative number of specific cell populations, 5 non-overlapping random fields per mouse were analyzed (×40 magnification) after capturing the images by confocal microscopes (see above).We counted the number of GFP + cells or tdTomato + cells co-immunostained with specific antibodies for each field.Hoechst co-staining was used to determine the cell number for each lineage as identified by differentiation markers.Those analyses were processed using ImageJ (NIH).

Statistical analysis
analysis was performed using Prism 8. Data acquired by performing biological replicas of two or three independent experiments are presented as the mean ± SD.Statistical significance was determined using a two-tailed t-test and unpaired one-or two-way analysis of variance with the Tukey post hoc test.*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: non-significant.

Figure 1
Figure 1 continued on next page

Figure supplement 1 .
Figure supplement 1. Foxa2-lineage gradually increased in the mesenchyme and endothelial cells during mouse lung development.

Figure 2 continuedFigure 3 .
Figure2 continued distal lung bud epithelium (SOX9), mesenchyme (TBX4), and nucleus (Hoechst) markers.The budding structures expressed SOX9 and NKX2-1 (asterisk), and monolayer cells expressed TBX4.Scale bar = 100 μm.(D) RT-PCR analyses of lung mesenchyme and epithelium markers in time course according to the protocol shown in (A).Each plot showed a different biological experiment (n = 3 independent experiments).Error bars represent mean ± standard deviation (SD).(E) Flow cytometry (FCM)-based protein kinetic analyses during definitive endoderm (DE) and lateral plate mesoderm (LPM) induction; MIXL1 expression preceded compared to PDGFRα or FOXA2.FOXA2 appearance in the subset of the PDGFRα + population (red asterisk) (n = 3 independent experiments).(F) Representative IF imaging of 36-hr-cultured hiPSCs.Scale bar = 100 μm.(G) Schematic summary of FCM analysis.(H) Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR) analyses further confirmed the preceded MIXL1 induction and subsequent expression of PDGFRα and FOXA2.All graphs: Data normalized by undifferentiated induced pluripotent stem cells (iPSCs).Each plot showed a different biological experiment (n = 3 independent experiments).Error bars represent mean ± SD.The online version of this article includes the following source data and figure supplement(s) for figure 3: Source data 1.qRT-PCR analyses of lung mesenchyme and epithelium markers in time course of directed differentiation.Source data 2. qRT-PCR analyses of mesendoderm markers in time course of directed differentiation.

Figure supplement 1 .
Figure supplement 1. Endodermal and mesodermal lung progenitors develop together in the directed differentiation using hiPSC.

Figure
Figure supplement 1-source data 2. % GFP in CD45 + hematopoietic cells analyzed by flow cytometry.

Figure supplement 4
Figure supplement 4-source data 1.Morphometric analysis of E17.5 chimeric lungs in each lung mesenchyme marker.

Figure supplement 5
Figure supplement 5-source data 1.Morphometric immunofluorescence (IF) analysis of the chimeric trachea: % chimerism of tracheal epithelium and mesenchyme.

Figure 4 continued
Figure 4 continued

Figure supplement 1 .
Figure supplement 1. Summary of the results and proposed models.
Figure 5 continued