The article describes step wise directed differentiation of induced pluripotent stem cells to three-dimensional whole lung organoids containing both proximal and distal epithelial lung cells along with mesenchyme.
Human lung development and disease has been difficult to study due to the lack of biologically relevant in vitro model systems. Human induced pluripotent stem cells (hiPSCs) can be differentiated stepwise into 3D multicellular lung organoids, made of both epithelial and mesenchymal cell populations. We recapitulate embryonic developmental cues by temporally introducing a variety of growth factors and small molecules to efficiently generate definitive endoderm, anterior foregut endoderm, and subsequently lung progenitor cells. These cells are then embedded in growth factor reduced (GFR)-basement membrane matrix medium, allowing them to spontaneously develop into 3D lung organoids in response to external growth factors. These whole lung organoids (WLO) undergo early lung developmental stages including branching morphogenesis and maturation after exposure to dexamethasone, cyclic AMP and isobutylxanthine. WLOs possess airway epithelial cells expressing the markers KRT5 (basal), SCGB3A2 (club) and MUC5AC (goblet) as well as alveolar epithelial cells expressing HOPX (alveolar type I) and SP-C (alveolar type II). Mesenchymal cells are also present, including smooth muscle actin (SMA), and platelet-derived growth factor receptor A (PDGFRα). iPSC derived WLOs can be maintained in 3D culture conditions for many months and can be sorted for surface markers to purify a specific cell population. iPSC derived WLOs can also be utilized to study human lung development, including signaling between the lung epithelium and mesenchyme, to model genetic mutations on human lung cell function and development, and to determine the cytotoxicity of infective agents.
The lung is a complicated, heterogeneous, dynamic organ that develops in six distinct stages – embryonic, pseudoglandular, canalicular, saccular, alveolar, and microvascular maturation1,2. The latter two phases occur pre and postnatally in human development while the first four stages occur exclusively during fetal development unless preterm birth occurs3. The embryonic phase begins in the endodermal germ layer and concludes with the budding of the trachea and lung buds. Lung development occurs in part via signaling between the epithelial and mesenchymal cells4. These interactions result in lung branching, proliferation, cellular fate determination and cellular differentiation of the developing lung. The lung is divided into conducting zones (trachea to the terminal bronchioles) and respiratory zones (respiratory bronchioles to the alveoli). Each zone contains unique epithelial cell types; including basal, secretory, ciliated, brush, neuroendocrine, and ionocyte cells in the conducting airway5, followed by alveolar type I and II cells in the respiratory epithelium6. Much is still unknown about the development and response to injury of the various cell types. iPSC derived lung organoid models enable the study of mechanisms that drive human lung development, the effects of genetic mutations on pulmonary function, and the response of both the epithelium and mesenchyme to infectious agents without the need for primary human lung tissue.
Markers corresponding to the various stages of embryonic differentiation include CXCR4, cKit, FOXA2, and SOX17 for definitive endoderm (DE)7, FOXA2, TBX1, and SOX2 for anterior foregut endoderm (AFE)8, and NKX2-1 for early lung progenitor cells9. In embryonic lung development, the foregut divides into the dorsal esophagus and ventral trachea. The buds of the right and left lungs appear as two independent outpouchings around the tracheal bud10. During branching morphogenesis, the mesenchyme surrounding the epithelium produces elastic tissue, smooth muscle, cartilage, and vasculature11. The interaction between the epithelium and mesenchyme is essential for normal lung development. This includes the secretion of FGF1012 by the mesenchyme and SHH13 produced by the epithelium.
Here, we describe a protocol for the directed differentiation of hiPSCs into three-dimensional (3D) whole lung organoids (WLO). While there are similar approaches that incorporate isolation of lung progenitor cells via sorting at the LPC stage to make alveolar-like14,15 (distal) organoids or airway16 (proximal) organoids, or generate ventral-anterior foregut spheroids to make human lung organoids expressing alveolar-cell and mesenchymal markers and bud tip progenitor organoids17, the strength of this method is the inclusion of both lung epithelial and mesenchymal cell types to pattern and orchestrate lung branching morphogenesis, maturation, and expansion in vitro.
