Standardized pipeline for establishing, expanding, and differentiating airway and alveolar organoids from human BAL fluid

Abstract

Lung organoids are versatile experimental models, but their broader use in studying human disease is limited by the scarcity of starting material and the complexity of current methods. To align organoid technology with common clinical practice, we developed airway and alveolar organoids using cells obtained from patients’ bronchoalveolar lavage (BAL) fluid. Building on existing techniques, we showed that BAL is a reliable, accessible source of primary human epithelial cells, yielding airway and alveolar organoids within 10 days. Organoids can then be expanded over many passages for downstream analysis. Our streamlined methods do not require cell sorting or other complex procedures, all cells are derived from a single patient, and media are based on serum-free, chemically-defined formulations. Here, we present detailed protocols for organoid establishment, standardized passaging and phenotyping, and differentiation of both airway and alveolar models. We provide a time course of BAL-derived airway organoid differentiation at air-liquid interface, and we demonstrate proof of principle for differentiation of BAL-derived alveolar organoids in 3D culture. These methods can be readily adapted to generate and characterize organoids from lung tissue, tracheobronchial specimens, or other primary cells from humans or mice, expanding the potential to use lung organoids for disease modeling.

NEW AND NOTEWORTHY

We provide streamlined protocols to generate both airway and alveolar epithelial organoids from a single, clinical BAL specimen. From standardized specimen collection to organoid plating to passaging and differentiation, we show that rare, primary epithelial cells in BAL can give rise to all the major airway and alveolar cell types. Our serum-free, feeder-free, sorting-free methods offer a simplified starting point for using patient-derived organoids to model lung disease.

Graphical Abstract

INTRODUCTION

Organoids are three-dimensional (3D), self-assembled groups of cells that mimic a particular tissue, incorporating multiple cell types and often recapitulating form and function in ways that traditional 2D culture cannot (1). Human lung organoids vary widely in their starting material, cellular composition, culture format, media ingredients, and matrix material (2). Here, we focus on human adult organoids from primary cells that are grown in conditions that favor either airway or alveolar epithelial cell expansion. Airway organoids were first generated from basal stem cells in primary bronchial epithelial tissue (3). Alveolar organoids were first established using alveolar type 2 (AT2) stem cells from lung tissue, requiring MRC5 cells (a fetal human lung fibroblast cell line) for support (4). These and subsequent systems opened new windows into human biology and disease, but there were key limitations. Cultures generally survived only a few passages. The media contained animal serum, which has unknown and variable components that can change cellular identity. The procedures were labor-intensive, calling for lung harvesting, digestion, and cell sorting on the same day, while supporting cells had to be grown separately and harvested in time to co-culture with epithelial organoids.

Long-term expanding airway organoids were developed in 2019 (5), and alveolar organoids followed in 2020 (6, 7), showing that primary human stem cells could be maintained for months in vitro without losing their identity. These newer methods used defined chemicals and growth factors to modulate specific pathways known to drive lung stem cell development, maintenance, and differentiation, supplanting the role of supporting cells. These “serum-free, feeder-free” approaches made cultures more reproducible and simplified the organoid workflow. Nonetheless, most methods required sorting cells from lung tissue, which is difficult to obtain and not available for most disease states. Organoids derived from induced pluripotent stem cells (iPSCs) provided a more flexible cell source but require complex manipulation (8–11). A major advance came with showing that culture media alone is sufficient to select for airway cell outgrowth, without cell sorting. It became possible to establish airway organoids from bronchoalveolar lavage (BAL) fluid (5), tracheal aspirates (12), and nasal brushings (13)—accessible sources of primary cells that contain only rare stem cells proficient at organoid formation. Our group further simplified and characterized airway organoids from human BAL fluid and established alveolar organoids from BAL as well (14). Combining serum-free, feeder-free, sorting-free approaches, we can now progress from BAL collection to organoid plating in approximately 4 hours using standard laboratory equipment and techniques. Here, we present updated protocols for culturing airway and alveolar BAL organoids that further standardize sample collection, passaging, and phenotyping. We also demonstrate examples of differentiating airway and alveolar stem cells in 2D and 3D formats, respectively, showing that BAL organoids remain capable of recapitulating a full array of relevant cell types for disease modeling and advanced applications.

MATERIALS AND METHODS

Materials

All materials and catalog numbers are listed in Table 1, along with antibody dilutions.

Table 1.

Complete list of materials and catalog numbers, with antibody dilutions.

