Generation of human antral and fundic gastric organoids from pluripotent stem cells

Abstract

The human stomach contains two primary domains: the corpus, which contains the fundic epithelium, and the antrum. Each of these domains has distinct cell types and functions, and therefore each presents with unique disease pathologies. Here, we detail two protocols to differentiate human pluripotent stem cells (hPSCs) into human gastric organoids (hGOs) that recapitulate both domains. Both protocols begin with the differentiation of hPSCs into definitive endoderm (DE) using activin A, followed by the generation of free-floating 3D posterior foregut spheroids using FGF4, Wnt pathway agonist CHIR99021 (CHIR), BMP pathway antagonist Noggin, and retinoic acid. Embedding spheroids in Matrigel and continuing 3D growth in epidermal growth factor (EGF)-containing medium for 4 weeks results in antral hGOs (hAGOs). To obtain fundic hGOs (hFGOs), spheroids are additionally treated with CHIR and FGF10. Induced differentiation of acid-secreting parietal cells in hFGOs requires temporal treatment of BMP4 and the MEK inhibitor PD0325901 for 48 h on protocol day 30. In total, it takes ~34 d to generate hGOs from hPSCs. To date, this is the only approach that generates functional human differentiated gastric cells de novo from hPSCs.

Introduction

The human stomach contains two main types of epithelium. The corpus contains a fundic epithelium with oxyntic glands that are enriched for acid-secreting parietal cells and protease-producing chief cells. The antrum abuts the pyloric/duodenal junction and can be identified by a unique population of gastrin-producing cells.

In this protocol, we describe how to generate hGOs that contain either antral or fundic epithelium to model components of the stomach. The hGOs are generated from hPSCs, which include both human embryonic stem cells (hESCs) and human induced pluripotent stem cells (iPSCs). We have previously used this method to generate antral and fundic organoids to study the molecular pathogenesis triggered by Helicobacter pylori infection and to modulate acid secretion by physiologic and pharmacologic factors1,2.

Development of the protocol

Gastric organoids can be generated from primary human stomach tissue3,4,5,6; however, this requires access to surgical or cadaveric samples. Moreover, it has been challenging to generate gastric organoids that contain functional parietal cells from adult tissues. Mouse tissue has been used to overcome the hurdle of lack of access to human surgical samples; however, the mouse stomach is different from the human both architecturally and in pathologic responses7. For example, the largest domain in the mouse stomach is the forestomach, which is composed of non-glandular squamous epithelium similar to that of the esophagus. Humans do not have this anatomical structure.

We decided instead to bypass the need for surgical samples by starting with established hPSCs and differentiating those into hGOs. To develop an unlimited source of human gastric tissue with functional differentiated cell types, we developed two protocols to differentiate hPSCs into hGOs that contain either antral or fundic epithelium1,2. To do this, we manipulated signaling pathways that direct normal stages of embryonic stomach development in a stepwise manner. We started by differentiating hPSCs into DE by adding activin A. Activin is a TGFβ-family member that stimulates Nodal signaling, a highly conserved pathway that is required for endoderm formation across vertebrate species8. Endoderm was then patterned into anterior and foregut endoderm by inhibiting BMP signaling through the use of Noggin. BMP signaling is known to promote a posterior endoderm fate in vertebrates ranging from frog to human9,10,11,12. Foregut spheroids were directed into posterior foregut by activation of the retinoic acid (RA) signaling pathway, which was shown to be required for development of the posterior foregut13,14,15. At this stage of stomach development, the presumptive antral and fundic domains can be distinguished by the expression of regional markers. However, the signaling pathways that patterned the developing stomach into antral versus fundic domains had not been identified. Using mouse genetics and human foregut tissues derived from hPSCs cultures, we determined that canonical Wnt signaling was required for promoting the development of fundic epithelium, but not for that of antrum2. This enabled us to generate human antral gastric organoids (hAGOs), which contained the expected diversity of endocrine cells, particularly the antrum-specific gastrin-producing G cells, as well as mucous neck and pit cells1. Wnt activation promoted development of human fundic gastric organoids (hFGOs); however, this was not sufficient to promote parietal cell development. We found it difficult to differentiate gastric cells into functional parietal cells and therefore used hFGOs to screen multiple signaling pathways to see what promoted parietal cell development. We determined that concurrent MEK inhibition and activation of the BMP signaling pathway promoted differentiation of parietal cells that secreted acid upon histamine stimulation2.

Overview of the protocol

The entire protocol is schematically outlined in Fig. 1a. It takes ~34 d to perform a stepwise differentiation of hPSCs into both hAGOs and hFGOs with their respective differentiated and functional gastric cells. The procedures for hAGOs and hFGOs are identical for the first 6 d (protocol days 0–5) and only differ once the generated posterior foregut spheroids are embedded in Matrigel. The first stage requires 3 d of activin A treatment with BMP4 supplementation on the first day of treatment to differentiate hPSCs into a DE monolayer co-expressing the markers SOX17 and FOXA2. The second stage requires another 3 d and results in the formation of free-floating 3D posterior foregut spheroids expressing the markers HNF1β and SOX2. This spheroid generation requires the combined activity of FGF4, the GSK3 antagonist and Wnt activator CHIR99021 (CHIR), BMP antagonist Noggin, and RA. Free-floating spheroids are embedded in Matrigel (protocol day 6) and grown in 3D culture for the remainder of the protocol.

Fig. 1: Overview of the protocol.
figure1

a, Schematic representation of the protocol from pre-passage hPSCs to end hAGOs and hFGOs. b, Day −1 low-magnification image of hPSC colonies (~75–85% confluent) in a six-well dish, showing the optimal appearance of cells ready to be passaged as single cells into a 24-well dish for the start of the differentiation. Scale bar, 1 mm. The image brightness was increased to improve visibility. Refer to Table 2 for the specific details of the different culture media that are required at each stage of the protocol.

After Matrigel embedding, growth of hAGOs requires 3 additional days of treatment with EGF, Noggin, and RA, followed by continuous treatment with EGF for the remaining 3+ weeks. To grow hFGOs, after Matrigel embedding, the spheroids are treated with EGF, CHIR, Noggin, and RA for 3 d. It is important to note the patterning roles of signaling pathways that are manipulated during this process. Inhibition of BMP signaling with Noggin initiates the development of foregut spheroids, and RA further patterns the foregut into posterior foregut as distinguished by expression of HNF1B and GATA4. After the first 3 d in 3D culture, posterior foregut spheroids develop into hAGOs with no further patterning instructions. However, activation of canonical Wnt/β-catenin signaling is required to generate hGOs with a fundic epithelium, which is achieved using the small molecule CHIR. At protocol day 9, fundic spheroids are cultured in medium containing CHIR and EGF, which is additionally supplemented with FGF10 from protocol day 20 for the remaining 14 d. The EGF concentration is reduced from protocol day 30 onward to aid in the development of endocrine cell populations in both hAGOs and hFGOs. Finally, on protocol day 30, hFGOs are treated for 48 h with supplementary BMP4 and the MEK inhibitor PD0325901 to stimulate the production of parietal cells. By protocol day 34, both hAGOs and hFGOs have a broad range of functional cell types (see ‘Anticipated results’ for the expected characteristics of these hGOs), and we typically collect organoids for analysis around this time point. Continued culture of hFGOs beyond day 36 is not optimal, possibly due to harmful effects of acid and proteases produced by hFGOs.

Applications of the protocol

Very little is known about embryonic development of the stomach, relative to the lower gastrointestinal tract. Similarly, the pathways that control formation of all the gastric lineages from stem cells and progenitors are also largely unknown. hGOs are well suited to the study of normal and congenitally abnormal stomach development, as well as regeneration from stem and progenitor cells. Through its digestive and endocrine functions, the stomach controls the breakdown of food and eating behavior. As a result of the low pH and abundance of digestive enzymes, the stomach is capable of neutralizing most ingested pathogens. However, failure to do so can result in diseases ranging in severity from gastritis, to ulcers, to gastric cancer. Owing in part to the architectural differences in the stomach between mice and humans, many gastric disease progressions are not well recapitulated in mice. hGOs can be grown in large numbers, are genetically tractable, and can be pharmacologically manipulated. Therefore, hGOs can be used to study many of the above mentioned physiologic and pathologic processes. hGOs produce the hormone gastrin; they secrete acid and pepsin. Through pharmacologic or genetic manipulation, hGOs can be used to study these cellular processes. For example, one can increase or eliminate gastrin-producing cells by the genetic manipulation of the transcription factor NEUROG3 (refs 1,2), and then study the resulting effects on parietal cells. Because hAGOs produce gastrin as well as other hormones, it should be possible to use organoids to screen for drugs that modulate gastrin hormone secretion and to study the downstream effects on gastric cell types.

