We describe a method to engineer a retinal tissue composed of retinal pigment epithelial cells derived from human pluripotent stem cells cultured on top of human amniotic membranes and its preparation for grafting in animal models.
Several pathological conditions of the eye affect the functionality and/or the survival of the retinal pigment epithelium (RPE). These include some forms of retinitis pigmentosa (RP) and age-related macular degeneration (AMD). Cell therapy is one of the most promising therapeutic strategies proposed to cure these diseases, with already encouraging preliminary results in humans. However, the method of preparation of the graft has a significant impact on its functional outcomes in vivo. Indeed, RPE cells grafted as a cell suspension are less functional than the same cells transplanted as a retinal tissue. Herein, we describe a simple and reproducible method to engineer RPE tissue and its preparation for an in vivo implantation. RPE cells derived from human pluripotent stem cells are seeded on a biological support, the human amniotic membrane (hAM). Compared to artificial scaffolds, this support has the advantage of having a basement membrane that is close to the Bruch's membrane where endogenous RPE cells are attached. However, its manipulation is not easy, and we developed several strategies for its proper culturing and preparation for grafting in vivo.
RPE is crucial for the survival and homeostasis of the photoreceptors with which it is tightly associated1. Several pathological conditions alter its functionality and/or survival, including RP and AMD.
RP is a group of inherited monogenic mutations that affect the functions of photoreceptors or RPE cells or both2,3. It is estimated that mutations that affect specifically the RPE cells account for 5% of RP2. AMD is another condition where the RPE layer is altered, leading ultimately to central vision loss. AMD is caused by the complex interactions of genetic and environmental factors and affects the elderly4,5,6. According to projections, AMD will be a concern for 196 million patients worldwide by 20207. For these disorders, no effective cure exists, and one of the strategies proposed is the transplantation of new RPE cells in order to compensate for dead/nonfunctional preexisting RPE cells8.
The mode of formulation of the final product to be grafted is essential to ensure the best functional outcomes. RPE cells injected as a cell suspension, despite being an easy and straightforward method of delivery, raise concerns regarding their survival, integration, and functionality9,10,11,12,13. Scientists are now developing more complex formulations to deliver engineered retinal tissue9,13,14,15,16. In this context, we developed an original method to generate in vitro RPE tissue that could be used for transplantation9.
RPE cell banks derived from human embryonic stem (ES) cells are used in this protocol. However, alternative RPE cell banks from different cell sources (human-induced pluripotent stem cells, primary RPE cells, etc.) and differentiated with a different method are also suitable for this protocol. It includes directed differentiation protocols using cytokines and/or small molecules17,18,19,20,21,22.
To be transplanted, the engineered tissue should be prepared on a scaffold. In the past few years, different scaffolds were developed based on a polymer or on a matrix of biological origin13,23,24. Here, the biological substrate used is the hAM, but other substrates, like denuded Bruch membranes, could be implemented. The method described herein has the advantage of using a biological scaffold that is more relevant to the RPE native environment.
Human ES cell-derived RPE cells are cultured for at least 4 weeks in order to be fully organized as a cobblestone monolayer. At that stage, the epithelium obtained is functional and polarized9. Finally, as this tissue wrinkles easily, it is embedded in a thin layer of a hydrogel carrier to give it more rigidity and elasticity and to protect it during the injection procedure. This product is then stored at 4 °C until grafting.
All human materials used in this protocol were used in accordance with European Union regulations. The human ES cell line used in this study was derived from a unique embryo. The couple who had donated the embryo was fully informed and gave their consent for an anonymous donation. A clinical-grade human ES cell line was derived from this embryo, banked, qualified, and properly documented by Roslin Cells (UK). hAMs were procured under sterile conditions during a cesarean section in mothers who signed an informed consent for placenta donation according to hospital guidelines (APHP, Hôpital Saint Louis).
1. Preparation of Culture Media and Reagents
2. Thermolysin Treatment of Human Amniotic Membranes
3. Fixation of Human Amniotic Membranes on a Culture Insert
4. Thawing and Seeding of Retinal Pigment Epithelium Cells on Human Amniotic Membranes
5. Maintenance of Retinal Pigment Epithelium Cell Cultures on Human Amniotic Membranes
6. Preparation of the Retinal Pigment Epithelium Patch for Transplantation
NOTE: Starting at day 30 of the culture, the tissue is ready for transplantation.
hAMs contain an epithelial layer that should be removed before the seeding of RPE cells. An enzymatic treatment of the membrane is performed with the thermolysin under shaking. In order not to not lose the polarity of the membrane (the epithelium is on one side), it is fixed on a support which composition could be different depending on the provider (Figure 1A). Check the adhesion of the membrane to its support at this step and add clips if necessary. At the time of the fixation in the culture insert, work carefully to avoid making holes in the membrane and keep its polarity with the basement membrane facing up (Figure 1B). When cells are seeded on the membranes, they could escape from these holes and the final cell concentration will be reduced. The presence of holes could be visualized under a microscope or by adding PBS inside the culture insert and checking if PBS is leaking. Any membrane with holes should be discarded.
