Here we present a protocol to differentiate retinal pigment epithelium (RPE) cells from human pluripotent stem cells bearing patient-derived mutations. The mutant cell lines may be used for functional analyses including immunoblotting, immunofluorescence, and patch clamp. This disease-in-a-dish approach circumvents the difficulty of obtaining native human RPE cells.
Although over 200 genetic mutations in the human BEST1 gene have been identified and linked to retinal degenerative diseases, the pathological mechanisms remain elusive mainly due to the lack of a good in vivo model for studying BEST1 and its mutations under physiological conditions. BEST1 encodes an ion channel, namely BESTROPHIN1 (BEST1), which functions in retinal pigment epithelium (RPE); however, the extremely limited accessibility to native human RPE cells represents a major challenge for scientific research. This protocol describes how to generate human RPEs bearing BEST1 disease-causing mutations by induced differentiation from human pluripotent stem cells (hPSCs). As hPSCs are self-renewable, this approach allows researchers to have a steady source of hPSC-RPEs for various experimental analyses, such as immunoblotting, immunofluorescence, and patch clamp, and thus provides a very powerful disease-in-a-dish model for BEST1-associated retinal conditions. Notably, this strategy can be applied to study RPE (patho)physiology and other genes of interest natively expressed in RPE.
It has been documented that at least five retinal degenerative diseases are caused by genetic mutations in the BEST1 gene1,2,3,4,5,6,7,8, with the number of reported mutations already over 200 and still increasing. These BEST1-associated diseases, also known as bestrophinopathies, cause progressive vision loss and even blindness, and there are currently no effective treatments. The protein product of BEST1, namely BESTROPHIN1 (BEST1), is a Ca2+-activated Cl– channel (CaCC) specifically expressed in the retinal pigment epithelium (RPE) of the eyes5,6,8,9. Importantly, a clinical phenotype of BEST1-associated diseases is the reduced visual response to light stimuli, called light peak (LP) measured in the electrooculogram10,11; LP is believed to be mediated by a CaCC in RPE12,13,14. In order to better comprehend the pathological mechanisms of BEST1 mutations and to work towards potential therapies, it is essential to study mutant BEST1 channels endogenously expressed in human RPE cells.
However, obtaining RPE cells directly from live patients is highly impractical. Although native RPE cells can be harvested from biopsies of human cadavers and fetuses, the difficult accessibility to these sources significantly limits scientific research. Therefore, it is critical to have alternative RPE sources other than human eyes. This call has been answered by recent advancements in stem cell techniques, as functional RPE cells can now be differentiated from human pluripotent stem cells (hPSCs), including embryonic stem cells (hESCs) and induced pluripotent stem cells (hiPSCs), the latter being generated by reprogramming primary skin fibroblasts from donors16,17,18. Importantly, the self-renewal and pluripotency of hPSCs ensure a reliable source to generate RPEs, while the patient-specificity of hiPSCs and genomic modification potential of hESCs (e.g., by CRISPR) offer a versatile disease-in-a-dish model for desired BEST1 mutations.
hPSC-RPE has several advantages over mice RPE models: 1) BEST1 knockout mice do not display any retinal abnormality19, raising the possibility of different genetic requirement of BEST1 in RPE between mice and humans; 2) only 3% of human RPE cells are binucleate, in contrast to 35% in mice20; 3) hiPSC-RPE potentiates autologous transplantation in clinical treatment of retinal disorders21. Nevertheless, animal models are still indispensable for studying RPE physiology and pathology in a live system, and the oncogenic potential of hiPSC cannot be overlooked.
The procedure here describes a useful and moderately simple hPSC to RPE differentiation protocol that can be used for research and clinical purposes. This protocol uses nicotinamide (vitamin B3) to augment differentiation of hPSCs to neural tissue, which is further induced to differentiate into RPE by treatment with activin-A. Nicotinamide treatment has been shown to increase the number of pigmented cells (a sign of differentiation into RPE), possibly by attenuating the apoptotic activity of differentiating cells22. The resulting hPSC-RPE cells display the same key markers, cobblestone morphology, and cellular functionality as native human RPE cells22. Thus, in a research setting, the resulting hPSC-RPE cells are suitable for downstream functional analyses including immunoblotting, immunostaining, and whole-cell patch clamp, for which detailed experimental procedures are also provided. Clinically, RPE cells derived from stem cells have shown great potential for transplantation treatment of macular degeneration in both animal studies and human trials23.