The bacteria H. pylori is the main cause of gastric disease and has been effectively studied using hPSC-derived hGOs1. With combinations of cell-lineage tracing and live imaging, hGOs could be used to identify cells that are directly targeted by H. pylori and the resulting molecular changes that occur. For detailed methods describing how to study H. pylori using mouse organoids, see Engevik et al.16. In addition, genetic influences on gastric cancer initiation and progression in humans could be studied in real time using iPSC lines from patients with hereditary forms of cancer or through CRISPR-mediated editing of genes implicated in gastric cancer. Given that hFGOs have functional parietal cells that are drug responsive2, they can be used to screen for new classes of proton pump inhibitors and to pre-screen other drugs for adverse gastric effects before phase 1 clinical trials17.

Comparison with other methods

There have been several reports describing hGO models derived from patient tissues (primary gastric organoids)3,4,5,6. These models contain gastric epithelium and in the short term contain multiple cell types but lack an associated mesenchymal component. Long-term culture of human primary gastric organoids is possible using growth medium that favors the growth of stem and progenitor cells. However, robust differentiation into parietal cells in these cultures is lacking. Moreover, establishing these cultures requires access to human surgical samples, which are not commonly available to many laboratories. In addition, the quality of surgical samples is widely variable and is heavily dependent on timely access to tissue.

The method described here differs from the above methods in three fundamental ways. First, this method uses hPSCs (either hESCs or human iPSCs), which are established, quality-controlled cell lines that are available to all laboratories. Second, this method reproducibly results in the de novo formation of differentiated lineages, including acid-secreting parietal cells, protease-producing chief cells, and all gastric endocrine cell lineages. Gastric organoids from adult gastric glands can be passaged in culture, but re-differentiation of passaged organoids into functional parietal cells has proven challenging3,4,5,6. Third, unlike human primary gastric organoids derived from adult tissues, which are purely epithelial structures, hGOs have a population of undifferentiated mesenchymal cells surrounding the glandular epithelium. Finally, the most noteworthy advantages of iPSC-derived organoids are that iPSC lines are immortal and highly quality controlled, can be infinitely expanded, are pluripotent, can be used to generate any organoid type, and can easily be generated from any patient, starting with blood, fibroblasts, or even urine. In addition, the use of established iPSC lines does not require any institutional approval. These features make iPSC-derived hGOs an ideal model for facilitating personalized medicine. However, a drawback to the use of iPSC-derived hGOs is the requirement for substantial expertise in handing and differentiating hPSCs. Moreover, hPSC-derived in vitro–differentiated cell lineages such as pancreatic, liver, and intestinal cells are not as mature and functional as their adult organ counterparts18,19,20. Although hGOs have functional cell types that secrete acid, proteases and mucus, at a transcriptional level they appear to be less mature than human gastric tissue1,2.

Despite these shortcomings, hGOs derived from hPSCs or adult gastric samples have been effectively used to study gastric cell function, stem cells and the response of cells to injury and H. pylori infection in a human-specific manner1,2,3,4,5,6,21,22. By developing biobanks of genetically cataloged, quality-controlled primary or iPSC-derived organoids, institutions are capitalizing on organoid technologies to study human congenital defects, cancer-causing mutations, host–pathogen interactions, and probiotics, as well as to prescreen drugs for efficacy and toxicity, among other aims17.

Application of protocol to other PSC lines

To date, we have generated hGOs from two hESC lines (H1 and H9) and >18 different iPSC lines generated by Cincinnati Children’s Hospital Medical Center’s Pluripotent Stem Cell Facility (PSCF). When all parameters of the protocol were met, 90% of iPSC lines successfully generated hGOs. Performing the quality-control steps highlighted in Boxes 1 and 2 is recommended when first generating hGOs with new hPSCs. If an hPSC line is capable of forming an optimal DE monolayer, described in detail in the protocol, then its ability to spontaneously generate spheroids that develop into hGOs is near 100%. hPSCs that are proven to generate hGOs should routinely generate organoids, provided that the hPSCs are properly maintained in a high-quality state largely free of differentiation. We have applied this protocol only to hPSCs, so we do not know if it could be used to differentiate PSCs from other species.

Materials

Biological materials

  • hESCs or iPSCs. We have successfully used H1 and H9 human ESCs (WiCell International Stem Cell Bank, WiCell Research Institute) and iPSCs (generated by the PSCF, Cincinnati Children’s Hospital Medical Center). The human embryonic stem cell lines used here are cataloged at the NIH Human Embryonic Stem Cell Registry. H1 (WA01) has registration number 0043 and H9 (WA09) has registration number 0062

    Caution

    Use of human tissues must adhere to all relevant ethical guidelines. When appropriate, informed consent must be obtained from donors or their legal guardians.

    Caution

    When handling cells, take the necessary precautions. Commonly, use proper personal protective equipment (PPE) at biosafety level 2 (BSL-2) for tissue culture work, and perform the work within a class II biosafety cabinet.

    Critical

    All published iPSC lines generated by the PSCF are available with the consent of the principal investigator (PI) and a material transfer agreement. The PSCF has several lines that were generated exclusively by the facility and do not require PI consent. Requests for these iPSC lines can be sent to the PSCF at https://www.cincinnatichildrens.org/research/cores/pluripotent-stem-cell-facility. The PSCF also provides training courses that cover basic handling of hPSCs, iPSC generation, and organoid production. Although the PSCF was one of the first facilities of its kind, there are now quite a few commercial and academic facilities that provide services related to iPSC technologies.

Reagents

Caution

When handling reagents, take the necessary precautions. Commonly, use proper PPE at BSL-2 for tissue culture work and BSL-1 for non–tissue culture work. Refer to specific reagent material safety data sheets for additional information if unfamiliar with the reagents.

Growth medium and supplements

Critical

We have not tested medium or supplements from other vendors.

  • mTeSR1 medium (StemCell Technologies, cat. no. 85850)

  • Advanced DMEM-F12 (Invitrogen, cat. no. 12634-010)

  • RPMI medium 1640 (Invitrogen, cat. no. 11875-093)

  • hESC-qualified Matrigel (Corning, cat. no. 354277)

  • Matrigel (basement membrane matrix; Corning, cat. no. 354234)

  • MEM non-essential amino acid solution (NEAA; 100×; Invitrogen, cat. no. 11140-050)

  • Defined FBS (dFBS; Hyclone, cat. no. SH30070.02)

  • l-Glutamine (100×; Life Technologies, cat. no. 25030-081)

  • Penicillin–streptomycin (100×; Life Technologies, cat. no. 15140-122)

  • B27 supplement (50×, without vitamin A; Thermo Fisher Scientific, cat. no. 12587-010)

  • N2 supplement (100×; Invitrogen, cat. no. 17502-048)

  • HEPES buffer (1 M; Life Technologies, cat. no. 15630-080)

Enzymes and growth factors

Critical

We have not tested enzymes and growth factors from other vendors.

  • Accutase (Thermo Fisher Scientific, cat. no. A11105-01)

  • Dispase (Thermo Fisher Scientific, cat. no. 17105-041)

  • Activin A (Cell Guidance Systems, cat. no. GFH6)

  • Bone morphogenetic protein 4 (BMP4; R&D Systems, cat. no. 314-BP-050)

  • Fibroblast growth factor 4 (FGF4; R&D Systems, cat. no. 235-F4)

  • Fibroblast growth factor 10 (FGF10; R&D Systems, cat. no. 345-FG)

  • CHIR99021 (CHIR; Stemgent, cat. no. 04-0004-10)

  • Noggin (R&D Systems, cat. no. 6057-NG)

  • Epidermal growth factor (EGF; R&D Systems, cat. no. 236-EG-01M)

  • Retinoic acid (RA; Sigma-Aldrich, cat. no. R2625)

  • PD0325901 (StemCell Technologies, cat. no. 72182)

  • Y-27632 dihydrochloride (rho-associated protein kinase (ROCK) inhibitor; Tocris; cat. no. 1254)

Immunostaining reagents

  • Normal donkey serum (NDS; Jackson ImmunoResearch Laboratories, cat. no. 017-000-121)

  • Alexa Fluor donkey anti-goat 488 (Thermo Fisher Scientific, cat. no. A11055)

  • Alexa Fluor donkey anti-mouse 568 (Thermo Fisher Scientific, cat. no. A10037)

  • Alexa Fluor donkey anti-rabbit 647 (Thermo Fisher Scientific, cat. no. A31573)

  • Alexa Fluor donkey anti-rabbit 488 (Thermo Fisher Scientific, cat. no. A21206)

  • Alexa Fluor donkey anti-goat 568 (Thermo Fisher Scientific, cat. no. A11057)

  • Alexa Fluor donkey anti-mouse 647 (Jackson ImmunoResearch Laboratories, cat. no. 715-605-150)

  • Goat anti-SOX17 (R&D Systems, cat. no. AF1924)

  • Mouse anti-FOXA2 (Abnova, cat. no. H00003170-M01, clone 7E6)

  • Goat anti-SOX2 (Santa Cruz, cat. no. sc-17320)

  • Mouse anti-CDX2 (BioGenex, cat. no. MU392A-UC, clone CDX2-88)

  • Rabbit anti-β-catenin (CTNNB1; Santa Cruz, cat. no. sc-7199)

  • Rabbit anti-GATA4 (Santa Cruz, cat. no. sc-9053)

  • Goat anti-PDX1 (Abcam, cat. no. ab47383)