REAGENT OR RESOURCE	SOURCE	CATALOG NUMBER
Cell Culture Media Components
Advanced DMEM/F-12	Thermo Fisher	12634010
GlutaMAX Supplement	Thermo Fisher	35050061
HEPES (1X)	Thermo Fisher	15630080
Primocin	InvivoGen	ant-pm-1
B-27 Supplement (50X)	Thermo Fisher	17504044
N-2 Supplement (100X)	Thermo Fisher	17502048
Insulin-Transferrin-Selenium (100X)	Thermo Fisher	41400045
Nicotinamide	Millipore Sigma	N0636
N-Acetyl-L-cysteine	Millipore Sigma	A9165
Heparin	Millipore Sigma	H3149
Recombinant Human EGF	Thermo Fisher	AF-100-15
Recombinant Human R-Spondin-1	Thermo Fisher	120-38
Recombinant Human Noggin	Thermo Fisher	120-10C
Recombinant Human FGF-10	Thermo Fisher	100-26
Recombinant Human FGF-7 (KGF)	Thermo Fisher	100-19
Recombinant Human Heregulinβ-1	Thermo Fisher	100-03
Recombinant Human IL-1β	Thermo Fisher	200-01B
CHIR 99021	Tocris	4423
A 83-01	Millipore Sigma	SML0788
SB 202190	Tocris	1264
SB 431542	Tocris	1614
BIRB 796	Tocris	5989
ROCK Inhibitor Y27632	ATCC	ACS-3030
Amphotericin B	Millipore Sigma	A2942
Gentamicin	Thermo Fisher	15710064
PneumaCult-Ex Plus Medium	StemCell Technologies	05040
PneumaCult-ALI Medium	StemCell Technologies	05001
PneumaCult-ALI-S Medium	StemCell Technologies	05050
Hydrocortisone Stock Solution	StemCell Technologies	07925
Human Serum	Millipore Sigma	H4522
IMDM	Thermo Fisher	12440053
Ham’s F-12	Corning	10-080-CV
Bovine Serum Albumin (7.5% w/v solution)	Thermo Fisher	15260037
L-Ascorbic Acid	Millipore Sigma	A4544
1-Thioglycerol (MTG)	Millipore Sigma	M6145
TRULI (LATS1/2 Inhibitor)	MedChem Express	HY-138489
Dexamethasone	Millipore Sigma	D4902
8-Bromoadenosine 3’,5’-Cyclic Monophosphate Sodium Salt (cAMP)	Millipore Sigma	B7880
3-Isobutyl-1-Methylxanthine (IBMX)	Millipore Sigma	I5879
Antibodies and Dyes
Chicken Polyclonal IgY anti-Human KRT5 (1:500)	BioLegend	905901, RRID:AB_2565054
Goat Monoclonal IgG anti-Human KRT5 (1:500)	Abcam	ab321868
Goat Polyclonal IgG anti-Human KRT13 (1:200)	Abcam	ab79279, RRID:AB_2281128
Rabbit Monoclonal IgG anti-Human Acetyl-α-Tubulin (1:200)	Cell Signaling Technology	5335S, RRID:AB_10544694
Mouse Monoclonal IgG1k anti-Human SCGB1A1 (1:200)	Santa Cruz Biotechnology	sc-365992, RRID:AB_10915481
Rabbit Polyclonal IgG anti-Human SCGB1A1 (1:200)	Thermo Fisher	26909-1-AP, RRID:AB_2880680
Goat Polyclonal IgG anti-Human SCGB3A2 (1:200)	R&D Systems	AF3545, RRID:AB_2183543
Mouse Monoclonal IgG1k anti-Human MUC5AC (1:200)	Thermo Fisher	MA5-12178, RRID:AB_10978001
Rabbit Polyclonal IgG anti-Human pro-SPC (1:500)	Seven Hills Bioreagents	WRAB-9337, RRID:AB_2335890
Goat IgG anti-Human AGER (1:200)	R&D Systems	AF1145, RRID:AB_354628
Donkey anti-Chicken IgY Highly Cross Adsorbed, Alexa Fluor 647 (1:500)	Thermo Fisher	A78952, RRID:AB_2921074
Donkey anti-Rabbit IgG Highly Cross-Adsorbed, Alexa Fluor Plus 488 (1:500)	Thermo Fisher	A32790, RRID:AB_2762833
Donkey anti-Mouse IgG Highly Cross-Adsorbed, Alexa Fluor Plus 594 (1:500)	Thermo Fisher	A32744, RRID:AB_2762826
Donkey anti-Mouse IgG Highly Cross-Adsorbed, Alexa Fluor Plus 647 (1:500)	Thermo Fisher	A32787, RRID:AB_2762830
Donkey anti-Goat IgG Highly Cross-Adsorbed, Alexa Fluor Plus 594 (1:500)	Thermo Fisher	A32758, RRID:AB_2762828
Phalloidin, Alexa Fluor Plus 405	Thermo Fisher	A30104
DAPI	Millipore Sigma	D9542
PE anti-Human CD45 Antibody (1:100)	BioLegend	304008, RRID:AB_314396
PE/Cyanine7 anti-Human CD326 (EpCAM) Antibody (1:100)	BioLegend	324222, RRID:AB_2561506
APC anti-Human CD271 (NGFR) Antibody (1:100)	BioLegend	345108, RRID:AB_10645515
Mouse IgM anti-Human HT2-280 (1:50)	Terrace Biotech	TB-27AHT2-280, RRID:AB_2832931
Goat anti-Mouse IgM Cross Adsorbed Secondary Antibody, DyLight 488 (1:200)	Thermo Fisher	SA5-10150, RRID:AB_2556730
UltraComp eBeads Compensation Beads	Thermo Fisher	01-2222
Other Reagents
Matrigel Matrix for Organoid Culture, Phenol Red-free, LDEV-free	Corning	356255
Liberase TM	Millipore Sigma	5401127001
Deoxyribonuclease I (DNase)	Millipore Sigma	D4527
TrypLE Express Enzyme (1X), No Phenol Red	Thermo Fisher	12604021
Bambanker Cell Freezing Medium	Bulldog Bio	BB05
Cryo-SFM Plus Freezing Medium	PromoCell	C-29922
MycoFluor Mycoplasma Detection Kit	Thermo Fisher	M7006
Trypan Blue Solution, 0.4%	Thermo Fisher	15250061
DPBS, 1X without Calcium and Magnesium	Corning	21-031-CV
EDTA	Millipore Sigma	E6758
Dimethyl Sulfoxide	Millipore Sigma	472301
Neutral Buffered Formalin, 10%	Fisher Scientific	STL286001
Triton X-100	Millipore Sigma	T8787
Normal Donkey Serum	Jackson ImmunoResearch	017-000-121
ProLong Gold Antifade Mountant	Thermo Fisher	P36930
Other Materials
12-well Clear Flat Bottom TC-treated Plate, Sterile	Corning	353043
24-well Clear Flat Bottom TC-treated Plate, Sterile	Corning	353047
48-well Clear Flat Bottom TC-treated Plate, Sterile	Corning	353078
6.5 mm Transwell with 0.4 μm Pore Polyester Membrane Insert, Sterile	Corning	3470
Light Sensitive Amber 50 ml Centrifuge Tubes	VWR	89079-535
2-Chip Disposable Hemocytometer	Bulldog Bio	DHC-N51
2 ml External Threaded Polypropylene Cryogenic Vial, Self-Standing with Round Bottom	Corning	430659
CoolCell LX, Cell Freezing Container, for 12 x 1 mL or 2 mL Cryogenic Vials	Corning	432002
Greiner Bio-One Cell Strainer, 70 μm, for 1.5 to 15 ml tubes	Fisher Scientific	07-001-107
Greiner Bio-One Cell Strainer, 40 μm, for 1.5 to 15 ml tubes	Fisher Scientific	07-001-106
Superfrost Plus Microscope slides	Fisher Scientific	1255015
Micro Cover Glasses, 22x22 mm, No. 1.5	Fisher Scientific	12541016
Super HT PAP Hydrophobic Slide Marker	Research Products International	195506
5 mL Round Bottom Polystyrene Flow Cytometry Tube, with Cell Strainer Snap Cap	Corning	352235
Software
GraphPad Prism, version 10.4.2	https://www.graphpad.com/
Fiji/ImageJ, version 1.54p	https://imagej.net/software/fiji/
FlowJo, version 10.10.0	https://www.flowjo.com/
Microscope Setups
BZ-X810  - Image type: Brightfield  - Objectives: Plan Apo 2x/0.1, Plan Fluor 20x/0.4  - Temperature: Room temperature	Keyence
A1R Confocal Microscope  - Image type: Laser Scanning Confocal Fluorescence  - Objectives: Plan Apo 10x/0.45, Plan Fluor 40x/1.3 Oil  - Lasers: 405, 488, 561, 640 nm  - Filters: 450/50, 525/50, 595/50, 685/70  - Temperature: Room temperature	Nikon

We prepare stock solutions according to the manufacturers’ instructions, and we reconstitute proteins in Dulbecco’s phosphate-buffered saline (DPBS) with 0.1% (w/v) bovine serum albumin (BSA) as a stabilizing agent. We store small molecules at −20°C and proteins at −80°C and use within two freeze-thaw cycles. We use Matrigel for organoid culture (Corning 356255), which is tested by the manufacturer for the ability to form stable 50-μl domes. Once thawed, we store Matrigel at 4°C and use within 2 weeks.

Media

We previously developed our organoid media based on published formulations, adding our own variations (5, 6, 14, 15). For simplicity, we denote the media as Airway or Alveolar, and we streamlined the Base Media to be the same for both (Table 2). We also add HRG1-β1 to Airway Media (14) and IL-1β to Alveolar Media (6, 14, 15) for the first four days after initial plating (Passage 0) to promote cell survival.

Table 2.

Media recipes.

REAGENT	STOCK CONCENTRATION	FINAL CONCENTRATION	VOLUME FOR 50 ML
Base Media
Advanced DMEM/F12			47.9 ml
GlutaMAX	100X	1X	500 μl
HEPES	1 M	10 mM	500 μl
Primocin	50 mg/ml	100 μg/ml	100 μl
B-27	50X	1X	1 ml
BAL Buffer
Base Media			49.7 ml
Amphotericin B	250 μg/ml	250 ng/ml	50 μl
Gentamicin	10 mg/ml	50 μg/ml	250 μl
ROCK Inhibitor Y27632	10 mM	5 μM	25 μl
Passaging Buffer
Base Media			50 ml
ROCK Inhibitor Y27632	10 mM	5 μM	25 μl
Airway Media
Base Media			49.5 ml
Nicotinamide	1 M	5 mM	250 μl
N-Acetyl-L-cysteine	1.25 M	1.25 mM	50 μl
Recombinant Human R-Spondin-1	500 μg/ml	500 ng/ml	50 μl
Recombinant Human Noggin	100 μg/ml	100 ng/ml	50 μl
Recombinant Human FGF-10	100 μg/ml	100 ng/ml	50 μl
Recombinant Human FGF-7	100 μg/ml	25 ng/ml	12.5 μl
A 83-01	5 mM	500 nM	5 μl
SB 202190	5 mM	500 nM	5 μl
ROCK Inhibitor Y27632	10 mM	5 μM	25 μl
Recombinant Human Heregulinβ-1	50 μg/ml	50 ng/ml	1 μl per ml Airway Media, first 4 days only
Alveolar Media
Base Media			48.9 ml
N-2	100X	1X	500 μl
Insulin-Transferrin-Selenium	100X	1X	500 μl
N-Acetyl-L-cysteine	1.25 M	1.25 mM	50 μl
Heparin	50 μg/μl	5 μg/ml	5 μl
Recombinant Human FGF-10	100 μg/ml	10 ng/ml	5 μl
Recombinant Human EGF	1 μg/μl	50 ng/ml	2.5 μl
CHIR 99021	10 mM	3 μM	15 μl
SB 431542	50 mM	10 μM	10 μl
BIRB 796	10 mM	1 μM	5 μl
ROCK Inhibitor Y27632	10 mM	5 μM	25 μl
Recombinant Human IL-1β	10 μg/ml	10 ng/ml	1 μl per ml Alveolar Media, first 4 days only
Alveolar Differentiation Media (ADM, from ref. 6)
Base Media			45 ml
N-Acetyl-L-cysteine	1.25 M	1.25 mM	50 μl
Heparin	50 μg/μl	5 μg/ml	5 μl
Recombinant Human EGF	0.1 μg/μl	5 ng/ml	2.5 μl
Recombinant Human FGF-10	10 μg/ml	1 ng/ml	5 μl
Human Serum		10% (v/v)	5 ml
L-DCI Differentiation Media (from ref. 23)
IMDM			37.5 ml
Ham’s F12			12.5 ml
GlutaMAX	100X	1X	500 μl
Primocin	50 mg/ml	100 μg/ml	100 μl
B-27	50X	0.5X	500 μl
N-2	100X	0.5X	250 μl
Bovine Serum Albumin	7.5% (w/v)	0.05% (w/v)	333 μl
L-Ascorbic Acid	50 mg/ml	50 μg/ml	50 μl
1-Thioglycerol (MTG)	150 mM	450 μM	150 μl
TRULI (LATS1/2 Inhibitor)	50 mM	10 μM	10 μl
Dexamethasone	1.25 mM	50 nM	2 μl
8-Bromo-cAMP	50 mM	100 μM	100 μl
3-Isobutyl-1-Methylxanthine (IBMX)	100 mM	100 μM	50 μl

We use Base Media within 6 weeks, BAL and Passaging Buffer within 4 weeks, and complete Airway and Alveolar Media within 2 weeks.

