This protocol allows for the differentiation of human pluripotent cells into intestinal organoids. The protocol mimics normal human development by differentiating cells into a population of definitive endoderm, hindgut endoderm and then intestinal epithelium. This makes the protocol suitable for studying both intestinal development as well as disease modelling applications.
hiPSC-derived intestinal organoids are epithelial structures that self-assemble from differentiated cells into complex 3D structures, representative of the human intestinal epithelium, in which they exhibit crypt/villus-like structures. Here, we describe the generation of hiPSC-derived intestinal organoids by the stepwise differentiation of hiPSCs into definitive endoderm, which is then posteriorized to form hindgut epithelium before being transferred into 3D culture conditions. The 3D culture environment consists of extracellular matrix (ECM) (e.g., Matrigel or other compatible ECM) supplemented with SB202190, A83-01, Gastrin, Noggin, EGF, R-spondin-1 and CHIR99021. Organoids undergo passaging every 7 days, where they are mechanically disrupted before transfer to fresh extracellular matrix and allowed to expand. QPCR and immunocytochemistry confirm that hiPSC-derived intestinal organoids contain mature intestinal epithelial cell types including goblet cells, Paneth cells and enterocytes. Additionally, organoids show evidence of polarization by expression of villin localized on the apical surface of epithelial cells.
The resulting organoids can be used to model human intestinal development as well as numerous human intestinal diseases including inflammatory bowel disease. To model intestinal inflammation, organoids can be exposed to inflammatory mediators such as TNF-α, TGF-β, and bacterial LPS. Organoids exposed to proinflammatory cytokines display an inflammatory and fibrotic phenotype in response. Pairing of healthy versus hiPSCs derived from patients with IBD may be useful in understanding mechanisms driving IBD. This may reveal novel therapeutic targets and novel biomarkers to assist in early disease diagnosis.
The properties of pluripotent stem cells (PSCs), such as self-renewal and the ability to differentiate to any cell type of the human body, make them valuable tools in the study of development, disease pathology and drug testing1. Human induced pluripotent stem cells (hiPSC) are particularly useful for disease modelling studies as they can be derived from patients, directly capturing the genome responsible for the disease phenotype2,3. Such hiPSCs can be differentiated to the cell type affected by the genetic defect allowing careful examination of molecular mechanism of the disease4.
Differentiation protocols for human PSCs aim to direct differentiation of cells via the main developmental stages by activation or inhibition of specific signaling pathways governing lineage commitment and specification. Maintenance of hiPSC in a pluripotent state requires moderate levels of Activin A (Act-A) signaling, while a high dosage of Act-A for 3 days commits hiPSC to a definitive endoderm (DE) fate5,6. Act-A and Wnt pathways direct anterior-posterior identity of the DE. Signaling by Act-A induces foregut (FG) markers such as HHEX, HNF4α and GATA4, while blocking expression of hindgut (HG) genes such as CDX2. Wnt signaling induces posteriorization of the DE, which then adopts a HG gene expression profile7,8. Once HG cell identity is established, differentiation can be moved from 2D to 3D and directed towards the formation of intestinal organoids.
Intestinal organoids are typically cultured in a 3D extracellular matrix (e.g., Matrigel or other compatible ECMs) based culture system9, comprised of laminin, collagen IV and entactin and enriched with growth factors such as EGF, FGF, PDGF and IGF-1 to contribute to the support survival and proliferation. The organoids are cultured in a defined medium containing Gastrin, Noggin and CHIR to stimulate and support intestinal stem cell growth and proliferation during long-term culture.
After intestinal epithelial cells are embedded into the extracellular matrix, intestinal crypts begin to form and expand eventually forming spheroids. These mature into organoid structures that mimic physiological functioning of intestinal epithelium. Organoids can typically be cultured for more than 1 year without significant loss of functional and gene expression profiles. Passaging is required on a weekly basis using enzymatic digestion of the organoids into smaller fragments which then self-reassemble into complete organoids.