  • Mouse anti-E-cadherin (CDH1; BD Transduction Laboratories, cat. no. 610182, clone 36/E-cadherin)

  • DAPI (Sigma-Aldrich, cat. no. D9542-10MG)

  • Fluoromount-G fluorescent mounting medium (Southern Biotech, cat. no. 0100-01)

Other reagents and chemicals

  • Sterile PBS (Sigma, cat. no. P5368)

  • Triton X-100 (Sigma, cat. no. T8787)

  • BSA (Sigma-Aldrich, cat. no. A8806-5G)

  • Dimethyl sulfoxide (DMSO; Sigma-Aldrich, cat. no. D8418-250ML)

  • Paraformaldehyde (PFA; Thermo Fisher Scientific, cat. no. T353-500)

  • Optimal cutting temperature (OCT) compound (Tissue-Tek; VWR, cat. no. 25608-930)

  • Sucrose (Sigma, cat. no. S7903-1KG)

Equipment

  • Portable Pipet-Aid XP pipette controller (Drummond, cat. no. 4-000-101)

  • Pipetman single-channel pipettes (2, 20, 200, and 1,000 μl; Gilson, cat. nos. F144801, cat. no. F123615, cat. no. F123601, and cat. no. F123602)

  • Horizontal clean bench (Labconco, cat. no. 3600004)

  • Labgard class II type A2 biological safety cabinet (NuAire, cat. no. NU-425-400)

  • −80 °C Freezer (Thermo Scientific, cat. no. UXF700086D)

  • Cryostat (Leica, cat. no. CM1850)

  • Forma Steri-Cycle i160 CO2 incubator (Thermo Scientific, cat. no. 51030533)

  • Stereomicroscope (Leica, cat. no. S8APO)

  • Inverted microscope (Nikon, TMS model)

  • Pulled-glass Pasteur pipette (9 inches, pulled in lab; Fisherbrand, cat. no. 13-678-20D)

  • Nunclon delta surface tissue culture dish (six well; Nunc, cat. no. 140675)

  • Nunclon delta surface tissue culture dish (24 well; Nunc, cat. no. 142475)

  • Nunclon delta surface tissue culture dish (four well; Nunc, cat. no. 176740)

    Critical

    Nunclon delta surface tissue culture dishes are used to plate the Matrigel spheroid suspension in a drop (bead), which allows for 3D expansion of the tissue. It is unknown if other tissue culture surfaces allow the Matrigel bead to form or if the Matrigel will undesirably spread into a thin layer across the surface. If using tissue culture dishes from other manufacturers, it is important to test the formation of 3D Matrigel beads beforehand.

  • Serological pipettes (5 and 10 ml; Falcon, cat. nos. 357543 and 357551)

  • Sterilized filter pipette tips (10, 20, and 200 μl (Denville, cat. nos. P1096-FR, P1121, and P1122, respectively) and 1,250 μl (VWR, cat. no. 10017-092))

  • Millipore 0.22-μm conical sterilization tubes (Millipore, cat. no. SCGP00525)

  • Conical tubes (15 and 50 ml; Falcon; BD Biosciences, cat. nos. 352097 and 352070)

  • 1.7-ml Microcentrifuge tubes (Denville, cat. no. C2170)

  • 5-ml Microtubes (Argos, cat. no. T2076A)

  • Cell culture coverslips (sterile plastic; Thermanox; Thermo Scientific, cat. no. 174969)

  • Microscope slides (Superfrost Plus; Fisher Scientific, cat. no. 12-550-15)

  • Cover glass (Cardinal Health, cat. no. M6045-10)

  • Disposable cell lifter (Fisher Scientific, cat. no. 08-100-240)

  • Disposable base molds (15 × 15 × 5 mm; Fisherbrand, cat. no. 22-363-553)

  • Petri dish (sterile, 60 × 15 mm; Fisher Scientific, cat. no. FB0875713A)

  • Hydrophobic pen (ImmeEdge pen; Vector Laboratories, cat. no. H-4000)

  • Nikon A1 single-photon confocal microscope with NIS-Elements Advanced Research imaging software

Reagent setup

Human pluripotent stem cell lines

hPSCs include both hESCs and iPSCs. Culture hPSCs in feeder-free conditions on hESC-qualified Matrigel as previously described1,2,9,23,24,25. Briefly, hPSCs should be maintained in feeder-free conditions on hESC-qualified Matrigel-coated Nunclon delta surface six-well dishes in 2 ml of mTeSR1 medium per well at 37 °C in a 5% CO2 tissue culture incubator. mTeSR1 medium should be replaced daily. hPSCs will grow as colonies and should be dispase-passaged as cell clumps, commonly at a well/well ratio of 1:6, into new six-well dishes every 4 d for confluence maintenance. During routine maintenance culture, any noticeable differentiation should be removed by ‘picking’ or scraping the differentiation from the well floor with a pulled-glass Pasteur pipette and then aspirating during the daily mTeSR1 medium change. For additional information regarding proper hPSC maintenance, refer to ‘Stem Cell Protocols’ at http://www.wicell.org/.

Dispase

Resuspend dispase in advanced DMEM-F12 to obtain a final concentration of 1 mg ml−1. Next, filter-sterilize, using a filter-sterilization tube. Once filtered, divide into aliquots and store at −20 °C for up to 6 months.

Accutase

Thaw bottle on ice or overnight at 4 °C. Divide Accutase into aliquots and store at −20 °C for up to 6 months.

Aliquots of hESC-qualified Matrigel for hPSCs culture and basement membrane Matrigel for hGO culture (embedding Matrigel)

Thaw a 5-ml bottle of hESC-qualified Matrigel or a 10-ml bottle of basement membrane Matrigel on ice or overnight at 4 °C. Divide the Matrigel into aliquots in 1.7-ml microcentrifuge tubes and quickly transfer filled microcentrifuge tubes to ice; then store at −80 °C for up to 6 months. We recommend aliquot volumes that are sufficient to cover one or multiple 6- or 24-well dishes for hESC-qualified Matrigel. The actual volumes are lot-dependent and determined by the manufacturer. We recommend 500-μl aliquots for basement membrane Matrigel.

Critical

Matrigel will rapidly begin to solidify near room temperature (20–25 °C), so it is vital to work quickly and keep the Matrigel cold throughout the process of making aliquots.

hESC-qualified Matrigel-coated dishes for hPSC culture

Coat 4-, 6-, or 24-well Nunclon delta surface dishes with hESC-qualified Matrigel as detailed by the manufacturer. Briefly, thaw the Matrigel aliquot on ice and resuspend in cold advanced DMEM-F12 medium. Note that the final dilution required is lot-dependent and determined by the manufacturer. Transfer sufficient cold medium plus Matrigel to each well of the tissue culture dish so that the entire surface is covered. Usually, 1 ml per well is required for a 6-well dish and 500 μl per well is required for 4- or 24-well dishes. It is vital to incubate all Matrigel-coated dishes at room temperature for at least 1 h before passaging hPSCs. Matrigel-coated dishes can be stored at 4 °C for up to 2 weeks if the dishes are wrapped with Parafilm to prevent evaporation and the formation of dry spots.

Critical

Matrigel will rapidly begin to solidify and fall out of solution near room temperature, so it is vital to work quickly and add Matrigel to cold medium before coating dishes.

Growth factor and small-molecule reconstitution

Reconstitution information for all growth factors and small molecules can be found in Table 1. After reconstitution, growth factors can be stored at 4 °C for up to 1–2 weeks, whereas small molecules typically can be stored at 4 °C for longer than 2 weeks. Refer to manufacturers’ guidance for storage and stability instructions if any questions arise. For long-term storage (up to 6 months), make aliquots and store at −20 or −80 °C. After thawing an aliquot of growth factor or small molecules, store at 4 °C. Avoid freeze–thaw cycles.

Table 1 Growth factor and small-molecule reconstitution table

hGO protocol culture media

The components required for the culture medium used at different time points during hAGO and hFGO generation are listed in Table 2. Further instructions relating to their production are listed in the following sections. Until day 6, the various media used are the same for all hGOs. From day 6 onward, different media are required for hAGO and hFGO differentiation.

Table 2 hAGO and hFGO protocols culture media

Day 0 medium

Medium for first day of DE differentiation. Supplement RPMI 1640 with 1× MEM NEAA. Then add growth factors as found in Table 2. Medium complete with growth factors is best if made fresh each day, but it can be stored at 4 °C for a few days. Medium without growth factors can be stored at 4 °C for 2 weeks.

Critical

No serum is used in day 0 medium.

Day 1 medium

Medium for second day of DE differentiation. Supplement RPMI 1640 with 1× MEM NEAA and 0.2% (vol/vol) dFBS. Then add growth factors as found in Table 2. Medium complete with growth factors is best if made fresh each day, but it can be stored at 4 °C for a few days. Medium without growth factors can be stored at 4 °C for 2 weeks.