Collecting BAL

To develop these protocols and demonstrate their generalizability, we used BAL and lung tissue from a variety of sources (Supplemental Table S1). Most specimens were from lung transplant recipients at Mass General Brigham in Boston or the University of Wisconsin-Madison (UW-Madison). These patients underwent bronchoscopies for routine transplant surveillance or for a decline in clinical status due to post-transplant complications such as rejection or infection. In four of these cases, BAL was performed in the native, non-transplanted lung, providing samples of interstitial lung diseases. Control specimens included BAL from healthy volunteers and lung tissue from deceased donors whose organs could not be used for transplant. When we obtained specimens from other biorepositories, they were deidentified, so the work was classified as non-human subjects research and was exempted from Institutional Review Board (IRB) approval. We now also run biorepositories for lung transplant recipients and general pulmonary patients at UW-Madison, which are approved under IRB #2023–1445 and #2023–1772, respectively. In all cases, specimens were originally obtained with written informed consent from all donors or their legally authorized representative.

We prefer to use fresh, excess BAL from clinical bronchoscopies. The workflow for obtaining such samples will differ by institution, biorepository protocols, and patient population of interest. In our experience, when initiating the study, pathologists at each institution ascertain the volume of BAL required to cover all clinical testing and approve the collection of excess fluid directly from the bronchoscopy procedure location. Potential study participants are identified using the bronchoscopy schedule from the electronic medical record and pre-screened for antibiotic-resistant infections. For example, for biosafety purposes, we exclude patients with a history of (or high clinical suspicion for) methicillin-resistant Staphylococcus aureus (MRSA), mycobacterial, or fungal infection in the past year, or viral infection in the past month. We follow up the results of clinical testing to confirm no resistant organisms were found.

We recommend standardizing the BAL procedure. UW pulmonary clinicians follow guidelines from the International Society of Heart and Lung Transplantation (ISHLT), which call for instillation of two 50-ml aliquots of saline and pooling the return (16). A similar alternative is three 40-ml aliquots. We do not add additional volume for research purposes. We reserve the first 35 ml of BAL return for clinical lab testing, then transfer the excess to a sterile specimen container for our studies. We find that 5 ml works well to establish organoids (minimum 1 ml), but more volume is helpful for additional technical replicates or biobanking. We transport the specimen on ice and use or freeze the cell pellet within 4 hr.

Establishing Organoid Cultures

Here, we provide a step-by-step protocol for establishing organoid cultures from fresh BAL fluid. Our methods are derived primarily from Sachs et al. (5) but are further simplified. For example, we use enzymatic digestions that avoid passing cells through flamed Pasteur pipettes, and we use completely serum-free methods. We were the first group to establish BAL-derived alveolar organoid cultures using the same protocol. Our methods do not require feeder cells or cell sorting, which significantly reduces processing time and stress on the cells; this also means that our epithelial cell monocultures contain only cells from a single patient. Having now replicated this technique on over 60 biological specimens, plus additional controls, we note tips and troubleshooting strategies that have allowed for relatively easy uptake of these methods. Our full pipeline is readily scalable and adaptable; we briefly comment on protocol variations at the end of each section.

All steps are performed in a biosafety cabinet with sterile technique.

• Note: Because media is light-sensitive, we work with the hood light off, relying on indirect lighting from the room.

• Important: For infection control, we use BAL specimens from biorepositories that collect limited clinical data, so that specimens can be screened for viral infections or drug-resistant organisms, as described above. We also include Primocin, an antibacterial and antifungal cocktail, in our media (Table 2). If BAL results reveal any infections not covered by Primocin, the cells are bleached and discarded. We test all cultures for Mycoplasma at least once monthly.

• AApproximate time for establishing organoid cultures: 3–4 hr

  • A1) Pre-warm 12-well tissue culture plates to 37°C.

  • A2) Transfer fresh BAL fluid from the specimen collection container to a 15- or 50-ml conical tube. Spin BAL fluid at 800g for 10 min at 4°C. This speed helps to collect all cells, including small mucus plugs.

    • Optional: Save an aliquot of whole BAL fluid and/or supernatant for biobanking.

  • A3) Aspirate the supernatant and resuspend the cell pellet in a volume of BAL Buffer (Table 2) appropriate for counting, usually 1 ml. Count live and total cells (e.g. using Trypan Blue) and note the percent viability.

  • A4) Transfer up to 1 million live cells to a 15-ml conical. Bring the volume to 5 ml of BAL Buffer. Freeze any excess cells (see “Establishing Organoids: Protocol Variations” below).

    • Example: If 3,200,000 live cells were counted in 1 ml BAL Buffer at step A3, transfer 313 μl (1,000,000 cells) to a 15-ml conical for plating. Freeze the remaining cells.

  • A5) To the BAL cells in 5 ml BAL Buffer, add Liberase TM (1:100, final concentration 50 μg/ml) and DNase I (1:500, final concentration 20 U/ml). This enzymatic digestion breaks up clumps of cells and mucus to generate a single-cell suspension.

  • A6) Gently agitate at 37°C for 20 min. An end-over-end rotating incubator is ideal; a shaking incubator or shaking water bath can be used as well, with the tube positioned at a 45° angle. If cells settle to the bottom during incubation, manually mix by inverting every 10 min to ensure thorough digestion.

  • A7) Place a a sterile 70-μm mesh filter inside a clean 15-ml conical tube. Pre-wet the filter with 1 ml BAL Buffer.

  • A8) Pass digested cells through the 70-μm mesh filter into the clean conical tube. Use an additional 1 ml BAL Buffer to wash the filter and maximize yield.

  • A9) Perform remaining steps on ice or at 4°C.

  • A10) Spin at 200g for 10 min. This speed is optimized to collect live cells post-digestion while eliminating dead cells and debris.

  • A11) The BAL cell pellet should appear white. If the cell pellet is red, add the following red blood cell (RBC) lysis step: resuspend the cell pellet in 1 ml RBC Lysis Buffer (150 mM ammonium chloride, 10 mM potassium bicarbonate, and 0.1 mM EDTA). Incubate at room temperature for no more than 2 min. Add 5 ml BAL Buffer to stop the lysis and spin at 200g for 10 min.

    • Note: BAL cells can also appear gray or brown depending on the patient (e.g. with smoking). We do not use RBC lysis for these samples, since the pigmentation usually comes from macrophages.

  • A12) Resuspend the cell pellet in 5 ml BAL Buffer (wash #1). Do not use less than 5 ml, since it is important to remove any remaining enzymes and debris. Spin at 200g for 10 min.

  • A13) Repeat for another wash (wash #2).

  • A14) Resuspend cell pellet in a volume of BAL Buffer appropriate for counting, usually 1 ml. Count live and total cells and note the percent viability.

    • Note: The post-digestion count often differs from the pre-digestion count and may be higher due to cells freed from mucus clumps. Viability is also improved by washing and should be at least 90%.

    • Optional: To perform cytology staining on BAL cells, set aside 100,000 live cells in 400 μl media with 10% fetal bovine serum (FBS) for Cytospin. This is enough to make two slides of 50,000 cells each. We spin at 300 rpm for 3 min using a Cytospin 4 Centrifuge, then immediately fix the cells in 10% formalin for 10 min.

  • A15) Spin at 200g for 10 min.