This protocol uses small molecules and growth factors to direct the differentiation of pluripotent stem cells through definitive endoderm, anterior foregut endoderm, and lung progenitor cells. These cells are then induced into 3D whole lung organoids through important developmental steps, including branching and maturation. The summary of the differentiation protocol is shown in Figure 1a with representative brightfield images of endodermal and organoid differentiation shown in Figure 1b. Figure 1c,d show the gene expression details of endodermal differentiation as well as the gene expression of both the proximal and distal populations of lung epithelial cells after completing the differentiation.
This study protocol was approved by the Institutional Review Board of UCSD's Human Research Protections Program (181180).
1. Definitive endoderm induction from induced pluripotent stem cells (Day 1 – 3)
2. Anterior foregut endoderm (AFE) induction (Day 4 – 6)
3. Lung progenitor cell (LPC) differentiation (Day 7 – 16)
4. 3D lung organoid induction (Day 16 – 22)
5. 3D Lung organoid branching (Day 23 – 28)
6. 3D lung organoid maturation (Day 29 – 34)
7. 3D Lung organoid immunocytochemistry
8. Removal of whole lung organoids from GFR-basement membrane matrix medium for passage, FACS, or cryopreservation
24 hours after plating, day 1, iPSCs should be 50%-90% confluent. On day 2, DE should be 90%-95% confluent. During DE induction, it is common to observe significant cell death on day 4 but attached cells will retain a compact cobblestone morphology (Figure 2b). Discontinue differentiation if the majority of adherent cells detach and consider shortening exposure to DE media with activin A by 6-12 h. During AFE induction, cell death is minimal, and cells remain adherent, but will appear smaller and more heterogeneous. Passaging the cells on day 7 must only be done if the yield of double positive SOX2 and FOXA2 is >80%. After passaging into basement membrane matrix for 3D LPC induction, small spheroids will first appear, then grow and some may begin to branch. Gene expression profiles for successful endodermal differentiation include increased SOX17 at DE, increased FOXA2 and SOX2 with decreasing SOX17 and the first appearance of NKX2-1, and increased NKX2-1, along with the presence of SOX2 and FOXA2. Consistent with early embryonic development, the ventralization of AFE occurs for lung bud development (NKX2-1+) and dorsalization of AFE occurs for gastrointestinal development (SOX2+). Cultures at LPC will have a mix of both lung and gastric progenitors.
Lung organoid induction from LPC has been performed using various methods. Some groups sort the cells using NKX2-1 fluorescent reporters or a surface antigen proxy (CPM, CD26lowCD47high). But those lung organoids contain alveolar type II like cells without alveolar type I cells or mesenchyme. Other groups have collected cell clumps that bud off the AFE/LPC monolayer and embedded them into basement membrane matrix. Those organoids contain a mixed population of lung epithelial and mesenchymal cells but take months to culture19. Our protocol includes both the presence of epithelial and mesenchymal cells. The WLOs express proximal epithelial cell markers p63 and KRT5 (basal cells) and SCGB3A2 (club cells) as well as distal epithelial cell markers HOPX (ATI) and proSPC, SPB, and NKX2-1 (ATII). They also express the mesenchymal marker Vimentin at the LPC stage, as well as in the whole lung organoids. PDGFRα is a marker for fibroblasts that have an important function in the lung during sacculation and alveolarization20 and is co-expressed with the transcription factor important in distal cell differentiation, SOX9 (Figure 3).
Our method efficiently generates NKX2-1-expressing LPC 3D cultures using signaling molecules that occur in fetal lung development to form early lung organoids. When passaging LPCs into GFR-basement membrane matrix medium for lung organoid induction, it is imperative not to over-dissociate into a single cell suspension, but to instead to retain small clumps of cells (10 cells/clump). Cell counting will not be completely accurate, but still necessary to avoid over confluence during the 3-week lung organoid differentiation.
Lung organoid induction should yield small, branching organoids by day 6 of induction (day 23 of differentiation). These should continue to grow during the organoid branching step and maturation step. Twenty-four hours after the introduction of dexamethasone, cAMP, and IBMX, the branches should expand into transparent spheres. Whole lung organoid analysis can be performed at the end of the differentiation, or the WLOs can be passage into fresh basement membrane matrix with GFR or cryopreserved by freezing down in 10% DMSO.