Day 2–5 media

Media for third day of DE differentiation and spheroid generation. Supplement RPMI 1640 with 1× MEM NEAA and 2.0% (vol/vol) dFBS. Then add growth factors and small molecules as found in Table 2; note that these differ over time. Medium complete with growth factors and small molecules is best if made fresh each day, but it can be stored at 4 °C for a few days. Medium without growth factors can be stored at 4 °C for 2 weeks.

Gut medium

Medium for 3D spheroid and organoid growth. Supplement advanced DMEM-F12 with 1× N2, 1× B27 (without vitamin A), 2 mM l-glutamine, 100 units per ml (1×) penicillin–streptomycin, and 15 mM HEPES. Then add growth factors and small molecules as found in Table 2; note that these differ over time. Gut medium complete with growth factors and small molecules is best if made fresh, but it can be stored at 4 °C for a few days. Gut medium without growth factors can be stored at 4 °C for 2 weeks and is largely based on the medium described in Sato et al.26. Gut medium complete with all growth factors and/or small molecules listed in Table 2 per the specific protocol day can be refered to as gastric (antral or fundic) growth medium.

PBST

PBST consists of PBS and 0.5% (vol/vol) Triton X-100, which serves as the cell permeabilizing agent. PBST is commonly prepared ahead of time and can be stored at room temperature for 6 months.

Immunostaining blocking buffer

Add 5% (vol/vol) normal donkey serum (NDS) to PBST, which includes the cell permeabilizing agent (e.g., add 50 μl of NDS to 950 μl of PBST). NDS can be stored at 4 °C for up to 2 weeks. However, make fresh blocking buffer before each use and keep on ice.

Procedure

Single-cell passage of hPSCs (day −1)

Timing 2 h

  1. 1

    1 h or more before single-cell passaging of hPSCs, prepare a 24-well Nunclon delta tissue culture dish with hESC-qualified Matrigel as described in ‘hESC-qualified Matrigel-coated dishes for hPSC culture’ in Reagent setup.

    Critical step

    Optionally, plastic cell culture coverslips can be aseptically placed in the 24-well dish before Matrigel coating. This allows cells to be removed for subsequent immunostaining and analysis with confocal microscopy at a desired time point. The use of pre-treated plastic coverslips instead of glass is critical for hPSCs attachment and growth during the protocol.

    Critical step

    All tissue culture steps in the protocol that follow should take place in a sterile environment and be performed with aseptic technique to prevent culture contamination.

  2. 2

    Take a six-well dish of largely differentiation-free hPSC colonies that are ~75–85% confluent (Fig. 1b). Remove any visible spontaneous differentiation using a pulled-glass Pasteur pipette to pick or scrape the differentiated cells from the well floor. Then remove the mTeSR1 by aspiration from the wells to be passaged.

    Critical step

    Optimal confluency of the starting hPSCs and absence of substantial differentiation in the six-well dish is vital for obtaining optimal results in the following steps. Generally, three wells at 75–85% confluence from a six-well dish is sufficient to passage into a full 24-well dish with minimal extra hPSCs.

  3. 3

    Wash wells with 2 ml of advanced DMEM-F12 that was pre-warmed in a 37 ºC bath, then aspirate the medium.

  4. 4

    Add 1 ml of warm Accutase to each well and place the dish back into the tissue culture incubator at 37 °C and 5% CO2 until all colonies have dissociated and lifted off the well surface as small cellular chunks or single cells (this usually takes 7–10 min).

    Critical step

    Accutase incubation time should be optimized for different lots, as the enzymatic activity varies across lots.

  5. 5

    Add 5 ml of warm advanced DMEM-F12 to each well to sufficiently dilute the Accutase.

  6. 6

    Gently triturate each treated well two to three times before transferring the 6 ml of hPSCs and medium per well to a 50-ml conical tube. Repeat the process with the remaining wells until all cells are collected into the same 50-ml conical tube. If only one to two wells are treated, the total volume will be <15 ml, so a 15-ml conical tube can be used instead.

  7. 7

    Centrifuge the cells at 300g for 3 min at room temperature.

  8. 8

    Aspirate the medium from the conical tube.

    Critical step

    Make sure you do not aspirate or disturb the pellet at the bottom of the conical tube.

  9. 9

    Add warm mTeSR1 to the conical tube.

    Critical step

    The amount mTeSR1 added correlates to the number of wells of a 24-well dish that will be passaged. Typically, 3 wells at 75–85% confluence from a 6-well dish can be optimally passaged at a well/well ratio of 1:9, meaning that 1 well of the 6-well dish will make 9 wells of the 24-well dish, with each well of the 24-well plate receiving 500 μl. Further, three wells of single cells being passaged at a ratio of 1:9 require that 13.5 ml of mTeSR1 be added to the conical tube.

    Critical step

    If counting cells, typically 300,000 (±50,000) cells per well, with each well receiving 500 μl, is optimal. Add mTeSR1 accordingly to obtain a cell count per well that is within that range. The ideal cell count will probably vary slightly across different hPSC lines based on survivability as single cells and proliferation rate, among other factors. Optionally, a range of cells per well can be passaged to ensure that optimal confluence will be achieved on day 0.

  10. 10

    Supplement the mTeSR1 and single-cell solution in the conical tube with 10 μM ROCK inhibitor Y-27632. Gently swirl the conical tube manually and apply gentle trituration to dissociate the cell pellet and mix the solution adequately.

  11. 11

    Aspirate the Matrigel from the prepared Matrigel-coated 24-well dish (from Step 1) and then transfer one 500-μl single-cell aliquot to each well.

    Critical step

    A 5-ml serological pipette or a 1,000-μl pipette is recommend for passaging. While passaging, occasionally swirl the conical tube or gently triturate the cells to ensure an approximately equal number of single cells per well.

  12. 12

    Place the 24-well dish in a tissue culture incubator and ensure that the single cells are evenly distributed in the wells by gently moving the 24-well dish back and forth twice, then side to side twice, before gently closing the incubator door and incubating overnight, typically close to 24 h.

Confirmation that hPSCs have achieved the required confluency (day 0)

Timing 5 min

  1. 13

    The following day, observe the 24-well dish with passaged hPSCs under a stereomicroscope and/or inverted microscope. It is common to observe that many of the cells passaged the day before are floating in the medium. The monolayer of hPSCs should be 75–90% confluent. There is some variation in this key starting confluence among different hPSC lines.

    Fig. 2: Identifying appropriate confluence of target hPSCs to ensure optimal differentiation to DE and robust generation of posterior foregut spheroids.
    figure2

    a, Day 0 images of hPSCs 1 d after single-cell passaging at different ratios, resulting in too low, target, and too high confluence. Low-magnification images of hPSCs passaged at a too sparse confluence (top left), at target confluence (top center), and at a too dense confluence (top right). High-magnification images of hPSCs passaged at too sparse confluence (bottom left), at target confluence (bottom center), and at too dense confluence (bottom right). b, Day 3 images of a target confluence DE monolayer at the end of 3-d activin A treatment at low (top) and high magnification (center and bottom). c, Day 6 low-magnification images of anticipated spheroid generation with a poor (left), robust (center), and moderate (right) generation. a,b, Culture medium was changed to remove cellular debris pre-image capture. Scale bars, 1 mm (a,b (top and center),c); 250 μm (b, bottom). The brightness of the top images in a was adjusted to better match that of the other panels.

    Critical step

    Do not continue with the Procedure if the confluency of the hPSCs is not 75–90%. A spider web configuration with linkages two to three cells thick or less is a common indication of low confluence, and a monolayer with only a few small holes is a common indication of high confluence (Fig. 2a). Target confluence on day 0 is critical to developing a target day 3 DE monolayer (Fig. 2b), which is vital to producing robust spheroid generation on day 6 (Fig. 2c).

    troubleshooting

Differentiation of hPSCs into human DE (days 0–2)

Timing 3 d

  1. 14

    Aspirate the mTeSR1 and any floating debris from each well. Add 500 μl of warm day 0 medium to each well and place the 24-well dish back in the incubator.

    Critical step

    Washing the hPSCs with warm RPMI 1640 medium before starting DE differentiation is not necessary.

  2. 15

    After 24 h, aspirate the day 0 medium and replace it with 500 μl of warm day 1 medium.

    Critical step

    It is common to observe substantial amounts of floating cell debris during the first and second days of the DE differentiation process. Minor loss of confluence after the first day of DE differentiation is normal. However, heavy loss of cells, resulting in isolated clusters or islands, is not desired. Cultures should continue to have a web-like structure of connected cells after the first day of DE differentiation.

  3. 16

    After 24 h, aspirate the day 1 medium and replace it with 500 μl of warm day 2 medium (for the third day of DE differentiation).

    Critical step

    The monolayer should reach 95–100% confluence after the second day of DE differentiation.

  4. 17

    (Optional) 24 h later, on day 3, if the hPSCs have been passaged on plastic coverslips (Step 1), carefully remove the plastic coverslip and associated DE monolayer with forceps and undertake immunostaining as a quality-control step. SOX17 and FOXA2 co-staining can be used to determine the efficiency of the DE differentiation as described in Box 1. An example of optimal results is shown in Fig. 3.

