This protocol describes how to produce retinal pigment epithelial cells (RPE) from pluripotent stem cells. The method uses a combination of growth factors and small molecules to direct the differentiation of stem cells into immature RPE in fourteen days and mature, functional RPE after three months.
We describe a robust method to direct the differentiation of pluripotent stem cells into retinal pigment epithelial cells (RPE). The purpose of providing a detailed and thorough protocol is to clearly demonstrate each step and to make this readily available to researchers in the field. This protocol results in a homogenous layer of RPE with minimal or no manual dissection needed. The method presented here has been shown to be effective for induced pluripotent stem cells (iPSC) and human embryonic stem cells. Additionally, we describe methods for cryopreservation of intermediate cell banks that allow long-term storage. RPE generated using this protocol might be useful for iPSC disease-in-a-dish modeling or clinical application.
The retinal pigment epithelium is a monolayer of pigmented cells that provides crucial support for photoreceptors. Retinal pigment epithelial cells (RPE) have numerous functions in vision, including light absorption, nutrient and ion transport, retinoid cycling, photoreceptor outer segment phagocytosis, and growth factor secretion1. There are a variety of retinal dystrophies that affect the function of RPE and result in a loss of vision, including age-related macular degeneration and retinitis pigmentosa. Generation of RPE from pluripotent stem cells may facilitate research to understand these eye diseases, and can provide an unlimited source of RPE for cell therapies2. In fact, multiple clinical trials are underway using RPE derived from pluripotent stem cells3.
This differentiation protocol was originally described by Buchholz4 and was based on the previously published method from Clegg5. The procedure mimics the normal in vivo developmental process to direct undifferentiated pluripotent stem cells towards an RPE fate via manipulation of the insulin growth factors (IGF), basic fibroblast growth factor (FGF-2; FGF-basic), transforming growth factor beta (TGF-β), and WNT pathways4,5. The protocol was significantly improved by addition of a WNT pathway agonist late in the protocol, which yielded 97.77% ± 0.1% pre-melanosome protein (PMEL) positive cells, and has been adapted to xeno-free conditions6,7. The resulting RPE have been shown to express RPE markers at the transcript and protein levels, to secrete known RPE growth factors with appropriate polarity, and carry out phagocytosis of photoreceptor outer segments8. This protocol is more rapid and reliable than "spontaneous" protocols of differentiation that involve simple removal of basic fibroblast growth factor8. Furthermore, RNA sequencing data show that RPE obtained using this protocol are very similar to those obtained using the more common spontaneous approach8. The 14-day method generates RPE that fit the "5 P's" mentioned by Mazzoni9 (pigmented, polarized, phagocytic, post-mitotic, polygonal)9. While this procedure has proven to be reproducible in multiple labs, we wish to acknowledge several additional directed differentiation methods that have been published in recent years10,11,12,13.
1. Preparation of Reagents for Day 0 to Day 14 of the Protocol
2. Day 0: Day of Pluripotent Stem Cell Passage for Differentiation
3. Day 1 to 14: Addition of Growth Factors
4. Day of Enrichment to Passage 0 of RPE
5. Maturation: Passage 1 and 2 of RPE
NOTE: volumes are indicated for 1 well of a 6-well plate or a T75 flask as indicated by parentheses.
6. Creating an Intermediate Cell Bank: Cryopreservation of Passage 2 day 3-5 RPE
NOTE: Cryopreserve cells while they are subconfluent (~50%) and have not regained pigment.
This method results in the production of a homogeneous, pigmented, and cuboidal monolayer of RPE. The timeline in Figure 1 corresponds to the images depicted in Figure 2. As shown in Figure 2A, the stem cell colonies are tightly packed with defined edges and no fibroblastic cells between colonies or opaque cells within colonies. Figure 2B provides a representation of immature RPE that are subconfluent. If the cells are already confluent at this stage, they cannot extend projections that are critical to the differentiation process. Cells that are severely subconfluent will not be able to establish a monolayer and form tight junctions, characteristic of epithelium. Details on how to optimize this confluence are outlined in the discussion section. Figure 2C shows the morphology of the two most common types of non-RPE that may arise during this differentiation process: neural or fibroblastic patches. It is important to note that these neural patches appear especially opaque on a dissecting microscope, whereas the defined, fibroblastic-like patches are nearly translucent on a dissecting microscope. It can be helpful to mark these areas on a tissue culture plate with an ethanol-proof lab pen to more easily identify them on both a compound light microscope and dissecting microscope. Figures 2D-F show the characteristic bright borders, cobblestone morphology, and pigmentation that indicate a healthy, maturing culture of RPE. Figure 3 is a higher magnification image to show the different appearance of fully mature RPE depicted by phase contrast and bright field microscopy. At passage 3 day 30, the cells are ready for the characterization that has been described in previous publications, including RNA expression, protein expression, growth factor secretion, and phagocytosis2,4,6,7. These characterizations show that the cells represented in these images are not only pigmented and cuboidal, but also phagocytic, post-mitotic, and polarized.