Figure 1: Overall schematic of whole lung organoid (WLO) differentiation from hiPSCs and representative data. (a) Schematic of WLO differentiation from hiPSCs. Circles represent endodermal cell type with identifying markers. Timeline of differentiation is indicated in black bars. Growth factors and/or small molecules for induction of endodermal and lung organoid populations. In summary, stem cells are differentiated into definitive endoderm, anterior foregut endoderm and into lung progenitor cells in approximately 16 days. These cells are then passage into GFR-basement membrane matrix medium containing medium inserts and undergo lung organoid induction, branching, and maturation. The total differentiation takes approximately 35 days. (b) Representative phase contrast images of the cells at major endodermal stages and 3D images of whole lung organoids. Scale bar size as indicated in panel. (c) qRT-PCR analysis of lung development markers during endoderm and (d) whole lung organoid differentiation of proximal and distal cell markers. All data represents an average of 3-5 biological replicates. Error bars represent standard error of the mean and are normalized to actin. Please click here to view a larger version of this figure.
Figure 2: Characterization of endoderm differentiation by flow cytometry and immunocytochemistry. (a) Flow cytometry of definitive endoderm marker CXCR4. Left panel shows the gating against the unstained population while the middle panel shows the CXCR4 positive population. The right panel shows immunocytochemistry image of SOX17 (red) overlaid with nuclei (blue). (b) Immunocytochemistry image of AFE markers FOXA2 and SOX2 overlaid with nuclei (blue). (c) Endogenous expression of NKX2-1-GFP in a reporter cell line in 3D LPC. Images taken from live cell culture in brightfield and GFP. Flow cytometry of lung progenitor intracellular marker NKX2-1 after fixation and permeabilization. Scale bar size = 50 µM. Please click here to view a larger version of this figure.
Figure 3: Characterization of 3D whole lung organoids after 3-week differentiation by immunocytochemistry. (a) Proximal lung markers. Left panel shows SOX2 (white) and SOX9 (red) overlaid by nuclei (blue). These markers are important in branching morphogenesis and represent the proximal and distal epithelial populations. Middle panels show P63 (red) and KRT5 (red), both markers of basal cells. The right panel shows SCGB3A2, a marker of club cells. (b) Distal lung markers. Left panel depicts pro-SPC (PSPC), (green) and HOPX (red), markers of alveolar type II ad I cells, respectively, overlaid with nuclei (blue). Middle panel shows pro-SPC (PSPC) (green) and SPB (red), markers of alveolar type II cells, overlaid with nuclei (blue). The right panel shows NKX2-1 (red) and ZO1 (green) overlaid with nuclei (blue). (c) Markers of lung mesenchyme. Left panel shows PDGFRA (red) and SOX9 (white), representing distal mesenchyme. Right panel shows Vimentin (red), which is dispersed throughout the lung. Scale bar size = 50 µM. Please click here to view a larger version of this figure.
REAGENTS AND SOLUTIONS – For company names please see the Table of Materials List |
3D organoid induction medium (day 17-22) |
Serum-free basal medium (see recipe) supplemented with: |
FGF7 (10 ng/mL) |
FGF10 (10 ng/mL) |
CHIR99021 (3 μM) |
EGF (10 ng/mL) |
3D organoid branching medium (day 23-28) |
Serum-free basal medium (see recipe) supplemented with: |
FGF7 (10 ng/mL) |
FGF10 (10 ng/mL) |
CHIR99021 (3 μM) |
All-trans retinoic acid (0.1 μM) |
EGF (10 ng/mL) |
VEGF/PIGF (10 ng/mL) |
3D organoid maturation medium (day 29-34) |
Serum-free basal medium (see recipe) supplemented with: |
Dexamethasone (50 nM) |
Br-cAMP (100 μM) |
IBMX (100 μM) |
AFE induction medium (day 4-6) |
Serum-free basal medium (see recipe) supplemented with: |
SB431542 (10 μM) |
Dorsomorphin (2 μM) |
DE induction medium (day 1-3) |
48.5 mL RPMI1640 + Glutamax |
1 mL B27 without retinoic acid |
500 μl HEPES (1%) |
500 μl pen/strep |
Human activin A (100 ng/mL) |
CHIR99021 (5 μM) – only in the first 24 hours |
LPC induction medium (day 7-16) |
Serum-free basal medium (see recipe) supplemented with: |
BMP4 (10 ng/mL) |
All-trans retinoic acid (RA) (0.1 μM) |
CHIR99021 (3 μM) |
Quenching medium |
49 mL DMEM/F12 |
1 mL FBS |
Serum-free basal medium (SFBM) |
375 mL Iscove’s Modified Dulbecco’s Medium (IMDM) + Glutamax |
125 mL Ham’s F12 |
5 mL B27 without retinoic acid |
2.5 mL N2 |
500 μl ascorbic acid, 50 mg/mL |
13 μl monothioglycerol/1ml of IMDM” use 300ul of 0.4mM monothioglycerol per 100ml of serum free media |
3.75 mL bovine serum albumin (BSA) Fraction V, 7.5% solution |
500 μl pen/strep |
Stem cell passaging medium (day 0) |
500 mL DMEM/F12 |
129 mL Knockout serum replacement (KSR) |
6.5 mL Glutamax |
6.5 mL NEAA |
1.3 mL 2-mercaptoethanol |
6.5 mL pen/strep |
Table 1: Table of media.