1. Differentiation of hPSC to RPE
2. Isolation and Culture of Differentiated RPE Cells
3. RPE Fate Validation by Immunoblotting
4. Checking BEST1 Subcellular Localization by Immunostaining
5. Recording Ca2+-dependent Cl– Current in hPSC-RPE by Whole-cell Patch Clamp
The most technically challenging step is the manual isolation, which aims to achieve a high purity of a differentiated P0 hPSC-RPE population. After a successful isolation, >90% cells in the P0 population will grow and mature to display signature RPE morphologies (Figure 1C). The existence of a minor portion of non-RPE or immature RPE cells in the P0 population is almost inevitable, but will not interfere with the downstream experiments as long as the number of pigmented and cobblestone-shaped hPSC-RPE cells is the overwhelming majority.
With the hPSC-RPE based disease-in-a-dish approach, each patient-specific BEST1 mutation can be comprehensively characterized in vivo for its protein expression (immunoblotting, Figure 3), membrane trafficking (immunostaining, Figure 4), and ion channel function (whole-cell patch clamp, Figure 5). These results will provide critical information for elucidating the pathological mechanisms of the channel mutations, and for developing personalized medicine.
As BEST1 is predominantly expressed in RPE cells, it can be used as a cellular marker for the validation of a mature RPE status. It should be noted that some BEST1 mutations might affect its protein expression, so other well-established RPE markers such as RPE65 and CRALBP still need to be evaluated (Figure 3).
Matured P0 hPSC-RPE cells can be maintained for 2-3 months before splitting (to P1) for whole-cell patch clamp. P0 cells older than 3 months are not recommended for patch clamp but can still be used for immunoblotting and immunostaining experiments.
Figure 1: Differentiated hPSC-RPE cells at different stages. Representative images of pigmented hPSC-RPE clusters in culture plates at first appearance (A), before isolation (B), and after growth to maturity post isolation (C). Eye-view and 20X phase contrast images are shown in the top and bottom panels, respectively. Please click here to view a larger version of this figure.
Figure 2: Representative image of the micro cell scraper pulled from a Pasteur pipette. Please click here to view a larger version of this figure.
Figure 3: Validation of the mature RPE status of differentiated hPSC-RPE cells by immunoblotting. Blots showing the expression of RPE-specific protein markers BEST1, RPE65, and CRALBP in WT hPSC and hPSC-RPE cells. Please click here to view a larger version of this figure.
Figure 4: Immunostaining of BEST1 in WT hPSC-RPEs. (A) Representative immunofluorescent images showing the membrane localization of WT BEST1 in cobblestone-shaped hPSC-RPE cells. (B) Immunostaining negative control omitting the BEST1 primary antibody.
Figure 5: Whole-cell patch clamp recordings of hPSC-RPEs. (A) A single hPSC-RPE cell for whole-cell patch clamp. (B) Representative current traces recorded from a BEST1 WT hPSC-RPE and a BEST1 mutated hPSC-RPE at peak Ca2+. The voltage protocol used to elicit currents is shown in the Insert. Scale bar = 1 nA, 150 ms. See Tables 3 and Table 4 for patch clamp preparation details. Please click here to view a larger version of this figure.
Reagent | Amount |
Knock-Out (KO) DMEM | 500 mL |
KO serum replacement | 15% (75 mL) |
Nonessential amino acids | 1% (5 mL) |
Glutamine | 1% (5 mL) |
Penicillin-streptomycin | 1% (5 mL) |
Nicotinamide | 10 mM |
Human activin-A* | 100 ng/mL |
*Human activin-A is supplemented during days 15–28 of the differentiation protocol. |
Table 1: RPE Differentiation Medium.
Reagent | Amount |
MEM (α modification) | 500 mL |
Fetal Bovine Serum | 5% (25 mL) |
N1 supplement | 1% (5 mL) |
Glutamine-penicillin-streptomycin | 1% (5 mL) |
Nonessential amino acids | 1% (5 mL) |
Taurine | 125 mg |
Hydrocortisone | 10 µg |
Triiodo-thyronin | 0.0065 µg |
Table 2: RPE Culture Medium.
Reagent | Concentration |
CsCl | 130 mM |
MgCl2 | 1 mM |
EGTA | 10 mM |
Magnesium ATP | 2 mM |
HEPES (pH 7.4) | 10 mM |
CaCl2* | Varies |
Glucose** | ~ 5 mM |
*Add CaCl2 to obtain desired Ca2+ concentration | |
** Use glucose to adjust osmolarity to 290–295 mOsm/L |
Table 3: Patch Clamp Internal Solution.
Reagent | Concentration |
NaCl | 140 mM |
KCl | 5 mM |
MgCl2 | 1 mM |
CaCl2 | 2 mM |
HEPES (pH 7.4) | 10 mM |
Glucose* | ~ 5 mM |
*Use glucose to adjust osmolarity to 300–305 mOsm/L. |
Table 4: Patch Clamp External Solution.
