Generation of human iPSC-derived pancreatic organoids to study pancreas development and disease

hiPSC culture

hiPSCs (AICS-0090-39)¹⁹ were obtained from the Coriell Institute for Medical Research (CIRM) and maintained according to the manufacturer’s protocol (https://www.coriell.org/0/PDF/Allen/ipsc/AICS-0090-391_CofA.pdf) in Matrigel-coated (Corning) 6 well-plates with mTesr1 growth medium (STEMCELL Technologies) supplemented with 1% penicillin-streptomycin (P/S) in a humidified incubator (37°C, 5% CO₂). hiPSCs were thawed at 37°C in a water bath, and the cell suspension was then transferred dropwise to a Falcon tube containing enough growth medium for a 1:10 dilution. hiPSCs were centrifuged at 300 × RCF for 5 min, resuspended in growth medium containing 5μM Rho-associated protein kinase (ROCK) inhibitor, and then seeded onto Matrigel-coated plates for maintenance.

Cells were passaged as single-cells at 70-80% confluency using Accutase (Invitrogen) and medium supplemented with 5μM ROCK inhibitor Y-27632 (Sigma) on the day of passaging. Passages of hiPSCs exhibiting significant background differentiation were excluded from the differentiation experiments, and hiPSCs were passaged at least once after thawing and prior to commencing differentiation. The number of passages used for all experiments was less than 45. Cryopreservation of hiPSCs was carried out in freezing medium comprising 60% mTesr1, 30% knockout serum replacement (KoSR), and 10% dimethyl sulphoxide for 24h at -80°C using controlled-rate freezing. The cells were then transferred to liquid nitrogen for long-term storage.

hiPSC differentiation into pancreatic progenitors

hiPSCs were seeded at a density of 200,000-300,000 cells/cm² in Matrigel-coated 6-well plates and grown for 24h in mTesr1 supplemented with 1% P/S and 5 μM ROCK inhibitor. Cells were subsequently differentiated in a medium that was replenished daily and supplemented with cytokines to induce their differentiation into pancreatic progenitors. The cytokine and small compound concentrations used were as described by Barsby et al.²⁰

Briefly, the first Basal Medium (basal medium 1) used between days (D) 1 and 5 of differentiation was composed of MCDB131 medium supplemented with Glutamax (1x), glucose (2.5 M), NaHCO₃ (1.5g/l), and BSA (0.5%). The second Basal Medium (basal medium 2) used between D6 and the end of differentiation was made of MCDB131 medium supplemented with Glutamax (1x), glucose (2.5 M), NaHCO₃ (2.5g/l), BSA (2%), and ITS-X (0.5x). The reagents used are listed in Table 1.

The supplements used to differentiate the pancreatic progenitors were iteratively added to the basal media, as detailed in Table 2.

Formation and culture of pancreatic organoids

On D11 of differentiation, pancreatic progenitors (PP) were dissociated using TrypLE (Gibco) and left to aggregate in the Aggrewell microwell system (STEMCELL Technologies) for 72h. Progenitors were seeded in Aggrewell plates at a density of 1200 cells/microwell. After PP clusters had formed in the Aggrewell (72h), they were transferred to suspension plates, and the basal medium was supplemented with the cytokines and small compounds listed in Table 3.

Here, we describe a step-by-step protocol for the generation and culture of pancreatic organoids derived from pancreatic progenitors cultured in 6-well plates and collected on D11 of differentiation. The general steps of cell dissociation, aggregation, and suspension culture were similar to those described by Barsby et al. However, the timing and culture media have been modified and optimized to induce exocrine differentiation and branching morphogenesis, instead of endocrine differentiation. The reagents used in these steps are listed in Table 2.

Day 11

Preparation of the Aggrewell microwell system (24-well plate)

Dissociation and collection of pancreatic progenitors

Day 12-13

Media changes

Add 1 ml of fresh pre-warmed D11-D13 medium gently on the top side of the wells.

Day 14

Prepare D14 culture medium

On this day, cell clusters collected from 2 wells of the Aggrewell plate are pooled in one ultra-low attachment (ULA) well of a 6-well plate.

Transfer of pancreatic progenitor clusters to suspension culture

Day 15

Day 16-18

Day 19

Experiments were performed using pancreatic organoids or following the steps below to fix the organoids.