Problem | Solution |
DE differentiation not efficient | 24 hours after plating in stem cell medium, cells should be 50-70% confluent |
GSK3β inhibitor/Wnt activator CHIR99021 should be removed within 20-24 hours of day 1 DE induction | |
DE differentiation should not exceed a total of 72 hours | |
AFE differentiation not efficient | Ensure that DE differentiation was successful and the cells express > 80% CXCR4 |
Ensure fresh growth factors/small molecules are being added to the media daily | |
LPC differentiation not efficient | Ensure the AFE differentiation was successful and the cells express > 80% FOXA2/SOX2 |
Ensure the AFE cells are passaged as aggregates of 4-10 cells and not single celled | |
3D lung organoids not growing or differentiating | Ensure the LPC differentation was successful and the cells express > 30% NKX2-1 |
Ensure the LPCs were passaged as aggregates of 4-10 cells and not single celled | |
Ensure there is no residual matrigel from the LPCs during passaging | |
Add ROCK Inhibitor Y-27632 with each media change | |
Ensure the media is changed on time and fresh growth factors/small molecules are added | |
Ensure concentration of growth factors/small molecules is correct |
Table 2: Troubleshooting.
The successful differentiation of 3D whole lung organoids (WLO) relies on a multi-step, 6-week protocol with attention to detail, including time of exposure to growth factors and small molecules, cellular density after passaging, and the quality of hiPSCs. For troubleshooting, see Table 2. hiPSCs should be approximately 70%-80% confluent, with less than 5% spontaneous differentiation prior to dissociation. This protocol calls for "mTeSR plus" medium; however, plain "mTeSR" medium has also been used with comparable results and is less expensive. For the extracellular matrix, we use GFR basement membrane matrix medium. We passage the hiPSCs using ReLesR (see Table of Materials) to reduce differentiation.
During endoderm differentiation, cells should be visualized daily prior to and after media changes. Specified growth factors/small molecules should be added fresh daily to the base medium to prevent premature degradation. Cell death in definitive endoderm (DE) is common but should be limited during anterior foregut endoderm (AFE) and lung progenitor cell (LPC) induction. If there is a large die off on the third day of DE (day 4), decrease total time of DE by 6-12 h. New iPSC cell lines may need to be optimized for successful endoderm differentiation. Perform flow cytometry at DE for CXCR4 to confirm successful induction (>85% CXCR4 + cells). The cells should be relatively stable at AFE and will change morphologically with little die off.
Passaging into LPC is another process that must be optimized for cell type. Replating cells at too low a density (<50%) will result in inefficient differentiation. Confirm successful LPC induction with immunocytochemistry for NKX2-1 or flow cytometry for CPM18 or CD26low/CD47high 15. Successful LPC induction must have >40% NKX2-1, otherwise the organoids will have greater abundance of dorsal AFE. For LPC induction, growth factors must be added to base media with each media change. If the media becomes yellow later into LPC induction, consider increasing the volume of fresh media, or change the media every day. During 3D whole lung organoid induction, plating number and maintaining cell clusters are key to successful organoid growth. GFR-basement membrane matrix medium is difficult to handle and highly temperature sensitive, so always keep it on ice. If the GFR-basement membrane matrix medium gels too early, then the LPC cells will not integrate within it. We recommend thawing 1 mL aliquots of GFR-basement membrane matrix medium on ice 30 min prior to passaging. Once the cells/clusters have been counted and appropriate aliquots made, place the cell pellet on ice. We suggest preparing plates, labeling, and addition of cell culture membrane inserts prior to GFR-basement membrane matrix medium handling.