The most important procedure for the disease-in-a-dish approach is to differentiate hPSCs with a disease-causing mutation to the correct cell lineage, which is RPE for BEST1. Thus, after each differentiation experiment, the resulting hPSC-RPE cells should be carefully validated for their mature status by RPE-specific morphologies and protein markers16,17,18. To minimize clonal artifact, multiple hiPSC clones from the same patient or multiple hESC clones with the same mutation should be used whenever possible.
Both hiPSC and hESC lines can be differentiated to RPE with the same protocol. hiPSCs with or without a mutation in the BEST1 gene can be reprogrammed from the skin fibroblasts of BEST1 patients or healthy donors, respectively. hESCs bearing a mutation in BEST1 can be genetically engineered from the parental hESC line, which has the wild-type (WT) BEST1. The hiPSC-RPE and hESC-RPE routes have their own advantages and disadvantages. The former is more patient-specific and clinically relevant, particularly to personalized medicine. However, it is worth noting that there are several limitations for the hiPSC-RPE approach: 1) it is logistically difficult to access a full spectrum of BEST1 patient samples, as donations from patients cannot always be granted, and some mutations are very rare in the first place; 2) non-specific phenotypes may result from different genetic backgrounds and physical conditions of the donors; 3) the hiPSC to RPE differentiation efficiency varies for different donors, such that some cases can be technically challenging to obtain hiPSC-RPEs. On the other hand, the hESC-RPE route, although more artificial, offers a versatile platform to generate isogenic hESC-RPEs with desired mutations on demand for non-biased functional investigations.
Mature RPE cells are pigmented, polygonal, and connected by tight junctions in a monolayer. After splitting into a single cell population for patch clamp, the freshly reseeded hPSC-RPE cells will lose the polygonal shape, but still retain pigmentation for several days (Figure 5A). Patch clamp is performed 24–72 h after cell splitting, during which time the pigmentation can still be used as a visible marker to select cells with good RPE status. A gentle cell split resulting in a majority of well-separated single cells without significant cell death is key to the success of the patch clamp.
Several other differentiation protocols using different cell culture media and growth factors have been documented in the literature21,32. The time required for differentiation of hPSC to RPE is similar in all these protocols including ours, while it is hard to tell if there is any significant difference in the differentiation efficiency without a side-by-side comparison. It should be noted that in the current protocol, hPSC-RPE cells are grown in regular flat bottom plates but not on plates with membrane inserts, so the hPSC-RPE generated by this procedure may not best recapitulate the polarity of RPE cells in vivo. Therefore, the techniques described above are mostly suited in a research setting33, although the generated hPSC-RPE cells can also be cultured in Transwell plates to form monolayer RPE sheets for clinical purposes17.
The authors have nothing to disclose.
This project was funded by NIH grants EY025290, GM127652, and University of Rochester start-up funding.
Knock-Out (KO) DMEM | ThermoFisher | 10829018 | |
KO serum replacement | ThermoFisher | 10829028 | |
Nonessential amino acids | ThermoFisher | 11140050 | |
Glutamine | ThermoFisher | 35050061 | |
Penicillin-streptomycin | ThermoFisher | 10378016 | |
Nicotinamide | Sigma-Aldrich | N0636 | |
Human activin-A | PeproTech | 120-14 | |
MEM (a modification) | Sigma-Aldrich | M4526 | |
Fetal Bovine Serum | VWR | 97068-085 | |
N1 supplement | Sigma-Aldrich | N6530 | |
Glutamine-penicillin-streptomycin | Sigma-Aldrich | G1146 | |
Nonessential amino acids | Sigma-Aldrich | M7156 | |
Taurine | Sigma-Aldrich | T0625 | |
Hydrocortisone | Sigma-Aldrich | H0386 | |
Triiodo-thyronin | Sigma-Aldrich | T5516 | |
mTeSR-1 medium | Stemcell Technologies | 5850 | |
Matrigel | Corning | 356230 | |
Collagenase | Gibco | 17104019 | |
Trypsin | VWR | 45000-664 | |
M-PER mammalian protein extraction reagent | Pierce | 78501 | |
proteinase inhibitor cocktail | Sigma-Aldrich | 4693159001 | |
RPE65 antibody | Novus Biologicals | NB100-355 | |
CRALBP antibody | Abcam | ab15051 | |
BEST1 antibody | Novus Biologicals | NB300-164 | |
Beta Actin antibody | ThermoFisher | MA5-15739 | |
Alexa Fluor 488-conjugated donkey anti-mouse IgG | ThermoFisher | A-21202 | |
Goat anti-mouse IgG | ThermoFisher | SA5-35521 | |
Goat anti-Rabbit IgG | LI-COR Biosciences | 926-68071 | |
Hoechst 33342 | ThermoFisher | 62249 | |
HEKA EPC10 patch clamp amplifier | Warner Instruments | 895000 | |
Patchmaster | Warner Instruments | 895040 |