Figure 1:. Timeline for the addition of growth factors and maturation of RPE. Growth factors are added to 12-well plates from day 0-14. Maturing RPE are cultured in 6-well plates or T75 flasks from day of enrichment to 30 days post-thaw (passage 3 day 30). Arrows indicate enzymatic cell passaging. (A-E) below the timelines correspond to the images in Figure 2. Please click here to view a larger version of this figure.
Figure 2: Representative morphology and confluence of maturing RPE. Induced pluripotent stem cells immediately before passaging for differentiation (A). Immature RPE cells subconfluent at day 2 (B) and before pick-to-remove enrichment on day 14; non-RPE patches (indicated by white arrows) appear as patches or opaque "ribbons" (C). RPE at passage 0, 1, and 3 on day 30 (D, E, and F). Scale bar = 200 µm. Please click here to view a larger version of this figure.
Figure 3: Mature RPE at passage 3: Day 30 Cuboidal morphology depicted in phase contrast (A) and pigmentation depicted in bright-field (B). Scale bar = 50 µm. Please click here to view a larger version of this figure.
This protocol describes how to produce retinal pigment epithelial cells from pluripotent stem cells. The method was optimized using both human embryonic and induced pluripotent stem cells from a feeder-free, serum-free culture method. Since the initial isolation of human embryonic stem cells in 1998 and the derivation of induced pluripotent stem cells (iPSC) in 2007, a multitude of stem cell culture methods have been developed14,15,16,17. These methods should be sufficient for producing stem cell colonies that are susceptible to this differentiation. There are no known limitations to the applicability of this method to properly derived and maintained pluripotent stem cells.
The most critical steps are the passaging of stem cells to day 0 of differentiation (step 2.5) and the potential need for manual dissection at day 14 of the process (step 4.5). When picking to remove differentiated cells from the stem cell colonies, refer to the images in Kent18. As indicated, the fibroblastic cells between colonies and the opaque cells within colonies indicate differentiated cells that need to be removed before beginning this protocol18. Only undifferentiated, tightly packed colonies with defined edges should be passaged for differentiation.
The number of stem cells seeded per well (step 2.6.7) is complicated by the fact that the stem cells cannot be triturated into a single cell suspension upon passage and cannot be accurately counted using a hemocytometer. The approximation of 80% confluent stem cells is indicated for passaging 1 well of a 6-well plate into 4 wells of a 12-well plate. Differences between stem cell lines, such as growth rate, can affect how quickly the immature RPE reach confluence between days 0 to 4. The stem cells will produce RPE regardless of precise confluence, but the cell yield will be negatively affected if the cells are too sparse at this stage. The immature RPE cells should be approximately 40-50% confluent on day 1 and nearly 100% confluent by day 4. If the cells are not producing a confluent monolayer by day 4 or 6, the protocol should be repeated at a higher seeding density at day 0. For example, if 1 well of a 6-well plate was passaged to 4 wells of a 12-well plate at day 0 and the immature RPE are not 100% confluent at day 4, reduce the seeding to a 1:3 or 1:2 passage on day 0 or allow the stem cells to become more confluent before passaging. It is critical to establish a consistent seeding density when comparing multiple cell lines.
The manual dissection step at day 14 is only necessary when non-RPE cells are present in culture (Figure 2C). Since the addition of CHIR99021 to the protocol, many pluripotent stem cell lines require little to no manual dissection. Some preparations have a higher incidence of neural patches and it is critical to remove those cells. If the RPE are not viable at passage 0 through passage 3, it is possible to repeat the differentiation protocol taking sufficient time to remove all non-RPE cells. This does not happen often, but it is mentioned here to note that the dissection step on day 14 can be optimized when needed.