  • A16) Remove as much supernatant as possible without disturbing the cell pellet. Cells are now ready for plating. For Passage 0, each well will have 100,000 ± 10,000 live cells. Having started with no more than 1 million cells at step A4 (pre-digestion), and re-counting at step A14 (post-digestion), we would expect to be plating no more than 12 wells. Half the wells will receive Airway Media, while the other half will receive Alveolar Media, so that the same starting population of cells will undergo selection for different cell types.

  • A17) Carefully resuspend ~100,000 live cells per 50 μl Matrigel.

    • Example: If there are 930,000 live cells, resuspend in 450 μl Matrigel, which will be enough to plate 9 wells (e.g. 4 Airway, 5 Alveolar).

    • Important: Since Matrigel solidifies at temperatures above 10°C, keep all materials cold/on ice. Handle tubes containing Matrigel by the tops only to avoid exposing Matrigel to warm fingers. Hold the pipette tips against the cold plastic on the inside of the tube for 10 seconds before touching the Matrigel. The tube should stay in contact with ice at all times.

    • Important: Pipette slowly, since Matrigel is viscous, and pause for 3 sec to allow the full volume of Matrigel to fill the pipette. Mix the cells evenly and avoid bubbles.

  • A18) Plate 50-μl drops in the center of each well of the pre-warmed 12-well plate (Figure 1A). Pipette the Matrigel up and down between plating each drop to ensure an even cell mixture. At the end, distribute any remaining Matrigel into the wells. Carefully remove any bubbles.

    • Tip: Put the plate on top of a plastic microfuge tube rack so that it cools less quickly than on the metal surface of the biosafety cabinet. Pipetting Matrigel onto the pre-warmed plate helps to establish a stable drop.

    • Tip: Reverse pipetting can help to minimize bubbles. Depress the pipette plunger past the first stop to pick up a little extra Matrigel. Then, go up and down to the first stop to dispense 50-μl aliquots.

    • Troubleshooting: If drops consistently run to the side of the well, instead of staying domed in the middle, we recommend checking the following: make sure the surface is level; check that Matrigel is less than 2 weeks old and has a protein concentration of at least 8 mg/ml (Corning 356255, see each lot’s quality certificate); at step A16, remove as much supernatant as possible from the cell pellet to avoid diluting the Matrigel; finally, consider plating smaller drops (e.g. 40 μl at the same cell density).

  • A19) Carefully transport the plate to the incubator. Let drops solidify for 25 min at 37°C.

  • A20) While waiting, warm up 1 ml of Airway or Alveolar Media per well to 37°C (Table 2).

    • Important: Add human HRG1-β1 (1:1,000, final concentration 50 ng/ml) to Airway Media and human IL-1β (1:1,000, final concentration 10 ng/ml) to Alveolar Media.

    • Note: HRG1-β1 and IL-1β are for the initial plating only, when organoids are first established. They are not used for subsequent passages.

  • A21) Once Matrigel has solidified, gently add 1 ml of Airway or Alveolar Media down the side of each well.

  • A22) This is Passage 0 Day 0 (P0 D0). Incubate cells at 37°C, 5% CO2.

  • A23) Change media on Day 4 to remove HRG1-β1 and IL-1β and feed again every 3–4 days thereafter.

  • A24) Count and image airway and alveolar organoids at Passage 0 Day 10 (P0 D10). We quantify organoid forming efficiency (OFE) for each condition as the total number of organoids divided by the number of live cells plated.

Figure 1.

Key steps during organoid culture and expected results. (A) Example of cells plated in a 50-μl drop of Matrigel on a pre-warmed 12-well plate. Once solidified, the Matrigel drop can be covered with Airway or Alveolar Media. (B) To passage organoids, after incubation in TrypLE, organoids are mechanically sheared through a pipette tip pressed perpendicularly against the bottom of the plate (see Supplemental Video S1). (C) Example of BAL-derived airway (top) and alveolar (bottom) organoids at Passage 0 Day 10, Passage 1 Day 14, and Passage 2 Day 14. Organoid morphologies are variable at P0-P1 and become more uniform from P2 onward (BAL #31 is shown; see Supplemental Figure S1 for additional examples). Brightfield images were acquired on a Keyence BZ-X810 microscope. A z-stack was taken covering the entire Matrigel drop (at 2X magnification) or examples of organoids (at 20X magnification), and the Keyence “full focus” feature was used to flatten the stack into one 2D image. Scale bars are 1,000 μm at 2X and 100 μm at 20X. (D) Flow cytometry of airway and alveolar organoid cultures at Passage 2 Day 14–18 shows that the culture media successfully select for either NGFR+ airway basal cells or HT2–280+ alveolar type 2 cells. After gating for live, single cells, >95% of cells are EpCAM+ epithelial cells. The percentage of epithelial cells that are NGFR+ in airway cultures or HT2–280+ in alveolar cultures is given as mean ± SD for 8 and 6 biological replicates, respectively. Updated from ref. 14.

Establishing Organoids: Protocol Variations

• Instead of plating all cells fresh, BAL cells can be cryopreserved (pre-digestion). Spin the BAL specimen at 800g for 10 min at 4°C. Resuspend cell pellets in CryoSFM or Bambanker serum-free freezing medium, up to 1 million live cells per 1 ml, and transfer to cryovials (1 ml per vial). Cool the vials slowly at −80°C for at least 24 hr, then transfer cells to liquid nitrogen. We have found that cell viability is better preserved with CryoSFM or Bambanker, though 90% FBS with 10% DMSO can be used as well.

  • o To thaw cells, for each cryovial containing 1 ml (up to 1 million cells), transfer the cell suspension into 5 ml BAL Buffer. Spin at 800g for 10 min at 4°C, wash with another 5 ml BAL Buffer, spin, and resuspend in 5 ml BAL Buffer for digestion. Proceed with step A5.

• Lung tissue can be used instead of BAL fluid to establish both airway and alveolar organoid cultures, without the need for cell sorting. We start with approximately 1 cm³ of minced lung tissue, chopped using sterile scissors or a razor blade so that the pieces fit through a 5-ml serological pipette tip. Resuspend the minced tissue in 8 ml BAL Buffer in a 15-ml conical tube. Digest with Liberase TM and DNase at the concentrations above for 1 hr at 37°C. Proceed with step A7.

Passaging Organoid Cultures

We perform the first passage at P0 D10 and then every 14 days thereafter. The earlier first passage seems to capture the newly-established organoids in their active growth phase, helping to promote further expansion. In our initial publication, we described passaging at 1:1 to 1:4 ratios without cell counting (14). We have since standardized our protocol to make it more quantitative. Here, we describe our updated protocol, including a longer dissociation time to obtain a single-cell suspension. The cells are then counted, allowing for measurement of OFE and organoid growth rate at each passage.

• Important: Every time cells are exposed to TrypLE is one passage.

• Important: Each BAL specimen has now been grown in two separate conditions: Airway and Alveolar. Both sets of cells are passaged on the same day, using the same protocol, in separate tubes. To avoid redundancy, we will not state this at every protocol step.

• BApproximate time for passaging organoid cultures: 3 hr

  • B1) Pre-warm new 12-well tissue culture plates to 37°C.

  • B2) Aspirate media from the wells. To dissociate the Matrigel and organoids, use 1 ml of TrypLE per 2 wells. Add 1 ml TrypLE to the first well. While ejecting, use the pipette tip to scrape Matrigel gently off the bottom of the plate (Supplemental Video S1). Pipette up and down one time to break up Matrigel. Pipette up all the Matrigel pieces and transfer into a second well, scraping apart the Matrigel in the second well. Pipette up and down one more time.

    • Note: If desired, every well can be dissociated with 1 ml TrypLE. We combine two wells into one to save on the volume of media, so that all steps can be performed in a 15-ml conical tube.

  • B3) Incubate the plate at 37°C for 10 min.

  • B4) While cells are incubating, prepare a 15-ml conical tube with enough Passaging Buffer to dilute TrypLE by 5-fold. Place on ice.

    • Example: For 5 wells of organoids, we use 3 ml TrypLE and add 12 ml Passaging Buffer.

  • B5) Pipette up the organoids and shear them by ejecting while pressing the pipette tip perpendicularly against the bottom of the plate (Figure 1B and Supplemental Video S1). There should be mild resistance while ejecting, and the organoid suspension should squeeze steadily through the space between the pipette tip and the plate. Shear two times. Minimize bubbles.