Use pipette tips to add correct volume of liquid GFR-basement membrane matrix medium immediately to cell pellet, keeping on ice. Pipette up and down quickly but gently (nuanced handling) to not introduce bubbles, then place the tube back on ice. Add the cell and GFR-basement membrane matrix medium mixture to the apical portion of the transwell in prepared plates. The mixture should spread and coat the entire transwell; gently tilt the plate to ensure coating. After gelling in the incubator for 30-60 min, there should be visible cell clusters in the GFR-basement membrane matrix medium. Add appropriate lung induction media to the basal chamber supplemented with 10 µM ROCK Inhibitor Y-27632 and monitor organoid growth every other day.
Future applications of the organoids generated by this protocol include studying the molecular pathways that control early lung lineage commitment and cell fate specification21,22,23. The interaction between the epithelium and mesenchyme can be determined by utilizing gene knock out models24. The organoids could also be co-cultured with endothelial cells to determine the importance of tissue specific co-pattern signaling between the lung epithelium, mesenchyme, and the endothelium25. Lung development occurs in parallel with vascular development and that relationship may elicit important molecular mechanisms necessary for proper lung development. We have also shown that these whole lung organoids are functional through surfactant secretion assays after GFR-basement membrane matrix medium was removed followed by short-term culture in ultra-low attachment wells26. Other strategies include sorting the cells for cell surface markers such as NGFR (basal cells)27 and HTII-280 (ATII cells)28 and replacing them as homogenous organoids or a monolayer in air liquid interphase culture conditions. Whole, proximal, and distal lung organoids have also been used to study the cellular targets and pathophysiology of SARS-CoV-2 viral infection in order to better understand and screen for drugs that might combat COVID-19.
This protocol is robust and reproducible, but many challenges still exist. Many different iPSC and ESC lines have been tested (>20 lines) but the protocol must be optimized for each cell line. Despite robust DE and AFE induction, LPC induction may be difficult to achieve >40% of NKX2-1 + cells. Other protocols include a sorting step for surface markers of NKX2-1 cells15,18, but those only yield alveolar type II like organoids without mesenchyme and still contain gastric and hepatic cell populations despite purifying the lung progenitor population29. We have also noted a small amount of gastric and hepatic cells in both the LPC and whole lung organoids, possibly due to the presence of dorsal anterior foregut cells contaminating the LPCs. Therefore, the differentiation of pure lung organoids is yet to be achieved, and more research on the development of the lung progenitor cells in human tissue must be completed. Downstream assays must be vigorously benchmarked with gene and protein expression from primary human lung tissue. While, to date, the most fruitful use of lung organoids has been in modeling diseases and screening for drugs in vitro, the transplantation of hiPSC-derived lung organoids into patients for regenerative medicine has been contemplated as a future therapy for a range of conditions. However, prior to considering such interventions, a good deal of quality control must be perfected, including the identification and removal of contaminating, undesirable, potentially tumorigenic hiPSC derivatives. Furthermore, better functional assays in vitro and better animal models of pulmonary disease still need to be developed.
Specifically, the functionality and safety of hiPSC-derived cells must be confirmed. Undifferentiated cells need to be excluded since they have the capacity to generate teratomas. One method to determine undifferentiated stem cells in definitive endoderm is to sort the cells out using the pluripotency marker SSEA4. Marker genes for undifferentiated hiPSCs were recently detected using single cell RNA sequencing30. ESRG, CNMD, and SFRP2 can be used to validate undifferentiated cells at any differentiation step. Once purity is confirmed, the benefit of autologous iPSC derived therapies is the ability for the transplanted cells to avoid rejection since they come from the patient's own cells. The drawbacks include the time it takes to fully differentiate the cells, undergo rigorous clinical grade testing, and transplant the cells into a patient with an acute injury (respiratory distress syndrome, myocardial infarction, or spinal injury). The alternative is to utilize banked allogenic iPSC derived cells31. These may be stored and readily available for patients with human leukocyte antigen (HLA) matched donors and they will have undergone thorough testing for contamination. The biggest drawback is the possibility of immune rejection. Immunosuppression may be necessary in allogenic cell transplantation, which is the current reality of allogenic whole tissue transplants. Strategies are being devised to allow the allogenic iPSC derived cells to evade the immune response to safely transplant them into patients32.
