Differentiation protocol for generating functional pancreatic β cells from human pluripotent stem cells


 The efficient generation of pancreatic β cells from human pluripotent stem cells may allow us to study their biological characteristics and use them for the treatment of type I diabetes. The protocol we present in the study provides an efficient method for producing β cells using either human embryonic stem cells or human induced pluripotent stem cells as the starting material.

However, the application potential of current methodologies remains limited because of several major issues, including moderate rate of e ciency (10%~40% NKX6.1+/INS+), high heterogeneity accompanying with high percentage of unwanted cell types (such as dual-hormonal GCG+/INS+ cells), and highly variable e ciencies depending on the origin of the manipulated cell line. To overcome these issues, we have developed a novel stepwise protocol by creating an e cient method for making pancreatic progenitor 3-D clusters, introducing an extra step to potentiate pancreatic progenitors with a cocktail of 10 chemicals, and identifying new factors or chemicals or their combinations for generating β cells from pancreatic progenitors. Using these improvements, we are able to generate ~60%-80% of β cells from hPSCs, and these β cells can rapidly reverse hyperglycemia in the mouse model of diabetes.  Reagent setup -MS12 medium (800ml) is made with 744ml MCDB131 medium, 3.2ml 45% glucose, 20ml 20% fatacid free BSA, and 16ml 7.5% sodium bicarbonate.
3. Dissociated single cells were rinsed twice with DMEM/F12 and spun at 300g for 3 min. The resulting cells were re-suspended in mTeSR™1 which was supplied with 10μM Y27632 (Sigma-Aldrich), and seeded on 1:30 diluted Matrigel-coated dishes at a density of ~55,000 cells/cm 2 .
4. The next day, the medium was exchanged for mTeSR™1 and maintained for one more day prior to differentiation initiation.
Tip: (1) 90% con uence (at the moment for starting differentiation) is important as the cell density too low or too high can greatly reduce the nal production of the β cells.
(2) Some cell line may grow a little slowly, thus increasing the seeding cell quantity or extending the cell growth time will facilitate these cells to reach the appropriate con uence. 9. After PBS rinse, cells are exposed to MS12 medium with KGF (50ng/ml), B27 (100x), Vitamin C (Vc, 0.25mM) and dorsomorphin (0.75µM). Cells were fed with fresh medium daily.
Tip: Handle the cells very gently at the beginning of this stage, as cells are prone to detach from the dish bottom. 14. The resulting cells were re-suspended in aggregation medium (made of 5a-Medium supplied with 10μM Y27632).
15. Cell solution with 0.1-0.4 (according to experimental design) million cells was added into each well of V-bottom 96-well plate, followed by a spun at 300g for 3 min.
16. The plate was put into 37°C incubator for 8-12 hours for cluster formation.
Tip: If cells do not settle to the bottom of plate, centrifuge again by 500g for 3min. This step does not affect the cell viability. 17. Pancreatic progenitor clusters assembled in V-bottom 96-well plate were collected and rinsed twice using DF12 medium, and loaded on 6-well air-liquid interface trans-well.
18. About 1.3ml stage-5 medium (which will enhance the propensity of pancreatic progenitor for generation of β cells) was added for each well.
19. Cells are fed with fresh medium every other day.
Tip: For transferring the clusters to the air-liquid interface, we use the wide bore 200ul pipette tips. 27. Cells were fed with fresh medium every other day. Troubleshooting 1. The hPSCs are heterogeneous and show early sign of differentiation. As this issue will greatly affect the nal e ciency of β cell differentiation, so it must be con rmed before determining whether to proceed or not. If only a few of colonies are differentiated ( attened, enlarged in cell size, or having neurites-like branches), the culture can still be rescued by removing those "bad colonies" using pipette tips before splitting. Repeat this procedure until the whole culture is homogeneous. If too many hPSCs colonies show the propensity of differentiation, start again with a new vial of cells and re-establish the culture.
2. Some hPSC lines may grow a little slowly than others and cannot reach 90% con uence two days after passage (before using for differentiation). Two solutions: (1) Increasing the initial seeding density; (2) Lowering the Activin-A concentrations. Instead of the typical 115-110-100 ng/ml gradient for stage-1, use a gradient of 110-105-95 ng/ml.  2) The starting point cell con uence (right before stage 1) is not ideal. We typically use ~90% cell con uence to start differentiation. If the con uence is lower than that, the e ciency of NKX6.1+/PDX1+ cells at stage 4 will be low, and the production of β cells at the nal stage will be greatly reduced. If the starting point cell con uence is too high, cell differentiation at stage 1 may not be complete. The percentage of de nitive endoderm cells at stage 1 will be lower than expected, and thus the β cell e ciency at the nal stage will be lower. (3) Reagent related issues, such as quality issues, preparation issues and storage issues, could cause inconsistent experimental results across different experimental batches. Here are some examples: Activin-A (and possibly other growth factors used in the protocol) from different vendors or produced by different approaches (bacteria or eukaryotic cells) or from different batches may have different potency. Therefore, it is better to order a large batch of Activin-A from a certain vendor, aliquot it after dissolving it, and store it at -80°C for long-time usage. (4) Taking the cell culture out of the incubator too often or for too long. Avoid taking the culture dish (with cells) out of the incubator too often (for example to check the cell status), as it would negatively affect the nal result. Especially, such practice will signi cantly increase the cell detachment at stage 1. One possible solution is to prepare two culture replicates with one for checking cell status and the other one for the actual experiment. 4. Stage 3, posterior foregut. Cells proliferate quickly and gradually become dense. At the end of this stage, ~90% of the cells are expected to express PDX1. 5. Stage 4, pancreatic progenitor. Cells gradually reduce proliferation. At the end of this stage, >80% of the cells are expected to be PDX1+/NKX6.1+. 6. Making pancreatic progenitor 3D-clusters in V-bottom plate. After 8-12 hours incubation, each well will form a 3D-cluster. There might be some cells (~1% -5%) not integrating into the clusters, which will not affect the later stage of cell differentiation. The percentage of PDX1+/NKX6.1+ cells will be comparable to that of stage 4. 7. Stage 5, posing pancreatic progenitor for generation of β cells. After the treatment, the clusters will become a little more bright and compact. The percentage of PDX1+/NKX6.1+ cells will be comparable to that of stage 4. Typically less than 3% NEUROD1+ cells (representing endocrine related cells) should be observed. Few cells might detach from the clusters at this stage.