    Fig. 3: Analysis of differentiation efficiency of day 3 DE and day 6 posterior foregut monolayer cultures by immunofluorescence.
    figure3

    a, Optimal differentiation of hPSCs into DE displays ~85–90% SOX17 (green) and FOXA2 (red) double-positive expression by day 3. b, Optimal day 6 posterior foregut monolayer displays near-ubiquitous SOX2 (green) and minimal-to-absent CDX2 (red) expression. DAPI (blue), CTTNB1 (white). Scale bars, 100 μm. Images were taken on a Nikon A1 single-photon confocal microscope and processed with NIS-Elements Advanced Research imaging software provided by the CCHMC Confocal Imaging Core. The brightness of the images in b was adjusted after image capture to improve visibility.

    Critical step

    When observed under a stereomicroscope, a confluent, flat, opaque, and near-uniform layer of DE is expected (Fig. 2b, top). When observed under an inverted microscope, a crowded and compact layer of DE is expected (Fig. 2b, center and bottom).

    Critical step

    If the DE monolayer is too sparse (not uniformly opaque) or too dense (not flat) on day 3, then the resulting spheroid generation will probably be poor or moderate instead of robust (Fig. 2c). Do not proceed with the Procedure if the DE monolayer is not confluent and opaque under a stereomicroscope, as it will not typically generate budding spheroids.

    troubleshooting

Spontaneous budding of posterior foregut spheroids from human DE (days 3–5)

Timing 3 d

  1. 18

    On day 3, aspirate the day 2 medium and replace it with 500 μl of warm day 3 medium (spheroid generation medium).

  2. 19

    After 24 h, aspirate the day 3 medium and replace it with 500 μl of warm day 4 medium (spheroid generation medium; this is the same as the medium used on day 3).

    Fig. 4: Morphological changes during protocol days 4–20 of culture associated with posterior foregut spheroid generation, and outgrowth of hAGOs and hFGOs in 3D Matrigel suspension.
    figure4

    a (Top), Day 4 low- (left) and high-magnification (right) images of the spheroid-generating monolayer as the 3D budding morphology develops across the well. (Center) Day 5 low- (left) and high-magnification (right) images of the spheroid-generating monolayer as the 3D morphology forms numerous attached budding spheroids with a few free-floating spheroids. (Bottom) Day 6 low- (left) and high-magnification (right) images of the spheroid-generating monolayer with numerous free-floating spheroids ready for collection and a monolayer largely devoid of 3D budding morphology. b, Day 6 high-magnification images of individual spheroids free-floating in medium. Spheroids from the same generation vary in shape and size, but all are embedded in Matrigel for future growth. c, Low-magnification images of antral and fundic gastric spheroids in 3D Matrigel suspension developing into hGOs on day 9 (top), day 13 (second row), and day 20 (third row). Insets show magnified images of individual spheroids from the suspension. (Bottom row) Day 20 high-magnification images of hAGOs and hFGOs with substantial unintended neural growth (green arrows). Such hGOs should be avoided when selecting hGOs to culture further by reducing density. Scale bars, 1 mm (a,c), 250 μm (b). The white boxed insets are 5× digital zooms; the width of the insets is ~833 μm (c). The brightness of the images in the top row of c was adjusted to match that of the other panels.

    Critical step

    On day 4, the emergence of a 3D morphology that precedes spheroid formation should be evident across the surface of the well (Fig. 4a), top).

  3. 20

    After 24 h, aspirate the day 4 medium and replace it with 500 μl of warm day 5 medium (spheroid generation medium), which is photosensitive because of the RA.

    Critical step

    Because the medium is photosensitive, turn the biosafety cabinet light off and wrap the tube of medium and RA in foil to prevent excessive light exposure.

    Critical step

    On day 5, the 3D morphology should have numerous attached budding spheroids and a small number of free-floating spheroids (Fig. 4a, center).

  4. 21

    After 24 h, on day 6, observe the wells. There should be a large number of posterior foregut spheroids that are free-floating in the medium. In addition, the monolayer should appear to be largely flat and devoid of the 3D budding morphology (Fig. 4a), bottom).

    Critical step

    The quality of the posterior foregut spheroid generation and resulting ability to develop into hGOs has been observed to positively correlate with the number of spheroids generated on day 6. Spheroids from a robust generation tend to result in more successful hGO generation than those from a moderate or poor generation (Fig. 2c).

    troubleshooting

  5. 22

    Collect posterior foregut spheroids under a stereomicroscope, using a 200-μl barrier pipette tip, without ripping or tearing the base monolayer; then pool the spheroids collected from one well into a 1.7-ml microcentrifuge tube.

    Critical step

    Posterior foregut spheroids substantially in size and shape (Fig. 4b). Although it is not known if the size of the posterior foregut spheroid dictates successful organoid formation, the majority of embedded spheroids (Fig. 4c, top) go on to form hGOs (Fig. 4c), third row), so all spheroids from a target well are embedded. We have noticed that iPSC-derived posterior foregut spheroids are commonly larger than those derived from hESCs, although this trend does not have any apparent consequence.

    Critical step

    One well of spheroids to one microcentrifuge tube assumes a robust generation (600–800 spheroids per well). If a more moderate generation resulted, combine two to three wells in one microcentrifuge tube.

    Critical step

    We typically carry out spheroid collection and embedding (Steps 22–28) on a horizontal clean bench for easy access, but these steps can also be performed in a biosafety cabinet.

  6. 23

    Optionally, on day 6, if hPSCs were passaged on plastic coverslips (Step 1), after collecting spheroids, carefully remove the plastic coverslip and associated posterior foregut spheroid–generating monolayer with forceps for quality-control immunostaining. SOX2 and CDX2 staining can be utilized to determine monolayer patterning, and therefore spheroid patterning, as described in Box 1. An example of optimal results is shown in Fig. 3b.

Development of spheroids into hGOs (days 6–20)

Timing 1 h to plate spheroids; 14 d to grow tissue

  1. 24

    Thaw, at 4 °C or on ice, the necessary number of aliquots of embedding Matrigel. The time for Matrigel to thaw varies according to the size of the aliquots. We recommend thawing Matrigel aliquots hours before use or overnight at 4 °C.

    Critical step

    Keep the Matrigel aliquots cold, as they will solidify quickly near room temperature.

  2. 25

    Once all posterior foregut spheroids have been collected (Step 22), place the spheroid microcentrifuge tubes upright in a tube rack for ~5 min. The spheroids will quickly settle at the bottom of the tubes via gravity sedimentation, so no centrifuging is required.

  3. 26

    While observing the spheroids under a stereomicroscope, use a 200-μl barrier pipette tip to remove as much excess medium as possible without disturbing the spheroids at the bottom of the microcentrifuge tube.

    Critical step

    Maintaining excessive spare medium will dilute the embedding Matrigel and may cause issues with Matrigel bead formation and/or proper 3D suspension of spheroids.

  4. 27

    Cut the tip off a 1,000-μl barrier pipette tip. Using the cut 1,000-μl tip, collect 350 μl of ice-cold Matrigel, enough to plate 6 × 50 μl of Matrigel beads with working spare, and transfer it to a spheroid-containing microcentrifuge tube. Once the cold Matrigel has been added, quickly stir with the pipette tip and pipette the suspension up and down several times with a cut 200-μl barrier pipette tip to mix thoroughly.

    Critical step

    Work quickly, because the Matrigel begins to solidify as it warms up, usually in under a minute. Concurrently, be mindful to avoid introducing air bubbles into the Matrigel-spheroid suspension.

    Critical step

    This step assumes a robust spheroid generation (Fig. 2c, center) with intent to embed in Matrigel and plate spheroids at a 1:6 well/well ratio, resulting in ~100 spheroids per Matrigel bead. Adjustments will be necessary if spheroid generation and/or desired plating density differ.

  5. 28

    After mixing, collect 50 μl of the suspension, using the cut 200-μl barrier pipette tip, and pipette into the center of a well in the 24-well dish. Keep the pipette tip slightly elevated from the well floor and ensure that the Matrigel bead does not touch the well wall while pipetting. Quickly, plate six beads in total from the suspension-filled microcentrifuge tube and then repeat Steps 27 and 28 as needed.

    Critical step

    After plating the spheroids, avoid significant disturbances to the dish for ~5 min to avoid undesirable spreading of the Matrigel bead across the well while the Matrigel begins to solidify.

    Critical step

    Nunclon delta surface tissue culture dishes (24- or 4-well typically) are important for the formation of an ideal 3D Matrigel bead. Matrigel may spread out on dishes from other manufacturers, thus limiting 3D spheroid and organoid expansion. If using tissue culture dishes from other manufacturers, it is important to test the formation of 3D Matrigel beads beforehand.

    Critical step

    Place the 24-well dish cover on top of the dish when not actively pipetting spheroids into wells. This will prevent the Matrigel bead from drying out or desiccating before application of culture medium.

    troubleshooting

  6. 29

    When each well of the 24-well dish contains a Matrigel bead, place the dish into a tissue culture incubator for 10–15 min to allow the Matrigel to completely solidify before adding culture medium.

    Critical step

    When the Matrigel is completely solidified, it should be possible to invert or mildly shake the dish without the Matrigel bead moving or disruption of its 3D structure.