  • B6) Check digestion under a microscope at low magnification (e.g. 4X objective). Initially, organoids will still be largely intact, but the Matrigel should be dissolved. Incubate for another 10 min.

  • B7) Repeat steps B5-B6 until the organoids are mostly single cells, up to 30–40 min total.

    • Note: Stop if the cells form strands that do not break up with pipetting, as this indicates they are dying and clumping up. Normally, this should not happen.

  • B8) Transfer cells into the prepared tube of Passaging Buffer to inactivate TrypLE. Use buffer to wash the wells to recover as many cells as possible.

  • B9) Filter cells through a 40-μm mesh into a new conical tube. Wash the filter as before to maximize cell recovery.

    • Exception: If there were <10 organoids to begin with, skip filtering so as not to lose precious cells.

  • B10) Spin at 200g for 10 min.

  • B11) Resuspend cell pellet in a volume of Passaging Buffer appropriate for counting. Count live cells.

  • B12) Spin at 200g for 10 min.

  • B13) Remove as much supernatant as possible without disturbing the cell pellet. Cells are now ready for plating. For Passage 1 onward, each well will have 20,000 live cells. We plate up to 6 wells each in Airway and Alveolar conditions for P1, up to 12 wells each for P2, and up to 3 wells each for higher passages. Freeze any excess cells.

    • Note: The cell count includes all types of cells, but under the selective media conditions, immune cells are eliminated while epithelial cells proliferate. The rate of epithelial cell expansion varies by specimen, so it is possible to have fewer wells in P1 than in P0.

    • Note: Our pipeline describes continual organoid culture, and excess cells are banked at each passage. Cells from organoids can be frozen in either CryoSFM or Bambanker with up to 300,000 live cells per 1 ml per cryovial. To restart organoid cultures, thaw cells and transfer into 5 ml Passaging Buffer. Spin at 200g for 10 min at 4°C, wash with another 5 ml Passaging Buffer, spin, and proceed with step B11.

  • B14) As described in steps A17-A21, carefully resuspend 20,000 live cells per 50 μl Matrigel. Plate 50-μl drops in the pre-warmed 12-well plate. Let drops solidify for 25 min at 37°C and add 1 ml per well of pre-warmed Airway or Alveolar Media (without HRG1-β1 and IL-1β).

  • B15) This is Passage 1 Day 0. Incubate cells at 37°C, 5% CO2. Feed with fresh media every 3–4 days.

  • B16) Count and image airway and alveolar organoids at Day 14 of every passage.

By Passage 2 Day 14, both airway and alveolar organoids should have reached maturity and expanded enough for desired applications (Figure 1C). They can be differentiated, harvested for flow cytometry or “-omics” analysis (14), fixed, lysed for DNA/RNA isolation, passaged further, cryopreserved, etc. (Logistically, we tend to passage at Day 14 but distribute analysis steps over Days 14–18. If larger organoids are desired, they can be grown to Day 21–28.) In our experience, airway organoids can be passaged at least 8 times (approximately 4 months), while alveolar organoids can be passaged 3–4 times; approximately 1 in 8 alveolar lines have survived to passage 8 or higher.

Differentiating Airway Organoids at Air-Liquid Interface (ALI)

Airway organoids grown under the conditions above consist of >80% basal cells (KRT5+ TP63+ NGFR+) (ref. 14 and Figure 1D). While the maintenance of stem cells is desirable for long-term organoid expansion, native airways contain mostly differentiated cell types such as secretory (also called club), goblet, and ciliated cells, which are uncommon in the 3D organoid cultures. To mimic native airways more closely, air-liquid interface (ALI) culture is a well-established model system, in which large or small airway epithelial cells are grown on a 2D semi-permeable membrane, with the apical side exposed to air and the basolateral side exposed to media (17, 18). Most methods involve expanding primary or commercial basal cell lines in traditional 2D culture, then transferring cells into a 24-well Transwell insert for ALI. These 2D-to-ALI methods require significant numbers of stem cells to establish the initial cultures and become less reliable at higher passages (18). A few groups have expanded cells as 3D organoids instead, then moved into ALI (5, 12, 13, 19, 20). We and others have shown that these 3D-to-ALI methods allow for the use of starting material that contains only rare stem cells—for example, BAL, tracheal aspirates, and nasal turbinate brush specimens. Here, we expand on this finding by establishing the first time course of differentiation from BAL-derived organoids to determine whether BAL-derived stem cells retain full differentiation potential.

We briefly describe our step-by-step protocol using commercially-available, serum-free PneumaCult-Ex Plus (basal cell expansion), PneumaCult-ALI (large airway differentiation), and PneumaCult-ALI-S (small airway differentiation) media. Media were prepared according to the manufacturer’s instructions and supplemented with Primocin antimicrobials to a final concentration of 100 μg/ml.

• CApproximate time for setting up ALI: 1 hr

  • C1) Dissociate airway organoids (usually Passage 3) to a single-cell suspension as described in the passaging protocol, steps B2-B12.

  • C2) Resuspend the cell pellet in Complete PneumaCult-Ex Plus Media to a concentration of 40,000 cells per 200 μl.

    • Note: We do not recommend using all cells for ALI, since ALI cultures cannot be passaged further. We set aside a portion of cells at P3 for ALI and either continue to expand the rest in 3D organoid culture or cryopreserve the excess cells.

  • C3) Add 200 μl of cell suspension to the apical (top) side of a Transwell insert in a 24-well plate, for as many wells as desired. Add 500 μl of Complete PneumaCult-Ex Plus Media (without cells) to the basolateral (bottom) side. Incubate cells at 37°C, 5% CO2.

    • Note: We use tissue culture-treated, polyethylene terephthalate (PET) Transwell membranes with 0.4-μm pores (Corning 3470), which have not required additional coating. If using untreated inserts or if cells do not adhere consistently to the membrane, many ALI protocols recommend pre-coating the inserts with human type IV collagen or related extracellular matrix material (18, 21).

  • C4) Feed with Complete PneumaCult-Ex Plus Media to both apical and basolateral sides every 2–3 days until cells are 90–100% confluent, usually 3 days.

  • C5) Gently aspirate media from the apical side, being careful not to touch the cells. On the basolateral side only, replace PneumaCult-Ex Plus with 500 μl PneumaCult-ALI or ALI-S, leaving the apical side of cells exposed to air. This is termed “air lift” and is Passage 3 Day 0 of ALI.

    • Tip: Lift the inserts with sterile forceps and tilt them to aspirate while keeping the pipette tip away from the cell layer. Stabilize the inserts so that they do not move when aspirating or pipetting.

  • C6) After air lift, change media on the basolateral side every Monday/Wednesday/Friday (2–3 days).

  • C7) Once per week, add 200 μl of pre-warmed DPBS to the apical side and gently aspirate to wash away excess mucus.

  • C8) Fix ALI at desired time points: Transfer the Transwell insert to a new 24-well plate; do not allow cells to dry out. Wash both apical and basolateral sides with DPBS (200 μl on top, 500 μl on bottom), add 10% formalin to both sides, and incubate at room temperature for 30 min. Wash with DPBS, then fill the well with 2 ml DPBS to store the fixed Transwell inserts. Store at 4°C.

  • C9) Quantify the major airway cell types using immunofluorescence (described below).

ALI: Protocol Variations

• Instead of dissociating airway organoids and plating directly into ALI culture, the single-cell suspension can be sorted using flow cytometry for NGFR+ basal cells, as previously described (14).

• A variation of airway organoid media has been developed that promotes differentiation into ciliated cells in 3D culture, though secretory and goblet cells were lost (13). Thus, 2D ALI remains the method of choice to recapitulate the diversity of differentiated airway cell types, but the 3D method offers an alternative model for ciliary diseases.

Differentiating Alveolar Organoids in 3D Culture

Similarly, we aimed to determine whether BAL-derived alveolar organoids, which contain >80% alveolar type 2 cells (SFTPC+ HT2–280+) (Figure 1D), could be differentiated into alveolar type 1 (AT1) cells. Traditionally, alveolar progenitor cells have been grown in 2D culture, on plastic or cover slips, in serum-containing media, which induces differentiation to AT1 cells (22). More recently, it was shown that alveolar organoids could be kept in 3D culture but switched into media containing 10% human serum (named Alveolar Differentiation Media, ADM) to produce AT1 cells in the organoids (6). Another group reported the first serum-free method to differentiate alveolar organoids derived from iPSCs, likewise keeping organoids in 3D culture and switching the media (23). The serum-free media was called L-DCI for its key components, with “L” denoting LATS inhibitor, which induces nuclear activation of YAP/TAZ; this was shown to be sufficient for AT2-to-AT1 differentiation.