Following the fixation on the culture insert, the presence of residual cells in a phase-contrast microscope is evaluated (Figures 1C and 1D). Classically, a few dead cells could remain on the surface of the membrane, but those cells will be eliminated when the culture medium is changed (Figure 1C). The fibers of the basement membrane could be seen at a higher magnification (Figure 1D) if no cells remain. If that is not the case, the timing of the incubation with thermolysin might be adjusted.
In the days following the seeding of the RPE cells, check if the cells adhere. Depending on the microscope used, it could be difficult to clearly see the cells during the first few weeks, but if the cells do not adhere, cells will be seen floating around in the cell culture medium. Once the epithelium is formed and starts to mature, it becomes easier to see it under a microscope (Figures 2A, 2B, and 2C). At 3 weeks, the cells form a complete monolayer epithelium typical of RPE (cobblestone organization). After 4 weeks, the epithelium is enough mature for its preparation for implantation (Figure 2D). However, it could be kept in culture for more weeks for logistical reasons.
The preparation of the graft for implantation is described in Figure 3. Upon apposition of the membrane to the 20% gelatin block, aspirate all the excess cell culture medium. This step is important, as any remaining medium could preclude the adhesion of the 8% gelatin to the RPE and the membrane. Indeed, the medium will form a film of liquid between the layers of gelatin.
The gelatin used for the embedding of the implant could be of a varying strength, depending on its Bloom index (i.e., a quality control test, performed and provided by suppliers, that evaluates the strength of a gel at a standardized temperature and concentration). If the gelatin used is not rigid enough and disaggregates during the grafting, the used gelatin reference should be changed to another one with a higher strength. The concentration of gelatin could also be changed, based on experimentation, to adjust its strength and elasticity.
Figure 1: Representative images of the hAM before and after the thermolysin treatment and fixation on a culture insert. (A) Representative image of a membrane supplied by a tissue bank. (B) Representative image of a membrane fixed on a culture insert. (C) Representative image of a membrane treated with thermolysin at a low magnification. (D) Representative image of a membrane treated with thermolysin at a high magnification. Please click here to view a larger version of this figure.
Figure 2: Representative images of the hAM upon seeding the RPE cells. (A) Representative image of the membrane before seeding.(B and C) Representative images of RPE cells on membranes at 3 weeks post-seeding. The membrane may not be completely planar as seen in panel C. Scale bar = 100 µm (C), 20 µm for the magnification. (D) Representative image of the RPE cells on a membrane at 30 days post-seeding, corresponding to the time of embedding them in gelatin. Please click here to view a larger version of this figure.
Figure 3: Scheme describing the sequential steps for the inclusion of the engineered retinal tissue containing the RPE cell layer on the hAM inside a gelatin film. Please click here to view a larger version of this figure.
We described a method for the culture of RPE cells on a biological scaffold and its preparation for implantation in animal models. One of the critical steps of the protocol is the maintenance of the orientation of the hAM all along the procedure until its inclusion into gelatin. Indeed, the native epithelium of the membrane is removed and its basement membrane becomes exposed9. The RPE cells have to be seeded on top of this basement membrane. Upon preparation for gelatin embedding, it is crucial to work with all the products at the defined temperature. Indeed, gelatin has the property to be rigid at 4 °C and liquid at body temperature (37 °C)9. If the temperature is not respected, the gelatin could solidify or liquefy at a step where this effect is not desired.
Several biological scaffolds have been proposed, like Descemet's membranes25 or hAMs26. In particular, hAMs, from a cesarean section27, were demonstrated to be well-tolerated in the subretinal space, causing limited inflammation and reducing choroidal neovascularization26. The membrane successfully supports the culture of human RPE cells9,28. Moreover,these membranes have also a long history in clinics29, making them good candidates for a scaffold for RPE cell therapy. Other biological supports could be easily implemented with this protocol. Synthetic scaffolds based on polymer are already rigid and might not need a gelatin embedding prior to implantation13,30,31,32,33. Other systems for implantation have been recently developed for the subretinal delivery in the human eye of an hESC-derived RPE monolayer on a rigid polyester scaffold34 or on a synthetic parylene substrate designed to mimic Bruch's membrane35. Even though these strategies are promising, we believe that biological scaffolds might provide the best platform for RPE tissue engineering.