Immunofluorescence and RT-qPCR

Immunostaining of cells

In addition to the 6-well plates used to generate pancreatic progenitor-derived organoids, hiPSCs were also cultured in 24-well plates in 1 ml differentiation medium at a cell seeding density of 200,000-300,000 cells/cm². These plates were used to confirm the efficiency of hiPSC differentiation into pancreatic progenitors by immunostaining for acinar and ductal progenitor markers at D14. Cells were washed in 1ml PBS, fixed for 20 min at 4°C with 4 % paraformaldehyde, and washed again. Fixed cells were then blocked for 45min in a blocking buffer composed of PBS, Donkey serum (Sigma) (3%), and Triton-X (0.1%). Primary antibodies were added to the blocking buffer at dilutions listed in Table 4 and incubated overnight at 4°C. The next day, cells were washed multiple times in PBS with Triton-X (0.1%) and incubated for 1 h in the dark with the secondary antibodies diluted in blocking buffer ( Table 4) and counterstained with Hoechst. Before imaging, the cells were washed multiple times with PBS.

Immunostaining of organoids

Throughout the staining procedure, the solutions were changed by centrifuging the organoids at 200 g for 2 min, removing the supernatant, and adding the next solution.

First, fixed organoids were blocked for 24h in a blocking buffer made from PBS, Donkey serum (3%), and Triton-X (0.5%) at 4°C on a rotating platform. Primary antibodies were added to the blocking buffer at the dilutions listed in Table 4 and incubated for 24h under similar conditions. The next day, organoids were washed three times for 45min in PBS with Triton-X (0.1%), and then incubated for 24h in the dark at 4°C on a rotating platform with the secondary antibodies diluted in blocking buffer ( Table 4), and counterstained with Hoechst. Before imaging, organoids were washed twice in PBS with Triton-X (0.1%) for 45min and once in PBS.

Imaging and image analysis

Immunostained organoids were placed in a 96-well plate with a glass bottom and imaged using a Zeiss LSM 700 confocal microscope. Using ImageJ,²¹ we quantified cell numbers using the in-built cell counter tool and measured PDX1 staining intensity levels in the nuclei by creating a mask on the Hoechst channel. PDX1 intensity was quantified by measuring the mean pixel intensity in the region of interest (ROI) delineating the AMY+ and NKX6.1+ cells.

RT-qPCR

Cells from the planar culture were dissociated and harvested with TrypLE, and organoids/cell clusters were collected with a trimmed pipette tip pre-coated with 1mL of medium previously removed from the well. These were centrifuged for 3 min at 300 RCF or for 2 min at 200 RCF, respectively. The supernatant was aspirated, and the pellets were washed once with PBS (and then frozen at -80°C). A High Pure RNA Isolation Kit (Roche) was used for RNA extraction and all samples were treated with DNase to remove genomic DNA. cDNA was synthesized using a Transcriptor First Strand cDNA Synthesis Kit (Roche). RT-qPCR was performed on a LightCycler (Roche) with FastStart Essential SYBR Green Master Mix (Roche). The 2-ΔΔCt method²² was used to evaluate the relative gene expression, and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used for normalization. The cDNA dilution was modified to return Cts of approximately 20 for GAPDH, and missing Ct values were set to 40 when the threshold for detectable fluorescence was not met. The primers used for RT-qPCR are listed in Table 5.

Quantification and statistical analysis

All statistical analyses were performed using the R package rstatix. The elements shown in the boxplots are the 25% (Q1, upper box boundary), 50% (median, black line within the boxes), and 75% (Q3, lower box boundary) quartiles, and the whiskers represent a maximum of 1.5X × interquartile range. P-values were calculated using a two-tailed t-test. Normality was assessed using the Shapiro-Wilk normality test. Statistical significance was calculated using unpaired Student’s t-tests for data with normal distributions, and Mann-Whitney U-tests otherwise.

Generation of pancreatic progenitors

hiPSCs were differentiated into pancreatic progenitors as described by Barsby et al.²⁰ with minor modifications, which are outlined in the Materials and Methods (Figure 1A). The efficiency of pancreatic progenitor differentiation was monitored by assessing the morphology of differentiating cells with a phase contrast light microscope between day (D8) and D10 of differentiation and by immunostaining for the pancreatic progenitor markers PDX1 and NKX6.1, at D11 of differentiation (Figure 1B). Pancreatic progenitors were predominantly double-positive for PDX1 and NKX6.1, and clusters were efficiently formed within 24 h in the Aggrewell plate.

Figure 1. Differentiation of hiPSCs into branched pancreatic organoids.

A. Schematic overview and media composition for the differentiation of hiPSCs into pancreatic organoids. The differentiation was carried out as a monolayer until day 11 (D11), then transferred to the Aggrewell microwell system from D12 to D14, and finally performed in suspension until D19. B. Representative phase contrast image (left panel) and confocal image (right panel) of D11 pancreatic progenitors immunostained with the PDX1 and NKX6.1 pancreatic markers. C. Representative phase contrast images of pancreatic clusters cultured in the Aggrewell microwell system on D14 of the differentiation (left panel) and in suspension on D19 of the differentiation (right panel). AA: Activin A, RA: Retinoic Acid.