Eventually, iPSC derived whole lung organoids will be utilized to study patient-specific disease models, tailor therapeutics, and enhance regenerative medical research.
The authors have nothing to disclose.
This research was supported by the California Institute for Regenerative Medicine (CIRM) (DISC2-COVID19-12022).
Cell Culture | |||
12 well plates | Corning | 3512 | |
12-well inserts, 0.4um, translucent | VWR | 10769-208 | |
2-mercaptoethanol | Sigma-Aldrich | M3148 | |
Accutase | Innovative Cell Tech | AT104 | |
ascorbic acid | Sigma | A4544 | |
B27 without retinoic acid | ThermoFisher | 12587010 | |
Bovine serum albumin (BSA) Fraction V, 7.5% solution | Gibco | 15260-037 | |
Dispase | StemCellTech | 7913 | |
DMEM/F12 | Gibco | 10565042 | |
FBS | Gibco | 10082139 | |
Glutamax | Life Technologies | 35050061 | |
Ham’s F12 | Invitrogen | 11765-054 | |
HEPES | Gibco | 15630-080 | |
Iscove’s Modified Dulbecco’s Medium (IMDM) + Glutamax | Invitrogen | 31980030 | |
Knockout Serum Replacement (KSR) | Life Technologies | 10828028 | |
Matrigel | Corning | 354230 | |
Monothioglycerol | Sigma | M6145 | |
mTeSR plus Kit (10/case) | Stem Cell Tech | 5825 | |
N2 | ThermoFisher | 17502048 | |
NEAA | Life Technologies | 11140050 | |
Pen/strep | Lonza | 17-602F | |
ReleSR | Stem Cell Tech | 5872 | |
RPMI1640 + Glutamax | Life Technologies | 12633012 | |
TrypLE | Gibco | 12605-028 | |
Y-27632 (Rock Inhibitor) | R&D Systems | 1254/1 | |
Growth Factors/Small Molecules | |||
Activin A | R&D Systems | 338-AC | |
All-trans retinoic acid (RA) | Sigma-Aldrich | R2625 | |
BMP4 | R&D Systems | 314-BP/CF | |
Br-cAMP | Sigma-Aldrich | B5386 | |
CHIR99021 | Abcam | ab120890 | |
Dexamethasone | Sigma-Aldrich | D4902 | |
Dorsomorphin | R&D Systems | 3093 | |
EGF | R&D Systems | 236-EG | |
FGF10 | R&D Systems | 345-FG/CF | |
FGF7 | R&D Systems | 251-KG/CF | |
IBMX (3-Isobtyl-1-methylxanthine) | Sigma-Aldrich | I5879 | |
SB431542 | R&D Systems | 1614 | |
VEGF/PIGF | R&D Systems | 297-VP/CF | |
Primary antibodies | Dilution rate | ||
CXCR4-PE | R&D Systems | FAB170P | 1:200 (F) |
HOPX | Santa Cruz Biotech | sc-398703 | 0.180555556 |
HTII-280 | Terrace Biotech | TB-27AHT2-280 | 0.145833333 |
KRT5 | Abcam | ab52635 | 0.180555556 |
NKX2-1 | Abcam | ab76013 | 0.25 |
NKX2-1-APC | LS-BIO | LS-C264437 | 1:1000 (F) |
proSPC | Abcam | ab40871 | 0.215277778 |
SCGB3A2 | Abcam | ab181853 | 0.25 |
SOX2 | Invitrogen | MA1-014 | 0.180555556 |
SOX9 | R&D Systems | AF3075 | 0.180555556 |
SPB (mature) | 7 Hills | 48604 | 1: 1500 (F) 1:500 (W)a |
SPC (mature) | LS Bio | LS-B9161 | 1:100 (F); 1:500 (W) a |