To test whether our BAL-derived AT2 organoids could differentiate into AT1 cells, especially using a serum-free method, we grew Passage 3 alveolar organoids for 7 days in our Alveolar Media, then transitioned into ADM or L-DCI for another 7 days. As a control, we treated lung tissue-derived organoids in the same manner. At Passage 3 Day 14, organoids were fixed and stained as described in the next section.

Characterizing Intact Organoids using Immunofluorescence

After experimenting with many different methods of fixing and staining organoids, we have found the following protocols to be optimal for maintaining organoids’ intact structure while removing Matrigel. Antibody dilutions that we have tested are included in Table 1, and the antibody panels used in the figures are given in Table 3.

Table 3.

Antibody panels used with these protocols.

PRIMARY ANTIBODIES	PRIMARY AB DILUTION	SECONDARY ANTIBODIES	SECONDARY AB DILUTION
Airway Differentiation
Chicken Polyclonal IgY anti-Human KRT5 (BioLegend 905901)*	1:500	Donkey anti-Chicken IgY Highly Cross Adsorbed, Alexa Fluor 647 (Thermo A78952)	1:500
Rabbit Monoclonal IgG anti-Human Acetyl-α-Tubulin (Cell Signaling 5335S)	1:200	Donkey anti-Rabbit IgG Highly Cross-Adsorbed, Alexa Fluor Plus 488 (Thermo A32790)	1:500
SCGB1A1 panel: Mouse Monoclonal IgG1k anti-Human SCGB1A1 (Santa Cruz sc-365992) MUC5AC panel: Mouse Monoclonal IgG1k anti-Human MUC5AC (Thermo MA5–12178)	1:200 1:200	Donkey anti-Mouse IgG Highly Cross-Adsorbed, Alexa Fluor Plus 594 (Thermo A32744)	1:500
*As the chicken anti-KRT5 antibody has become unavailable, we tested an alternative panel:
Goat Monoclonal IgG anti-Human KRT5 (Abcam ab321868)	1:500	Donkey anti-Goat IgG Highly Cross-Adsorbed, Alexa Fluor Plus 594 (Thermo Fisher A32758)	1:500
Rabbit Monoclonal IgG anti-Human Acetyl-α-Tubulin (Cell Signaling 5335S)	1:200	Donkey anti-Rabbit IgG Highly Cross-Adsorbed, Alexa Fluor Plus 488 (Thermo A32790)	1:500
SCGB1A1 panel: Mouse Monoclonal IgG1k anti-Human SCGB1A1 (Santa Cruz sc-365992) MUC5AC panel: Mouse Monoclonal IgG1k anti-Human MUC5AC (Thermo MA5–12178)	1:200 1:200	Donkey anti-Mouse IgG Highly Cross-Adsorbed, Alexa Fluor Plus 647 (Thermo A32787)	1:500
Alveolar Differentiation
Rabbit Polyclonal IgG anti-Human pro-SPC (Seven Hills WRAB-9337)	1:500	Donkey anti-Rabbit IgG Highly Cross-Adsorbed, Alexa Fluor Plus 488 (Thermo A32790)	1:500
Goat IgG anti-Human AGER (R&D Systems AF1145)	1:200	Donkey anti-Goat IgG Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor Plus 594 (Thermo A32758)	1:500

• Important: For steps involving transfer of large organoids, use a 1,000-μl pipette with about 3 mm cut off to widen the tip. Cut tips using a sterile blade in a sterile Petri dish.

• DApproximate time for fixing: 2–3 hr

  • D1) For each line of organoids to be fixed, coat all sides of a 5- or 15-ml tube with 1 ml 1% BSA in DPBS, which helps prevent cells from sticking to the plastic. Aspirate dry.

  • D2) Aspirate culture media (save if desired). Wash away residual media by adding 1 ml of room temperature DPBS to each well, without disturbing the Matrigel drop. Aspirate DPBS, revealing the intact Matrigel dome.

  • D3) Add 1 ml of cold DPBS + 10 mM EDTA per well. While ejecting, use the pipette tip to scrape Matrigel gently off the bottom of the plate. Pipette up and down one time to break up Matrigel without disrupting the organoids. Transfer well contents to the BSA-coated tube. Wash well(s) with another 1 ml DPBS-EDTA to recover as many cells as possible.

    • Note: This step is similar to adding TrypLE when passaging organoids at step B2 (Supplemental Video S1). Two wells of the same condition may be combined into one. The difference is that cold DPBS with EDTA will break up Matrigel without dissociating the cells.

  • D4) Gently agitate at 4°C for 1 hr to dissolve Matrigel. An end-over-end rotator can be used, or any platform rocker/nutating mixer with the tube laid on its side. The liquid should flow gently from one end of the tube to the other to ensure thorough dissociation.

  • D5) Let tube sit upright for 10 min so that organoids sink by gravity. Do not spin, since spinning will crush the organoids’ structure. The organoids should form a loose pellet, visible to the naked eye.

  • D6) Carefully aspirate DPBS-EDTA.

  • D7) Add 1 ml 10% formalin without touching the organoids, which will become sticky. Use the flow of fluid out of the pipette to resuspend the organoids. Add another 1 ml formalin (total 2 ml).

  • D8) Let sit for 30 min (maximum 2 hr) at room temperature to fix. Most organoids should sink.

  • D9) Remove formalin to a designated waste container. Add 3 ml DPBS without touching the organoids.

  • D10) Let sit for 10 min, aspirate, add another 3 ml DPBS and store at 4°C until ready for staining.

• EApproximate time for immunofluorescence day 1: 3 hr

  • E1) Start with fixed organoids in DPBS.

    • Important: Allow organoids to sink by gravity; do not centrifuge. Some loss is inevitable, so start with extra organoids if possible.

  • E2) Carefully aspirate old DPBS.

  • E3) Resuspend organoids in 500 μl Permeabilization buffer (0.2% Triton X-100 in DPBS) and transfer to a labeled 48-well plate. Use extra buffer to collect as many organoids as possible. Let the organoids settle to the bottom of the 48-well plate for at least 2 min, then aspirate any excess volume, leaving approximately 700 μl of Permeabilization buffer per well.

  • E4) Gently shake at room temperature for 15 minutes on a platform rocker or nutating mixer.

  • E5) While waiting, prepare Blocking buffer (5% normal donkey serum + 0.1% Triton X-100 in DPBS). Calculate how much will be needed: 500 μl/well for blocking, 200 μl/well for primary antibody incubation, and 200 μl/well for secondary antibody incubation. Make enough for 1–2 extra wells to allow for pipetting error.

  • E6) Aspirate Permeabilization buffer. Always leave a cushion of 150–200 μl to avoid aspirating the organoids (e.g. if starting with a total of 700 μl, remove no more than 550 μl).

    • Tip: To aspirate, use a 200-μl pipette. Remove bubbles first. Then, place the 200-μl tip just below the meniscus of the solution and follow it down while aspirating.

    • Tip: Pipette 200 μl of buffer into an empty well as a reference for how much volume to leave on top of the organoids.

    • Tip: Look inside the pipette tip to make sure no organoids were accidentally picked up. If they were, eject, let the organoids settle, and try again.

  • E7) Add 500 μl Blocking buffer per well. Eject the buffer down the side of each well, without touching the well contents. This helps mix the organoids without having them stick to the pipette tip.

  • E8) Gently shake at room temperature for 1 hour.

  • E9) Make a solution containing all primary antibodies combined in Blocking buffer. Remember to make enough for 1–2 extra wells. Negative controls will get Blocking buffer only, or an isotype control if available.

    • Note: The 150–200 μl cushion will result in the primary antibody solution being diluted by up to 2-fold. Consider making the stock antibody solution at 2X the desired concentration, so that the final antibody incubation happens at the correct dilution. For example, make 1:250 stock to target 1:500 for the incubation.

  • E10) Aspirate Blocking buffer. Add 200 μl of primary antibody solution to each well (not including negative controls). Add 200 μl of Blocking buffer to negative control wells.

  • E11) Gently shake at 4°C overnight.

• FApproximate time for immunofluorescence day 2: 5–6 hr

  • F1) Aspirate primary antibody solution, keeping the 150–200 μl cushion over the organoids at all times.

  • F2) Add 500 μl Wash buffer (0.1% Triton X-100 in DPBS) down the side of each well, again using the flow of liquid to mix the organoids without touching them.