Established organoid lines can be used as a reliable model of numerous disorders linked to the intestine, including Crohn’s Disease, Ulcerative Colitis and colorectal cancers10,11,12,13. This is a preferred model to animal cells as they express human genes linked to these disorders and respond to external stimuli more closely resembling that which occurs in human tissue in vivo.
NOTE: All tissue culture work detailed below should be done in a Class II laminar flow hood.
1. Human induced pluripotent stem cell (hiPSCs) differentiation to the hindgut endoderm
2. Passaging of intestinal organoids
A schematic of the differentiation protocol is shown in Figure 1A. On day 1 of the protocol, hiPSCs should be compact and forming small colonies with total confluence approximately 50-60%. 24 hours following induction of differentiation towards DE cells begin to migrate away from the stem cell colonies to form a monolayer of cells. This continues over the following 3 days and should form a complete monolayer on D3 of DE differentiation (Figure 1). Gene expression should be monitored over the course of differentiation with pluripotency markers (OCT4, NANOG, SOX2) highly expressed on D0 and rapidly downregulated during DE differentiation. During DE differentiation T expression should peak on D1, followed by EOMES and MIXL on D2. On D2 DE genes (SOX17, FOXA2, GATA4, CXCR4) should begin to be expressed and peak at D3 (Figure 2 and Figure 3). Cells should be a monolayer by DE D3 and can then be posteriorized into hindgut endoderm. During the posteriorization event 3D structures will start to form as early as D2. However, sometimes they will only start to appear at D4 or not at all; this is not always indicative of whether or not the cells will proceed and form intestinal organoids (Figure 1). During HG specification CDX2 and HNF4a expression should be induced and increase over time (Figure 3).
Following transfer of 2D sheets of cells into extracellular matrix, clumps of 2D cells will be observed for the first 24 h. After 48 h sheets of cells should begin to auto organize into more compacted 3D spheroid structures that are initially small (Figure 4A) then gradually increase in size and complexity over 7-10 days of culture (Figure 4B & Figure 4C). Organoids should not be passaged until they have achieved a clear organoid/spheroid morphology with obvious epithelium with the lumen facing towards the center of the organoid/spheroid (Figure 4D). At this stage immunocytochemistry can be used to confirm expression of intestinal markers such as villin and CDX2 (Figure 5). Not all clumps of 2D cells will develop into organoids and there will be some contaminating dead cells within the extracellular matrix. These dead sheets of cells should be ignored until the surviving cells have formed large organoids and are ready for passaging.
To model inflammation, TNFα can be added to the tissue culture medium for 24-48 h. Following incubation with proinflammatory molecules, organoids are harvested using the same technique used for their isolation and passaging and then lysed using a cell buffer compatible application such as QPCR or western blotting. If shorter exposures are required, organoids should first be removed from the extracellular matrix and exposed to TNFα in suspension using a 1.5 mL tube. Treatment of intestinal organoids with TNFα for 48 h typically induces expression of pro-inflammatory markers (TNFα, IL1B, IL8, IL23) while negatively affecting expression of intestinal epithelial markers (LGR5, VIL) (Figure 6).
Endoderm basal media | 50 mL | |
RPMI 1640 | 48.5 mL | |
B27 Supplement | 1 mL | |
1% NEAA | 0.5 mL |
Table 1: Composition of endodermal basal media for endoderm differentiation.
Intestinal Basal Media | 50 mL | |
Advanced DMEM/F12 | 46.5 mL | |
HEPES Buffer | 0.5 mL | |
GlutaMAX | 0.5 mL | |
Nicotinamide | 0.5 mL | |
N2 Supplement | 0.5 mL | |
B27 Supplement | 1.0 mL | |
Pen/Strep | 0.5 mL |
Table 2: Composition of intestinal basal media for culture of intestinal organoids
Figure 1: Morphological changes during differentiation of hiPSC via definitive endoderm to the hindgut lineage.