There are a variety of RPE differentiation protocols that vary in cost as well as culture methods, efficiency, quantification, and functional assessment, the latter of which has been reviewed thoroughly2. We prefer the 14-day method detailed here because of its efficacy, adaptability, and applicability to a wide range of cell lines4,7,8. The cryopreservation step in this protocol also provides a major advantage in creating an intermediate cell bank for future use, avoiding lot-to-lot variability in experiments. Starting with only 4 wells of a 12-well plate, it is possible to expand into 6-well plates at passage 0 and T75 flasks at passage 1 and 2. At passage 2 day 3-5, when the cells are still subconfluent and have not regained pigment, it is possible to cryopreserve tens of millions of cells and then thaw the mature RPE, designated passage 3 day 30, to check RNA expression, protein expression, growth factor secretion, phagocytosis, etc. We have also established protocols to expand RPE for up to 13 passages 19.
Looking forward, this method will be useful for iPSC modeling of ocular disease and for generation of RPE for cellular therapy. With regards to iPSC disease modeling, this protocol is currently being used in the lab to produce RPE from CRISPR-corrected lines with non-corrected controls from the same patient. Furthermore, this protocol is adaptable to synthetic substrates and xeno-free conditions that are useful for adhering to the good manufacturing practices required for a cellular therapy.
The authors have nothing to disclose.
This work was supported by the Garland Initiative for Vision, the California Institute for Regenerative Medicine (CIRM; grants DR1-01444, CL1-00521, TB1-01177, FA1-00616 and TG2-01151), The Vermont Community Foundation, The Breaux Foundation, and the Foundation Fighting Blindness Wynn-Gund Translational Research Acceleration Program.
SterilGARD III laminar flow biosafety cabinet | Baker | model: SG603A-HE, type: A2, class: 2 | 6' Baker laminar flow biosafety cabinet |
Dissection Hood | Labconco | Model 3970405 | laminar flow bench top |
dissecting microscope | Nikon | SMZ 1500 | heated stage |
air-jacketed CO2 incubator | Sanyo | MCO-17AIC | 37 oC and 5% CO2 |
inverted phase contrast microscope | Olympus | IX53 | |
Name | Company | Catalog Number | Comments |
Media Components | |||
DMEM/F12 | Gibco | 10565042 | |
N2 Supplement | Gibco | 17502048 | |
B27 Supplement | Gibco | 17504044 | |
NEAA | Gibco | 11140050 | |
Name | Company | Catalog Number | Comments |
Growth Factors and Reagents | |||
Nicotinamide | Sigma | N0636 | |
Recombinant mouse noggin | R&D systems | 1967-NG-025 | |
Recombinant human DKK-1 | R&D systems | 5439-DK-010 | |
Recombinant IGF-1 | R&D systems | 291-G1-200 | |
FGF-basic | Peprotech | 100-18B | |
Recombinant human/mouse/rat Activin A | Peprotech | 120-14E | |
SU5402 FGF inhibitor | Santa Cruz Biotechnology | sc-204308 | |
Name | Company | Catalog Number | Comments |
Substrates | |||
Matrigel Basement Membrane Matrix, Phenol Red-Free, LDEV-Free | Corning | 356237 | extracellular matrix-based hydrogel (ECMH) |
Matrigel hESC-Qualified Matrix, LDEV-Free | Corning | 354277 | growth factor reduced ECMH |
Name | Company | Catalog Number | Comments |
Other reagents | |||
1X Versene (EDTA) | Gibco | 15040066 | |
DPBS | Gibco | 14190250 | |
1X PBS (no calcium, no magnesium) | Gibco | 10010023 | |
TrypLE (trypsin-like dissociation enzyme, TDE) | Gibco | 12563011 | |
X-VIVO 10 (RPE supporting medium) | Lonza | BW04-743Q | |
Y-27632 | Tocris | 12-541-0 (1254) | |
CryoStor CS10 | BioLife Solutions | 210102 | cryopreservation medium |
1.2 mL Cryogenic Vial | Corning | 430487 | |
Mr. Frosty (freezing container) | Nalgene | 5100-0001 | freezing container |
Normocin | Invivogen | ant-nr-2 | antimicrobial reagent |
Name | Company | Catalog Number | Comments |
Other Equipment | |||
Pipet-aid | Drummond | 4-000-101 | |
12-well culture plate | Corning | CLS3516 | Used during differentiation. |
T75 flask | Corning | 430641 | Used during RPE maturation. |
6-well culture plate | Corning | CLS3513 | Used during RPE maturation. |
cell scraper | Corning | 08-771-1A | Used during passages. |
cell strainer | Falcon | 352340 | Used during passages before cell count. |