  • F3) Gently shake at room temperature for 5 minutes. Aspirate Wash buffer.

  • F4) Repeat steps F2-F3 for a total of 5 washes.

    • Note: Given the volume left over each time, each wash is really a dilution, so it is important to complete 5 washes.

  • F5) Make a solution containing all secondary antibodies combined in Blocking buffer. Remember to make enough for 1–2 extra wells. This time, include negative controls.

    • Important: Protect fluorescent antibodies from light by wrapping tubes and plates in aluminum foil.

  • F6) Add 200 μl of secondary antibody solution to each well (including negative controls).

  • F7) Gently shake at room temperature for 2 hr.

  • F8) Aspirate secondary antibody solution.

  • F9) Repeat steps F2-F3 for a total of 5 washes. If desired, add DAPI to the wash buffer for the fourth wash (final concentration 1 μg/ml).

  • F10) Using 500 μl of DPBS, transfer the organoids from the 48-well plate to individual microfuge tubes. Use another 500 μl of DPBS to wash out the wells, collecting as many organoids as possible. Let the organoids sink to the bottom.

  • F11) While waiting for organoids to settle, cut about 3 mm off of 200-μl pipettes to widen the tips. Label glass microscope slides and draw a circle (about 2 cm in diameter) on each slide with a hydrophobic marker.

  • F12) Aspirate as much supernatant as possible without losing organoids.

  • F13) Using the cut-off tips, add 30 μl of Prolong Gold Mountant to each sample. Gently resuspend the organoids in mountant and transfer into the circle on the pre-labeled slide. Use the pipette tip to move the majority of organoids toward the middle of the circle. Aspirate bubbles if necessary.

  • F14) Place a coverslip on each slide, pressing very gently to remove any bubbles.

  • F15) Store slides in the dark at room temperature overnight to let the mountant dry. Image on a confocal microscope or other instrument as desired, using the same microscope settings and image analysis settings for all comparable samples. We prefer to image 1–2 days after slides have dried; slides can be stored in the dark at room temperature or 4°C for up to 2 months, but the main limitation is fluorescence fading over time.

Immunofluorescence: Protocol Variations

• We used a modified version of this protocol to stain ALI membranes for the differentiation time course. In this case, we stained for KRT5+ basal cells, SCGB1A1+ secretory cells, MUC5AC+ goblet cells, acetylated tubulin-positive (AcTub+) ciliated cells, plus F-actin to outline all cells. Due to the number of markers used, we separated the staining into two antibody panels (Table 3). After fixing the ALI cultures at each time point, we cut each membrane in half, then removed the halves from the Transwell using a razor blade. Each half was stained in a 48-well plate using the methods above. The main difference was that, since the cells were still attached to the membrane, all the volume could be aspirated with each wash, as long as the cells were not allowed to dry out. Instead of DAPI at step F9, phalloidin was added to stain F-actin, according to the manufacturer’s instructions; this made cell counting easier. Membranes were mounted onto slides with the cell side facing the cover slip (plastic membrane facing the slide). After imaging, cells were manually counted in an area of 25,000–100,000 μm².

• If only one antibody panel is desired, ALI membranes can be stained directly in the Transwell, adding reagents to the apical and basolateral sides, then cutting out and mounting the membrane as a last step. ALI membranes can also be paraffin-embedded and stained by standard methods.

• This protocol can be further adapted to stain Cytospin slides or formalin-fixed, paraffin-embedded (FFPE) organoids. FFPE samples require dewaxing and antigen retrieval. We have found that similar dilutions of the antibodies described work well across these various applications. Many other groups have described additional variations that we have not tried, including embedding organoids in optical cutting temperature (OCT) compound for sectioning (15), or growing and staining organoids in Matrigel on a chamber slide or cavity slide (7, 23). Organoids (most commonly, sorted epithelial-mesenchymal cell co-cultures) can also be grown in 50% Matrigel inside a Transwell insert and stained directly using prolonged incubation times and agitation (24, 25). Each method has advantages and disadvantages; for example, sectioning can help to visualize very large, hollow organoids or intraluminal markers such as HT2–280 in AT2 cells, while intact organoids allow for scanning through the entire 3D structure.

RESULTS AND DISCUSSION

We optimized and standardized a pipeline for establishing, expanding, and differentiating airway and alveolar organoids from human BAL fluid and showed that all major lung epithelial cell types can be derived from a single, accessible source of primary cells. We have intentionally designed all steps to be accessible to users with standard laboratory skills and resources. All media components are chemically-defined and serum-free, which improves reproducibility by reducing lot-to-lot variation and avoiding exposure to serum components that can change cellular identity.

We established organoids from a wide variety of BAL samples with a high rate of success. Starting with 2 to 21 ml of BAL fluid, we obtained cell counts ranging from 160,000 to 5,800,000 live cells (Supplemental Table S1). After plating fresh BAL specimens, we counted and imaged organoids at Passage 0 Day 10 (Figure 1C). Organoids have variable morphologies initially but have a well-defined border that distinguishes them from the background of immune cells; at P0 D10, this is often better visualized by eye under a light microscope. At this point, 42 of 45 specimens (93%) yielded organoids in Airway Media, and 40 of 43 (93%) yielded organoids in Alveolar Media (updated from ref. 14). The organoid forming efficiency (OFE, number of organoids per cells plated) ranged from 0.0012% to 1.1% in Airway Media and from 0.0014% to 0.57% in Alveolar Media. The reason for the wide variation in OFE is not immediately clear and is under ongoing investigation. For example, OFE does not correlate with total cell count in BAL (not shown), since BAL contains mostly immune cells. OFE may relate to BAL technique and/or the number of epithelial cells, or specifically basal cells and AT2 cells that are thought to give rise to airway and alveolar organoids, respectively (3, 4). Three additional examples of BAL-derived airway and alveolar organoids are shown in Supplemental Figure S1 to illustrate heterogeneity between specimens.

In our prior work, the steps through P0 D10 were standardized, and cultures were then passaged by partially digesting organoids and splitting into new wells at ratios of 1:1–1:4 depending on organoid confluency. To quantify organoids across multiple passages, we now digest organoids to single cells and plate 20,000 cells per well at each passage. We processed 26 airway organoid cultures using this updated pipeline and determined that 22 of 26 (85%) could be expanded to at least Passage 2; for alveolar organoid cultures, 21 of 24 (88%) grew to at least Passage 2. Immune cells drop out of the cultures after passaging, since the media selects for epithelial cell outgrowth, and the 200g spins help to eliminate dead cells and debris. As a result, the cell counts—and therefore number of wells plated or banked at each passage—were highly variable between specimens, but the OFE increased as cultures became predominantly epithelial (Figure 1C and Supplemental Figure S1). For airway organoids at Passage 1, 22,500–890,000 live cells yielded an OFE of 0.005–18.0%; at Passage 2, 7,500–1,300,000 live cells had an OFE of 0.46–32.0%. By Passage 2 Day 14–18, airway organoids are spherical, a mixture of solid and cystic depending on organoid density and days in culture. For alveolar organoids at Passage 1, 12,500–507,000 live cells yielded an OFE of 0.087–4.58%; at Passage 2, 2,500–990,000 live cells had an OFE of 0.035–17.0%. Alveolar organoids are usually solid with irregular borders; thin-walled spheres can also be observed.

We used flow cytometry on mature Passage 2 organoids to validate the cellular composition. Details of our flow cytometry methods are given in ref. 14 and summarized in Supplemental Figure S2. All cultures yielded >95% EpCAM+ epithelial cells, confirming effective expansion from BAL fluid, which starts with approximately 1% epithelial cells of any kind (26, 27). The epithelial population consisted, on average, of 94 ± 3% NGFR+ basal cells in airway organoids and 90 ± 8% HT2–280+ AT2 cells in alveolar organoids (Figure 1D). This was a substantial improvement from our previous work, in which the yield of AT2 cells ranged from 3–93% of epithelial cells. We speculate that the improvement of alveolar cell selection stems from standardization of BAL procedures. Though we cannot test this directly, the volume of BAL is known to influence the composition of cells in the return fluid. The first 50–60 ml aliquot generally samples small airway cells, while the second aliquot reaches alveolar spaces (28). Since almost all patients at our center receive 100–120 ml of saline instillation, and the return is pooled, we are likely capturing more alveolar cells, as well as improving normalization across specimens.