(A) Schematic overview of intestinal differentiation protocol. This cell line of hiPSC forms loose colonies of small cells with a high nucleus to cytoplasm ratio. As the differentiation proceeds, the cells undergo changes consistent with transition from the epithelial to the mesenchymal phenotype and by DE D3 form a uniform monolayer. Once appropriate signals are delivered, DE cells elongate and form a more densely packed monolayer with 3D spheroids appearing as soon as HD D3 but this is dependent upon cell line used and is not a requirement for transition to 3D culture (B). Scale bar: 100 µm. Please click here to view a larger version of this figure.
Figure 2: Differentiation of hiPSCs to HG endoderm induces expression of endodermal genes.
Immunostaining of hiPSC differentiating to definitive endoderm shows changes in expression of TFs at protein level. Pluripotency markers (NANOG and OCT4) are downregulated by DE D3 (A). Expression of mesendoderm marker BRA(T) is present at D1 of the protocol (B) and DE specific TFs SOX17 and FOXA2 appear on D2 (B & C). Scale bar: 200 mm. Please click here to view a larger version of this figure.
Figure 3: Gene expression changes by qPCR during hiPSC differentiation to hindgut endoderm (HG).
Genes associated with pluripotency are downregulated (OCT4, NANOG, SOX2) followed by transient expression of mesendoderm genes (T, EOMES, MIXL1), and finally expression of DE genes (SOX17, FOXA2, CXCR4) and hindgut genes (CDX2, GATA4, HNF4a). Data presented as mean ±SD. Please click here to view a larger version of this figure.
Figure 4: HG endoderm self-assembles to form 3D intestinal organoids in 3D extracellular matrix culture.
HG endoderm is transferred into a suitable 3D extracellular matrix culture and initially forms small solid clumps of cells (A). Clumps of HG endoderm expand over 7-10 days of culture (B) and then become asymmetrical and begin to form a more complex epithelium (C) eventually giving rise to organoids with clear epithelial morphology and a luminal surface facing towards the centre of the organoid (D). Scale bars = 50 µm Please click here to view a larger version of this figure.
Figure 5: Established hiPSC-derived intestinal organoids express intestinal markers.
Immunocytochemistry showing expression of CDX2 and Villin. Scale bars = 100 µm Please click here to view a larger version of this figure.
Figure 6: Effect of TNFα on the inflammatory profile & intestinal cell expression of healthy intestinal organoids.
Inflammatory profile of healthy colonic organoids following 48 hour treatment with TNFα (40 ng/mL). Expression of pro-inflammatory markers (TNFα, IL-8 & IL-23) increases following exposure to TNFα, while at the same time expression of intestinal epithelial markers (LGR5, VIL) are downregulated. Statistical analyses were performed by two-sided student’s t-test. Data are expressed as mean ± SD of each group. *P < .01; **P < .001; ***P < .0001. (n=3). Please click here to view a larger version of this figure.
Here, we describe a protocol for the differentiation of human pluripotent cells into human intestinal organoids. We demonstrate their use to study inflammation; however, this can be applied to various contexts, and coupled with CRISPR/Cas9 gene editing approaches14, on any genetic background. Once differentiated, following the natural developmental differentiation sequence of definitive endoderm, hindgut endoderm and then intestinal epithelium, the resulting organoids can be continually cultured and passaged for more than 12 months.
A critical aspect of this protocol is the initial plating density of undifferentiated stem cells prior to endoderm differentiation. If this is not optimized sufficiently cells will likely die during the initial DE differentiation step (if cells are too sparse) or reduce the efficiency of the DE differentiation (if the cells are too dense). The correct starting density needs to be optimized for the cell line being used, and the correct density should generate a monolayer by the end of DE D3. Flow cytometry should be used to determine efficiency of DE specification and we typically see >80% of cells positive for SOX17 and/or CXCR4. When the number of SOX17 positive cells is less than 60%, the efficiency of HG patterning is impacted, which results in fewer organoids forming when transferred to the extracellular matrix. This will ultimately cause the resulting organoid cultures to fail. To determine if patterning DE to HG has been successful, we assess the number of CDX2 positive cells by flow cytometry and typically expect to see >80% positive cells. Again, if the number of CDX2 positive cells falls below 50% this will have an adverse effect on the number of intestinal organoids that are generated when transferred to 3D extracellular matrix culture.