  7. 30

    Add 500 μl of gastric (antral or fundic) growth medium to each well as desired, ensuring that each Matrigel bead is covered, before placing the dish back into the tissue culture incubator.

  8. 31

    Replace gastric (antral or fundic) growth medium every 4 d, when the phenol red in the medium turns yellow, or when the feeding protocol calls for a change of growth factors (see Table 2 for days on which this is required), whichever comes first, until day 20 or when a reduction in hGO density is required. A reduction of density is necessary when the hGOs begin to grow into each other.

    Critical step

    Between days 9 and 20, antral- and fundic-patterned gastric spheroids develop at similar efficiency and rate (Fig. 4c), first three rows).

    troubleshooting

  9. 32

    Optionally, take aliquots of day 20 hAGOs and hFGOs for immunostaining to determine the efficiency of antral and fundic specification. The immunostaining procedure is described in Box 2, and examples of optimal results are shown in Fig. 5.

    Fig. 5: Analysis of differentiation efficiency of day 20 hAGOs and hFGOs by immunofluorescence.
    figure5

    a, On day 20, hAGOs and hFGOs both display near-ubiquitous epithelial GATA4 (green) expression. hAGOs display near-ubiquitous epithelial expression of PDX1 (red), whereas hFGOs displays much lower or even absent expression of PDX1. b, On day 20, hAGOs and hFGOs both display near-ubiquitous epithelial SOX2 (green) and minimal-to-absent CDX2 (red) expression. DAPI (blue), CDH1 and CTNNB1 (white). Scale bars, 100 μm. Images were taken on a Nikon A1 single photon confocal microscope and processed with NIS-Elements Advanced Research imaging software provided by the CCHMC Confocal Imaging Core.

    Critical step

    Collection of hGOs on day 20 for immunostaining is not a rigid time point, and immunostaining can be performed a few days earlier or later, depending on the density of hGOs in the Matrigel and convenience.

    Critical step

    By day 20, although the hFGOs are slightly smaller and have a thicker epithelium than hAGOs, both types of hGOs should possess clear lumens and have a small population of mesenchymal cells surrounding the epithelium (Fig. 4c, bottom two rows) as described1,2.

    Critical step

    Both types of hGOs, especially hFGO cultures, can have a low level of unintended neural growth on the outside the epithelium (Fig. 4c, green arrows). The neural growth of extending neurons resembles an expanding fuzzy cloud that often appears as dense netting between closely growing hGOs. High-quality hPSCs are essential to the success of any differentiation protocol. Under optimal conditions, the level of unintended neural growth has a substantial effect on only a small number of organoids. If desired, the levels of neural growth can be measured using the neural markers NESTIN and TuJ-1 (neuron specific beta3-tublin) via qPCR. In addition, immunostaining for the expression of TuJ-1+ and/or SOX2+ E-cadherin cells can identify neural growth in hGO cultures.

Reduction of density and re-embedding of hGOs for continued growth

Timing 30 min–1 h to reduce density per new 24-well plate; 14 d to grow tissue

  1. 33

    On or around day 20, reduce the density of hGOs in each well by re-embedding the hGOs in fresh Matrigel. This step can be done by two different methods, the cutting and the swirling methods. The cutting method (option A), allows for more control when reducing density and is best when working with relatively few hGOs (two wells or fewer than 200 total hGOs) but is labor and time intensive. The swirling method (option B), yields comparable results, is relatively faster, and works well with greater numbers of hGOs. First, regardless of method selected, thaw 500 μl of embedding Matrigel on ice for each approximately eight wells of anticipated re-embedded hGOs. We typically undertake hGO density reduction on a horizontal clean bench for easy access, but it can also be done in a biosafety cabinet.

    Critical step

    The recommended re-plating density is five to ten hGOs per new Matrigel bead. However, alternative densities can be used if they better suit your experimental requirements. hFGOs tend to develop better with fewer hFGOs per Matrigel bead.

    Critical step

    While reducing density, select against any unintended neural growth by not re-embedding neural-heavy tissue (Fig. 4c, green arrows).

    1. (A)

      Cutting method of reducing density

      Timing 1 h per new 24-well plate

      1. (i)

        Prepare a 60 × 15-mm Petri dish with 4 ml of advanced DMEM-F12 and place it in a tissue culture incubator to warm the medium. This dish will be used for cutting the day 20 hGOs Matrigel beads into smaller sections.

      2. (ii)

        Cut the tip of a 200-μl barrier pipette tip to increase the bore size of the pipette such that hGOs can pass through the tip without sustaining damage, ~3 mm in diameter.

      3. (iii)

        With the cut 200-μl barrier pipette tip, dislodge the Matrigel bead containing day 20 hGOs from the base of the well. This can be easily accomplished by nudging or applying unidirectional pressure to the outside of the Matrigel bead using the outside of the cut 200-μl barrier pipette tip.

      4. (iv)

        With the cut 200-μl barrier pipette tip, collect the Matrigel bead and transfer it to the dish containing warm advanced DMEM-F12. Repeat Step 33A(iii,iv) with as many wells as necessary to collect the desired number of hGOs.

        Critical step

        Use separate dishes, for example, for different lots of hGOs, hAGOs or hFGOs, or those generated from different hPSC lines. Dishes with hGOs not currently in use can be kept temporally in a tissue culture incubator.

      5. (v)

        Under a stereomicroscope, use a scalpel to cut the floating hGOs-embedded Matrigel beads into smaller Matrigel sections that contain five to ten hGOs each.

        Critical step

        Try to avoid cutting hGOs directly but rather cut the surrounding Matrigel to achieve the desired five to ten hGO Matrigel sections. However, most hGOs that are inadvertently cut reform an intact lumen and continue growing.

      6. (vi)

        With a cut 200-μl barrier pipette tip, collect the desired hGOs embedded in Matrigel sections into 1.7-ml microcentrifuge tubes. Repeat this process as needed and remove any excess medium from the tubes. For example, if making a new 24-well dish, then four microcentrifuge tubes filled with six to seven Matrigel sections containing five to ten hGOs each would be collected at one time.

        Critical step

        There are three key desirable attributes for all hGOs when selecting for re-embedding: visibly clear hGOs lumens, visibly independent hGOs (no excessively tightly packed groups), and minimal visible unintended neural growth (fuzzy appearance) (Fig. 4c, green arrows).

        Critical step

        Maintaining excessive spare medium will dilute the embedding Matrigel and may cause issues with Matrigel bead formation and/or proper 3D suspension of hGOs.

      7. (vii)

        Using a cut 1,000-μl tip, collect 350 μl of ice-cold Matrigel, enough to plate 6 × 50-μl Matrigel beads plus spare, and transfer it to a microcentrifuge tube filled with Matrigel sections. Once the cold Matrigel has been added, quickly stir and pipette the hGOs up and down several times with a cut 200-μl barrier pipette tip to mix thoroughly.

        Critical step

        Work quickly, because the Matrigel begins to solidify as it warms up, usually in under a minute. Concurrently, be mindful to avoid introducing air bubbles into the suspension.

      8. (viii)

        After mixing, collect 50 μl of hGO suspension with the cut 200-μl barrier pipette tip, visually ensuring you have five to ten hGOs, and pipette it into the center of a well in a 24-well or 4-well dish. Keep the pipette tip slightly elevated from the well floor and ensure that the Matrigel bead does not touch the well wall while pipetting. Quickly, plate six beads in total from the microcentrifuge tube and then repeat Step 33A(vi–viii) as needed, typically filling a 24-well dish or a set of 4 × 4-well dishes in one sitting.

        Critical step

        Before the Matrigel completely solidifies, ensure that the hGOs are in the center of the Matrigel bead by nudging them into place with the outside of the cut 200-μl pipette tip, if necessary. If the hGOs are not in the center of the Matrigel bead, they are at an increased risk of growing outside the Matrigel, in which case, they will flatten and lose the desired 3D morphology.

        Critical step

        Keep in mind the concerns regarding the type of tissue culture dish used, the introduction of air bubbles, and avoiding disturbing the dish immediately following plating Matrigel beads as discussed in Step 28.

    2. (B)

      Swirling method of reducing density

      Timing 30 min per new 24-well plate

      1. (i)

        Perform Step 33A(i–iii).

      2. (ii)

        With the cut 200-μl barrier pipette tip, aggressively pipette the well’s contents until the Matrigel bead has been broken up and only single and small groupings of hGOs remain.

        Critical step

        It is common to break some hGOs during the aggressive pipetting, but it typically is not an issue because of the number of hGOs per Matrigel bead.

      3. (iii)

        Repeat Step 33B(ii) with additional wells, ensuring that sufficient hGOs have been separated into single hGOs or small groupings, so they can be re-embedded at five to ten hGOs per well.

      4. (iv)

        With the cut 200-μl barrier pipette tip, transfer all separated hGOs to the dish containing warm advanced DMEM-F12.

        Critical step

        Use separate dishes for different lots of hGOs, hAGOs or hFGOs, or those generated from different hPSC lines. Dishes with hGOs not currently in use can be kept temporarily in a tissue culture incubator.