In this protocol, the seeded RPE cells were derived from human ES cells using a spontaneous differentiation method. However, different types of RPE cells could be used, either differentiated from human induced pluripotent stem cells or from primary RPE cells obtained from cadavers or even from an RPE cell line19,36,37,38. Moreover, if using pluripotent stem cells, several protocols were developed in the past years to obtain RPE cells based on a spontaneous differentiation or using small molecules to guide the differentiation18,22. RPE cells obtained from these different differentiation protocols could be also used with this method for tissue engineering.
One limit of this method is the stability of the embedded tissue at 4 °C. As it is included in gelatin just before the shipment to the surgery site, it has to be maintained at this temperature until the engraftment. In that context, the surgeries should be performed within 48 h.
This method could be easily transferred to the clinic. The RPE tissue embedded in gelatin is conserved at 4 °C in the CO2-independent medium. This medium could be substituted, if required, with others already approved by the US Food and Drug Administration (FDA) for the conservation of corneas obtained post-mortem and which are used for transplantations into humans. We successfully demonstrated the efficiency of this strategy for transplantation in a proof-of-concept study into rodent models9, and we are currently validating the surgical approach in non-human primates before implementing it for a clinical trial to treat patients with specific forms of RP affecting the RPE.
The authors have nothing to disclose.
The authors would like to thank Jérôme Larghero and Valérie Vanneaux (Hôpital Saint Louis, Paris, France) for their input during the setting-up of the method described here.
This work was supported by grants from the ANR [GPiPS: ANR-2010-RFCS005; SightREPAIR: ANR-16-CE17-008-02], the Fondation pour la Recherche Médicale [Bio-engineering program – DBS20140930777] and from LABEX REVIVE [ANR-10-LABX-73] to Olivier Goureau and Christelle Monville. It was supported by NeurATRIS, a translational research infrastructure (Investissements d'Avenir) for biotherapies in Neurosciences [ANR-11-INBS-0011] and INGESTEM, the national infrastructure (Investissements d'Avenir) engineering for pluripotent and differentiated stem cells [ANR-11-INBS-000] to Christelle Monville. Karim Ben M'Barek was supported by fellowships from DIM Stempole and LABEX REVIVE [ANR-10-LABX-73]. I-Stem is part of the Biotherapies Institute for Rare Diseases supported by the Association Française contre les Myopathies (AFM)-Téléthon.
Sterile biosafety cabinet | TechGen International | Not applicable | |
Liquid waste disposal system for aspiration | Vacuubrand | BVC 21 | |
CO2-controlled +37 °C cell incubator | Thermo Electron Corporation | BVC 21 NT | |
200 µL pipette: P200 | Gilson | F144565 | |
1 mL pipette: P1000 | Gilson | F144566 | |
Pipet aid | Drummond | 75001 | |
+4 °C refrigerator | Liebherr | Not applicable | |
Vibratome | Leica | VT1000S | |
Fine scissors | WPI | 501758 | |
Forceps (x2) | WPI | 555227F | |
Water bath | Grant subaqua pro | SUB6 | |
Precision balance | Sartorius | CP225D | |
Centrifuge | Eppendorff | 5804 | |
Microscope | Olympus | SC30 | |
Horizontal Rocking Shaker | IKA-WERKE | IKA MTS 214D | |
Vortex | VWR | LAB DANCER S40 | |
Disposable Scalpel | WPI | 500351 | |
plastic paraffin film | VWR | PM992 | |
0.200 µm single use syringe filter | SARTORIUS | 16532 | |
Syringe without needle 50 mL | Dutscher | 50012 | |
Bottles 250mL | Dutscher | 28024 | |
15 mL sterile Falcon tubes | Dutscher | 352097 | |
50 mL sterile Falcon tubes | Dutscher | 352098 | |
culture insert | Scaffdex | C00001N | |
60 mm cell culture disches: B6 | Dutscher | 353004 | |
12 well cell culture plate | Corning | 3512 | |
6-well culture plates | Corning | 3506 | |
Razor blades | Ted Pella, Inc | 121-9 | |
Cyanoacrylate glue | Castorama | 3178040670105 | |
PBS 1X (500 mL) | Sigma | D8537 | |
Thermolysine | Roche | 5339880001 | |
DMEM, high glucose, GlutaMAX | Invitrogen | 61965-026 | |
KSR CTS (KnockOut SR XenoFree CTS) | Invitrogen | 12618-013 | |
MEM-NEAA (100X) | Invitrogen | 11140-035 | |
b-mercaptoethanol (50 mM) | Invitrogen | 31350-010 | |
Penicillin/Streptomycin | Invitrogen | 15140122 | |
CO2-independent medium | GIBCO | 18045-054 | |
Gelatin | MERCK | 104078 | |
human amniotic membrane | Tissue bank St Louis hospital (Paris, France) | Not applicable |