Following transfer of 2D monolayers into 3D culture, small compact spheres should appear 24-48 h following transfer. Large sheets of dead cells can appear depending on the differentiation efficiency for the cell line used. Instead of immediately passaging cultures to remove these debris, we allow the organoids to fully form and develop their more complex, folded structure. Waiting 7-10 days before attempting the first passing ensures that sufficient dividing cells are present to generate many new intestinal organoids. Any debris that are still present in the culture can be easily removed during the passaging process by slowly spinning the dissociated organoid/debris mixtures in a tube with sufficient speed to pellet the organoids but leave sheet of cells floating in the media. Media and cellular debris can then be aspirated so that only the pellet of organoids remains.
The limitation of this approach is that hiPSC-derived cell types are not often fully mature in terms of gene expression and functional profiles. To determine if hiPSC-derived intestinal tissue is suitable for specific applications the organoids should be characterized for different cell types including enterocytes (VIL), enteroendocrine cells (neurog3), goblet cells (MUC2), transient amplifying cells (CD133), paneth cells (FZD5) and LGR5+ stem cells (LGR5) to determine the organoid cellular composition.
Overall, the major advantage of this protocol over many other organoid differentiation protocols is that this culture platform is very cost-effective due to replacement of several recombinant proteins and conditioned media preparations with small molecules15,16. The differentiation into HG is very simple and fast and can be applied to both human embryonic and induced pluripotent stem cells with identical outcomes. When followed rigorously, and optimized for the cell lines being used it provides a relatively simple model platform free of contaminating mesenchymal cells that can then be applied to study intestinal epithelium in a variety of context including inflammation, host pathogen interactions7. Intestinal fibrosis modeling could be investigated by providing pro-fibrotic stimuli and then assessing expression of extracellular matrix proteins such as collagen, laminin and fibronectin by QPCR, western blot and ELISA. Using CRISPR/Cas9 gene editing techniques on undifferentiated stem cell lines prior to differentiation allows for the creation of gene knockout or protein overexpression organoids that could be used to create disease specific organoids and more complex disease models14,17,18.
The authors have nothing to disclose.
NH is funded by the MRC (MR/S009930/1) and the Wellcome Trust (204267/Z/16/Z), PD is funded by the MRC PhD DTP, KLF is funded by the BBSRC iCASE.
A83-01 | Tocris | 2939 | |
Activin A | R&D | 338-AC | |
Advanced DMEM/F12 (1X) | Life Technologies | 12654-010 | |
B27 supplement | Gibco | 17504044 | |
CHIR99021 | Sigma | SML1046-5MG | |
Epidermal Growth Factor | R&D Systems | 236-EG-01M | |
Gastrin | Sigma Aldrich | G9145 | |
GlutaMAX (100X) | Life Technologies | 15630-056 | |
Growth Factor reduced Matrigel | BD | ||
HEPES Buffer solution (1M) | Life Technologies | 15630-080 | |
N2 Supplement (100X) | Gibco | 17502-048 | |
N-acetyl-cysteine | Sigma Aldrich | A7250 | |
Nicotinamide | Sigma Aldrich | N0636 | |
Noggin | R&D Systems | 6057-NG | |
Non-essential amino acids | Gibco | 11140-050 | |
Paraformaldehyde | VWR | 9713.5 | |
Penicillin/Streptomycin | Gibco | 15140122 | |
Phosphate Buffered Saline | Gibco | 14190-094 | |
Retinoic Acid | Sigma | 302-79-4 | |
ROCK inhibitor | Tocris | 1254/1 | |
ROCK inhibitor Y-27632 | Tocris | 1254 | |
RPMI | Sigma | R8758-500ml | |
R-Spondin-1 | Peprotech | 120-38 | |
SB202190 | Tocris | 1264 | |
TrypLe Express | Gibco | 12604-021 | |
Wnt 3a | R&D | 5036-WN |