      5. (v)

        Once all desired separated hGOs have been transferred to a dish, begin to gently swirl the dish in small, tight circles.

      6. (vi)

        During the gentle swirling, the desirable hGOs should naturally gravitate toward the center of the dish, while neural tissue, large Matrigel chunks and broken hGOs will gravitate away from the center of the dish.

        Critical step

        Swirling commonly takes anywhere between 30 s and 2 min to achieve the desired selection gradient, and it can be repeated as needed. Swirling works for both hAGOs and hFGOs, but works especially well for hFGOs. A learning curve is anticipated before a researcher’s efforts yield an optimal hGOs selection, and the selection is aided by larger numbers (>200) of hGOs per dish. Usually, the more single and near-single hGOs in the dish, the better the selection.

      7. (vii)

        With the cut 200-μl barrier pipette tip, collect desirable hGOs from the center of the dish and transfer them to 1.7-ml microcentrifuge tubes. Repeat this process as needed and remove any excess medium from the tubes once the hGOs have settled at the bottom. For example, if making a new 24-well plate, then four microcentrifuge tubes filled with 35–70 hGOs each would be collected at one time.

        Critical step

        Consider the three desirable attributes (described in Step 33A(vi)) when selecting hGOs to continue to culture.

      8. (viii)

        Perform Step 33A(vii,viii) to re-embed the hGOs. Repeat Step 33A(vii,viii) as needed for additional hGOs.

  2. 34

    After each well of the 24-well dish contains a bead of Matrigel with five to ten hGOs, place the dish into a tissue culture incubator for 10–15 min to allow the Matrigel to completely solidify.

    Critical step

    When the Matrigel is completely solidified, it should be possible to invert or mildly shake the dish without the Matrigel bead moving or disrupting its 3D structure.

  3. 35

    Add 500 μl of gastric (antral or fundic) growth medium to each well as desired, ensuring that each Matrigel bead is covered before placing the dish back into the tissue culture incubator.

  4. 36

    Replace gastric (antral or fundic) growth medium every 4 d, when the phenol red in the medium turns yellow or when the feeding protocol calls for a change of growth factors (see Table 2 for days on which this is required), whichever comes first. Between days 20 and 34, hAGOs and hFGOs develop differing morphological characteristics that are observable by stereoscope (Fig. 6).

    Fig. 6: Formation and differentiation of antral and fundic epithelium in hGOs grown in 3D Matrigel suspension after reducing density.
    figure6

    White arrows identify the development of internal glands in the hAGOs. Yellow arrows identify the development of external glandular buds in the hFGOs. Green arrows highlight an example of a neural cluster that was carried through after the density reduction step. a,b, Low- (left) and high-magnification (right) images of hAGOs and hFGOs in 3D Matrigel suspension developing on day 24 (top rows), day 27 (middle rows), and day 34 (bottom rows). a (Bottom), Developed hAGOs display clear lumens with robust internal glands. b (Bottom), Developed hFGOs display robust external glandular buds with lumens that may or may not contain much cellular debris. Scale bars, 1 mm. The brightness of the images in the bottom rows was adjusted to match that of the other panels.

    Critical step

    By day 24, the hAGO epithelium begins to form an internal glandular morphology that expands and develops over time (Fig. 6a, right panels, white arrows), whereas the hFGOs epithelium becomes wavy and forms external buds that expand and develop over time (Fig. 6b, right panels, yellow arrows).

    Critical step

    By day 34, whereas the hAGOs lumens remain clear, an increasing proportion of hFGOs lumens become filled with cellular debris (Fig. 6a,b, bottom rows). No substantial differences in hFGO quality have been seen between those with clear lumens and those filled with cellular debris.

    Critical step

    Unintended neural growth (Fig. 6a,b, left panels, green arrows) should be minimized in hGOs cultures post day 20 by density reduction and re-embedding.

    troubleshooting

Expanding and passaging of hGOs

Timing variable

  1. 37

    Once hGOs reach day 34, the tissue is considered developed and can be used for further experiments of choice. We have not yet developed a robust protocol for expanding and/or passaging hGOs substantially post day 34. We have found that hGOs have variable survivability at extended time points (~1–2 weeks post day 34), with the viability of hAGOs commonly exceeding that of hFGOs. Currently, developed hGOs have not been re-plated and cultured as a 2D monolayer after prolonged 3D culture.

Troubleshooting

Troubleshooting advice can be found in Table 3.

Table 3 Troubleshooting table

Timing

  • Steps 1–12, single-cell passage of hPSCs (day −1): 2 h

  • Step 13, confirmation that hPSCs have achieved the required confluency (day 0): 5 min

  • Steps 14–17, differentiation of hPSCs into human DE (days 0–2): 3 d

  • Steps 18–23, spontaneous budding of posterior foregut spheroids from human DE (days 3–5): 3 d

  • Steps 24–32, development of spheroids into hGOs (days 6–20): 1 h to plate spheroids; 14 d to grow tissue

  • Steps 33–36, reduction of density and re-embedding of hGOs for continued growth: 30 min–1 h to reduce density per new 24-well plate; 14 d to grow tissue

  • Step 37, expanding and passaging of hGOs: variable

  • Box 1, immunostaining to determine efficiency of DE differentiation (day 3) and posterior foregut specification (day 6): 2 d

  • Box 2, immunostaining to determine efficiency of antral and fundic specification: 4 d

Anticipated results

This protocol outlines an efficient method of directed differentiation that generates hAGOs and hFGOs from hPSCs in vitro. Each step in the differentiation process yields specific and robust results. DE differentiation (protocol day 3 (Fig. 2b)), yields a population of cells ~85–90% pure by SOX17 and FOXA2 double-positive staining (Fig. 3a)1,23,24. During the 3-d posterior foregut induction (protocol days 3–6), morphogenetic tissue movements of the monolayer give rise to 3D structures, culminating in the budding and release of free-floating posterior foregut spheroids. On protocol day 4, an elevated 3D morphology is visible across the surface of the well (Fig. 4a), top); this precedes spheroid formation. On protocol day 5, the 3D spheroid–generating morphology has numerous attached budding spheroids and a small number of free-floating spheroids (Fig. 4a, center). On protocol day 6, there are numerous free-floating spheroids, ~600–800 for a robust generation, whereas the remaining monolayer is largely devoid of the 3D budding morphology (Fig. 4a, bottom). The robustness of spheroid generation is dependent on the efficiency of DE differentiation and the density/confluence of the DE monolayer. The resulting posterior foregut spheroids are SOX2+HNF1β+CDX2 (ref. 1), with the remaining monolayer expressing the same markers at comparable levels (Fig. 3b).

Embedding of posterior foregut spheroids in Matrigel allows for the 3D growth and expansion into hGOs, with epithelium development from simple, to pseudo-stratified, to a complex columnar glandular structure with an apical–basal orientation1,2. On protocol day 9, after 3 d of gastric specification, antral spheroid epithelium is SOX2+GATA4+PDX1+ and fundic spheroid epithelium is SOX2+GATA4+ with low PDX1 expression due to Wnt/β-catenin signaling. Over the next 11 d, antral and fundic spheroids expand into hGOs at a similar rate and efficiency, with >80% of embedded spheroids developing into hGOs (Fig. 4c, top three rows)2. At protocol day 20, patterning is maintained, as both hAGOs and hFGOs contain epithelium expressing the gastric SOX2+GATA4+ signature in >90% of cells, with high levels of PDX1 expression being a hallmark of the hAGOs (Fig. 5)1,2. By protocol day 20, hFGOs are notably slightly smaller and have a thicker epithelium than hAGOs (Fig. 4c, bottom two rows). In addition, hGO cultures can develop some unintended neural growth on the outside of the epithelium that presents in the form of extending neurons or as a dense neuroepithelium (Fig. 4c, green arrows).

After reducing the density of the hGOs, steady growth continues over the next 14 d, with hGOs typically reaching a final size between 2 and 4 mm in diameter (Fig. 6a,b, bottom left)1,2. Around protocol day 24, hAGOs begin to undergo internal complex folding and gland formation (Fig. 6a, right-hand panels, white arrows), whereas the hFGOs develop organized external glandular buds (Fig. 6b, right-hand panels, yellow arrows). Concurrently, hAGOs lumens remain clear, whereas an increasing proportion of hFGO lumens gradually become filled with cellular debris (Fig. 6). No substantial differences in hFGO quality have been observed between those with clear lumens and those filled with cellular debris. Around protocol day 27, both types of hGOs begin to express differentiated gastric cell types1,2, except for parietal cells in hFGOs, which require a 48-h induced differentiation step on protocol day 30 (ref. 2). As recapitulated in the late embryonic to early post-natal mouse (E18.5–P12), epithelial SOX2 is downregulated in the hGOs as the epithelium forms glands. By protocol day 34, the glandular epithelium of both types of hGOs expresses CDH1 (E-cadherin), CTNNB1 (β-catenin), and the gastric-specific CLDN18, in addition to an external minor population of undifferentiated FOXF1+VIM+ mesenchymal cells1,2. In addition, by protocol day 34, both types of hGOs express their respective tissue-specific differentiated epithelial gastric cell types with varying degrees of maturity and functionality (Table 4). In addition to analyzing expression of cell-specific markers, the functionality of parietal cells and chief cells can be analyzed by production of acid and secretion of pepsinogens as described2.