The prevalence of airway and alveolar stem cells in the organoid cultures is highly desirable for continued expansion and studying cell proliferation phenotypes. However, to model the cellular diversity of the native airway or alveolar spaces, methods that promote differentiation into specific cell types are needed. We prioritized serum-free approaches. For differentiation of airway basal cells, 2D air-liquid interface culture remains the method of choice, using commercially-available serum-free media such as PneumaCult. In contrast, differentiation of AT2 cells traditionally requires serum in 2D methods, so we explored newer serum-free methods that maintain the 3D organoid format.

To demonstrate differentiation of airway basal cells, we transitioned BAL-derived airway organoids into 2D ALI cultures and generated a time course of differentiation into secretory, goblet, and ciliated cells (Figure 2). At Day 1 after air lift, the membranes were covered with flattened KRT5+ basal cells. We observed no difference when starting from unsorted airway organoid cells or NGFR+ sorted basal cells (not shown). This suggests that any non-basal cell types in the organoids do not adhere or grow in the initial ALI conditions, so airway organoids can be transitioned directly to ALI without sorting. Interestingly, a few bundles of AcTub were visible, corresponding to active mitosis and midbody formation, which is required for primary ciliary development (Supplemental Figure S3) (29, 30). By Day 7, SCGB1A1+ secretory cells, MUC5AC+ goblet cells, and AcTub+ ciliated cells were all emerging; some of these cells were double positive for KRT5+, likely representing intermediate states. Secretory and goblet cells peaked in number at approximately Day 14 though overall made up <10% of the total cell population. Ciliated cells predominated by Day 21, rising to approximately 40–60% of the total, and were slightly more prevalent in PneumaCult-ALI (large airway) compared to ALI-S (small airway) media, as expected. One sample was grown to Day 28, which showed further ciliary maturation, with an increase in AcTub+ cells and decrease in KRT5+/AcTub+ double positive cells. The time course of differentiation of BAL-derived organoids was similar to that for tissue-derived organoid controls (Supplemental Figure S3).

Figure 2.

Time course of differentiation of BAL-derived airway organoids after being transitioned to air-liquid interface. Passage 3 cells grown in (A) PneumaCult-ALI (large airway) media or (B) PneumaCult-ALI-S (small airway) media were fixed at Day 1, 7, 14, and 21, then stained using two antibody panels for KRT5 (white), SCGB1A1 (magenta), MUC5AC (yellow), acetylated tubulin (AcTub, green), and F-actin (blue). Images were acquired on a Nikon A1R confocal microscope at 40X magnification with 0.5-μm slices. Each row of representative images was acquired and analyzed using the same settings and shown as maximum intensity projections. Scale bars are 50 μm. Cells were manually quantified in Fiji/ImageJ and graphed as the mean ± SD from 5 biological replicates. One specimen was extended to Day 28 for reference.

The cell proportions we observe at ALI Day 21–28 align well with the Human Lung Cell Atlas of primary airway tissue, suggesting that the BAL organoid-to-ALI approach can mimic native airways (31, 32). However, our cell counts were limited by having to use two separate staining panels to accommodate five markers. We also did not stain for rare cell types such as respiratory airway secretory cells, ionocytes, etc., but estimate they would total <10% of all cells. Despite these limitations, we show overall that BAL-derived airway organoids can be differentiated as expected at ALI, and we identify the specific time points at which secretory, goblet, and ciliated cells are enriched.

To demonstrate differentiation of AT2 cells, we maintained BAL-derived alveolar organoids in 3D culture while modifying the media conditions. We tested two recently-published media formulations: serum-containing ADM (6) and serum-free L-DCI (23). As is typical for human alveolar organoids (4, 6, 7, 33), we rarely observed AT1 cells in maintenance Alveolar Media, and the organoid morphologies ranged from thin-walled hollow spheres to more solid, irregular, and lobulated in appearance (Figure 3A). After 7 days in ADM or L-DCI differentiation media, organoids exhibited only subtle morphologic changes: some developed flatter patches or small budding cells, while others appeared unchanged. As expected, there was heterogeneity within and between specimens. Differences were more apparent after organoids were fixed and stained for the AT2 cell marker surfactant protein C (SFTPC) and the AT1 cell marker advanced glycosylation end-product specific receptor (AGER) (Figure 3B and Supplemental Figure S4). In L-DCI media, 76 of 98 BAL-derived organoids (78%) contained a majority of AGER+ cells, with a corresponding decrease in SFTPC signal. ADM had a less pronounced and more variable effect, with only 21 of 54 organoids (39%) showing majority AGER expression. The extent of differentiation for BAL-derived organoids was the same or better than for tissue-derived controls. It is possible that differentiation in ADM would have been more extensive if allowed to continue for 10–14 days, but we stopped after 7 days to prevent cultures from becoming overgrown. Overall, we demonstrate proof of principle that BAL-derived organoids can respond to established, serum-free differentiation cues, such as air-liquid interface culture for airway cells and YAP/TAZ activation for alveolar cells.

Figure 3.

Differentiation of BAL-derived alveolar organoids. Passage 3 organoids were grown for 7 days in Alveolar Media, then switched into serum-containing (ADM, ref. 6) or serum-free (L-DCI, ref. 23) differentiation media for another 7 days, or kept in Alveolar Media. Tissue-derived organoids were used as controls. (A) Brightfield images of organoids at Day 14 were acquired on a Keyence BZ-X810 microscope at 20X magnification. The Keyence “full focus” feature was used to show examples of organoids for three BAL-derived cell lines and one tissue-derived control. (B) Organoids were freed from Matrigel, fixed intact, and stained for the AT2 cell marker SFTPC (green), the AT1 cell marker AGER (magenta), and DAPI (blue). Images were acquired on a Nikon A1R confocal microscope at 40X magnification with 2-μm slices, using the same settings throughout. Shown are maximum intensity projections of representative organoids, with the three individual channels underneath the larger merged image. Scale bars are 100 μm.

In summary, by integrating organoid technology with BAL, a common clinical procedure, we highlight the opportunity to obtain primary lung cells for research with minimal risk to patients. BAL can access small airway and alveolar cells from a variety of healthy and diseased states. Our results demonstrate that the rare epithelial cells shed into BAL fluid can form organoids that behave in all the ways expected of organoids grown from lung tissue or iPSCs, from passaging to differentiation. Importantly, this protocol is readily adapted to other starting material; we have used cryopreserved human BAL, murine BAL, and human and murine lung tissue. We have not tried alternative airway specimens such as bronchial brushings, tracheal aspirates, nasopharyngeal swabs, or transbronchial biopsies. We speculate that these might be feasible as well, since the number of airway epithelial cells in such specimens is expected to be higher than in BAL fluid (27, 34). Indeed, recent advances in the organoid field highlight the model’s versatility and its remarkable ability to generate organoids from unexpected sources of primary cells (35).

Our methods have several limitations. We prefer to collect fresh BAL and establish organoids the same day, which can be logistically challenging. Storing or shipping samples on ice, or using frozen BAL cell pellets, yields lower OFE (not shown) and is likely to require additional optimization. We also have limited experience with BAL containing infectious organisms or malignant cells, for example. Overall, we observe significant variability of organoid numbers and growth patterns, especially at P0-P1; the reasons for this remain unclear. In this article, we aimed to be inclusive of different types of BAL specimens in order to demonstrate the generalizability of the method. We did not have sufficient clinical details or sample sizes to determine whether organoid or differentiation phenotypes differ based on patients’ lung disease. In future applications of this pipeline, it will be critical to establish inclusion and exclusion criteria and to validate whether BAL-derived organoids recapitulate disease states.

Finally, while our models are so far limited to epithelial monocultures, they provide a platform to add further layers of complexity. Organoid and ALI cultures can both be exposed to chemicals, chemokines, or aerosols for mechanistic manipulations or drug discovery (12, 36–38). Matrix material can be altered to examine the effects of extracellular proteins and mechanical tension (39, 40). Epithelial cells can be co-cultured with immune, mesenchymal, or endothelial cells and infected with pathogens (41–44). Increasingly, cells from organoids are being altered genetically or transplanted (10, 45–48). We envision that the combination of accessible starting material and straightforward protocols may facilitate these advanced applications.

Supplementary Material

SUPPLEMENTAL MATERIAL

Supplemental Table S1, Figures S1-S4, and Video S1 are available at: https://doi.org/10.5061/dryad.6m905qgf3

DATA AVAILABILITY

The data supporting the findings of this study are available within the article and its supplemental materials. Raw data are available from the corresponding author, MYL, upon request.