Generation of branched organoids

The dissociated pancreatic progenitors were left to aggregate in Aggrewell plates (Figure 1C). After progenitor clusters were formed, they were transferred to suspension plates for the final stage of differentiation, using an optimized differentiation medium (see composition in Table 3). Specifically, we supplemented the medium with GSiXX, a Notch inhibitor known to promote pancreatic cell differentiation,^23,24 along with FGF7 and FGF10, which stimulate the growth and morphogenesis of the exocrine pancreas.²⁵ In contrast to Barsby et al.,²⁰ we excluded factors known to induce endocrine differentiation, such as thyroid hormone receptor agonists, TGF-β antagonists, and betacellulin. This medium, combined with the suspension culture format, triggered primary branch formation in organoids between D17 and D18. 93.5% (SD = 2.3) of organoids acquired a branched architecture by D19 of the differentiation, highlighting a high efficiency of morphogenesis (Figure 1C; 2A).

Figure 2. Physiological patterning of the pancreatic organoid branches.

A. Representative confocal images of D14 pancreatic clusters (left panel) and D19 pancreatic organoids (right panel) immunostained for PDX1, NKX6.1 and Amylase (AMY), which respectively mark all pancreatic progenitor cells, trunk progenitor cells and acinar cells. B. Ratio of PDX1+ cells positive for the trunk marker NKX6.1 at D14 and D19 (left panel). Ratio of PDX1+ cells positive for acinar marker Amy at D14 and D19 (right panel). N=13-15 organoids per stage analysed, from N=3 independent differentiation experiments. P-values generated from two-tailed t-tests are displayed on the boxplots.

To investigate the efficiency of our culture medium to induce exocrine differentiation in organoids, we performed immunostaining for tip and trunk markers (Figure 2A). Our whole-mount immunostaining highlights that the stalk region of the organoid branches is positively marked by the NKX6.1 trunk marker, whereas the tip regions are marked by the amylase acinar marker. Hence, the pattern of expression of both these markers is mutually exclusive and clearly defines the tip and trunk domains in organoids. In addition, our quantification showed that the acinar domain strongly increased in the last stage of differentiation between D14 and D19, demonstrating that our protocol supports acinar cell differentiation (Figure 2B). Altogether, our differentiation protocol recapitulated the patterning of the pancreatic branches observed in vivo, with acinar cells positioned at the tip of the branches.

Analysis of acinar cell differentiation

It has been previously shown that during in vivo pancreatic development, PDX1 expression levels decrease in differentiating acinar cells.²⁶ To confirm that this process was recapitulated in our organoid system, we quantified PDX1 staining intensity in AMY+ acinar cells compared to NKX6.1+ cells in organoids collected on D19 (Figure 3A). Our results showed that the PDX1 staining intensity was significantly lower in AMY+ acinar cells than in NKX6.1+ cells at this stage, thus recapitulating the decrease in PDX1 expression observed in vivo.

Figure 3. Differentiation of acinar cells in branched pancreatic organoids.

A. Representative confocal images of D19 pancreatic organoids immunostained for PDX1, NKX6.1 and AMY, which respectively mark all pancreatic progenitor cells, trunk bipotent cells and acinar cells (left panel). PDX1 is shown alone in the right panel. Dotted green and red lines respectively delineate the NKX6.1+ trunk and AMY+ tip regions. B. Quantification of PDX1 staining intensity in the NKX6.1+ and AMY+ cells. N = 17-22 organoids analysed, from N = 3 independent differentiation experiments. The p-value generated from two-tailed t-test is displayed on the boxplot. C. Representative qRT-PCR analysis (N = 1 shown) for the indicated pancreatic genes performed on cells, clusters and organoids collected at D0, D14 and D19 of the differentiation protocol. Data are represented as relative fold change. Values shown are mean ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; two-tailed paired t tests.

In addition, we performed RT-qPCR on dissociated cells collected at D0, D14, and D19 to investigate the efficiency of our protocol in inducing cell differentiation, specifically the acinar fate. Our results highlight that differentiating cells acquire a pancreatic fate by D14, as evidenced by the strong expression of the pancreatic progenitor markers NKX6.1 and CPA1 (Figure 3B). Between D14 and D19, we observed that the level of Amylase expression was strongly increased in organoids, and CPA1 expression was also maintained, indicating that our organoid culture medium promoted the differentiation of acinar cells. Altogether, the differentiation of acinar cells and the formation of a branched architecture demonstrated that our protocol enables the formation of pancreatic organoids with an exocrine identity.