Table 4 Differentiated gastric epithelial cell types obtained

We have seen variable survivability when culturing hAGOs and hFGOs 1–2 weeks post protocol day 34, with hAGO viability commonly exceeding that of hFGOs. Neither hAGOs nor hFGOs display any degree of peristaltic motion, as there is no surrounding well-differentiated layer of smooth muscle nor an incorporated enteric nervous system. hGOs are capable of being serially passaged, as hFGOs have successfully undergone two rounds of passaging by fragmentation through a 20-gauge syringe followed by Matrigel re-embedding as described2. However, although the passaged hFGOs maintained expression of the lineage markers MUC5AC, MUC6, PGC, and ghrelin, the hFGOs were increasingly cystic in morphology, did not contain parietal cells, and were refractory to induced parietal cell differentiation2.

Data availability

The data that support the findings of this study are available from the corresponding author upon request.

References

  1. 1.

    McCracken, K. W. et al. Modelling human development and disease in pluripotent stem-cell-derived gastric organoids. Nature 516, 400–404 (2014).

    CAS  Article  Google Scholar 

  2. 2.

    McCracken, K. W. et al. Wnt/beta-catenin promotes gastric fundus specification in mice and humans. Nature 541, 182–187 (2017).

    CAS  Article  Google Scholar 

  3. 3.

    Bartfeld, S. et al. In vitro expansion of human gastric epithelial stem cells and their responses to bacterial infection. Gastroenterology 148, 126–136.e6 (2015).

    Article  Google Scholar 

  4. 4.

    Bertaux-Skeirik, N. et al. CD44 plays a functional role in Helicobacter pylori-induced epithelial cell proliferation. PLoS Pathog. 11, e1004663 (2015).

    Article  Google Scholar 

  5. 5.

    Dedhia, P. H., Bertaux-Skeirik, N., Zavros, Y. & Spence, J. R. Organoid models of human gastrointestinal development and disease. Gastroenterology 150, 1098–1112 (2016).

    Article  Google Scholar 

  6. 6.

    Schlaermann, P. et al. A novel human gastric primary cell culture system for modelling Helicobacter pylori infection in vitro. Gut 65, 202–213 (2016).

    CAS  Article  Google Scholar 

  7. 7.

    Willet, S. G. & Mills, J. C. Stomach organ and cell lineage differentiation: from embryogenesis to adult homeostasis. Cell. Mol. Gastroenterol. Hepatol. 2, 546–559 (2016).

    Article  Google Scholar 

  8. 8.

    Zorn, A. M. & Wells, J. M. Molecular basis of vertebrate endoderm development. Int. Rev. Cytol. 259, 49–111 (2007).

    CAS  Article  Google Scholar 

  9. 9.

    Munera, J. O. et al. Differentiation of human pluripotent stem cells into colonic organoids via transient activation of BMP signaling. Cell Stem Cell 21, 51–64.e6 (2017).

    CAS  Article  Google Scholar 

  10. 10.

    Rankin, S. A. et al. Timing is everything: reiterative Wnt, BMP and RA signaling regulate developmental competence during endoderm organogenesis. Dev. Biol. 434, 121–132 (2018).

    CAS  Article  Google Scholar 

  11. 11.

    Stevens, M. L. et al. Genomic integration of Wnt/beta-catenin and BMP/Smad1 signaling coordinates foregut and hindgut transcriptional programs. Development 144, 1283–1295 (2017).

    CAS  Article  Google Scholar 

  12. 12.

    Roberts, D. J. et al. Sonic hedgehog is an endodermal signal inducing Bmp-4 and Hox genes during induction and regionalization of the chick hindgut. Development 121, 3163–3174 (1995).

    CAS  PubMed  Google Scholar 

  13. 13.

    Bayha, E., Jorgensen, M. C., Serup, P. & Grapin-Botton, A. Retinoic acid signaling organizes endodermal organ specification along the entire antero-posterior axis. PLoS ONE 4, e5845 (2009).

    Article  Google Scholar 

  14. 14.

    Molotkov, A., Molotkova, N. & Duester, G. Retinoic acid generated by Raldh2 in mesoderm is required for mouse dorsal endodermal pancreas development. Dev. Dyn. 232, 950–957 (2005).

    CAS  Article  Google Scholar 

  15. 15.

    Wang, Z., Dolle, P., Cardoso, W. V. & Niederreither, K. Retinoic acid regulates morphogenesis and patterning of posterior foregut derivatives. Dev. Biol. 297, 433–445 (2006).

    CAS  Article  Google Scholar 

  16. 16.

    Engevik, K. A., Matthis, A. L., Montrose, M. H. & Aihara, E. Organoids as a model to study infectious disease. Methods Mol. Biol. 1734, 71–81 (2018).

    CAS  Article  Google Scholar 

  17. 17.

    Takebe, T., Wells, J. M., Helmrath, M. A. & Zorn, A. M. Organoid center strategies for accelerating clinical translation. Cell Stem Cell 22, 806–809 (2018).

    CAS  Article  Google Scholar 

  18. 18.

    Finkbeiner, S. R. et al. Transcriptome-wide analysis reveals hallmarks of human intestine development and maturation in vitro and in vivo. Stem Cell Rep. 4, 1140–1155 (2015).

    CAS  Article  Google Scholar 

  19. 19.

    Kroon, E. et al. Pancreatic endoderm derived from human embryonic stem cells generates glucose-responsive insulin-secreting cells in vivo. Nat. Biotechnol. 26, 443–452 (2008).

    CAS  Article  Google Scholar 

  20. 20.

    Si-Tayeb, K. et al. Highly efficient generation of human hepatocyte-like cells from induced pluripotent stem cells. Hepatology 51, 297–305 (2010).

    CAS  Article  Google Scholar 

  21. 21.

    Schumacher, M. A. et al. Helicobacter pylori-induced sonic hedgehog expression is regulated by NFκB pathway activation: the use of a novel in vitro model to study epithelial response to infection. Helicobacter 20, 19–28 (2015).

    CAS  Article  Google Scholar 

  22. 22.

    Wroblewski, L. E. et al. Helicobacter pylori targets cancer-associated apical-junctional constituents in gastroids and gastric epithelial cells. Gut 64, 720–730 (2015).

    CAS  Article  Google Scholar 

  23. 23.

    McCracken, K. W., Howell, J. C., Wells, J. M. & Spence, J. R. Generating human intestinal tissue from pluripotent stem cells in vitro. Nat. Protoc. 6, 1920–1928 (2011).

    CAS  Article  Google Scholar 

  24. 24.

    Spence, J. R. et al. Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro. Nature 470, 105–109 (2011).

    Article  Google Scholar 

  25. 25.

    Workman, M. J. et al. Engineered human pluripotent-stem-cell-derived intestinal tissues with a functional enteric nervous system. Nat. Med. 23, 49–59 (2017).

    CAS  Article  Google Scholar 

  26. 26.

    Sato, T. et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459, 262–265 (2009).

    CAS  Article  Google Scholar 

  27. 27.

    Teo, A. K. et al. Activin and BMP4 synergistically promote formation of definitive endoderm in human embryonic stem cells. Stem Cells 30, 631–642 (2012).

    CAS  Article  Google Scholar 

Download references

Acknowledgements

We thank D. Kechele for comments on the manuscript. This work was supported by grants from the National Institutes of Health (R01DK092456, U19AI116491, and P01HD093363 to J.M.W.). We also acknowledge core support from the Pluripotent Stem Cell Facility of Cincinnati Children’s Hospital Medical Center. We acknowledge core support from a Cincinnati Digestive Disease Center award (P30DK0789392).

Author information

Affiliations

Authors

Contributions

J.M.W., K.W.M., and T.R.B. conceived the study and experimental design. T.R.B. and J.M.W. co-wrote the manuscript. T.R.B. produced all images for the figures, and J.M.W. produced the protocol schematic. K.W.M. and T.R.B. analyzed the data and performed experiments.

Corresponding author

Correspondence to James M. Wells.

Ethics declarations

Competing interests

J.M.W. and K.W.M. are listed on the following patent applications: PCT/US2015/032626, ‘Methods and systems for converting precursor cells into gastric tissues through directed differentiation’, and PCT/US2017/031309, ‘Methods for the in vitro manufacture of gastric fundus tissue and compositions related to same’. T.R.B. declares no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Related links

Key references using this protocol

McCracken, K. W. et al. Nature 516, 400–404 (2014): https://doi.org/10.1038/nature13863

McCracken, K. W. et al. Nature 541, 182–187 (2017): https://doi.org/10.1038/nature21021

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Broda, T.R., McCracken, K.W. & Wells, J.M. Generation of human antral and fundic gastric organoids from pluripotent stem cells. Nat Protoc 14, 28–50 (2019). https://doi.org/10.1038/s41596-018-0080-z

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing