Publication protocol
Generation of embryoid bodies in agarose microwell arrays to regulate size: Agarose microwell arrays were prepared in ‘MicroTissues 3D Petri Dish micro-mold spheroids’ molds (size L, 9 × 9 array; Sigma, #Z764019-6EA) with 500 μl of 2% agarose (Thermo Fisher Scientific, #R0491) in DMEM with GlutaMax (GIBCO, #10569-010). The solidified molds were transferred to 12-well plates (Corning, #3513) and equilibrated with 1.5 mL mTeSR medium. The molds were stable for several weeks at 4°C and were warmed in the incubator at 37°C before use. To seed iPSCs into the agarose microwell arrays, a well of iPSCs was washed twice with 1 mL PBS without CaCl2 and MgCl2 (GIBCO, #14190094), incubated for 5 minutes with 600 μl 0.5 mM EDTA working solution (0.5M EDTA, Invitrogen, #15575020 in PBS without CaCl2 and MgCl2) at 37°C after which EDTA was aspirated. 500 μl of Accutase (Thermo Fisher Scientific, #00-4555-56) were added, the plate incubated for 3 minutes at 37°C, 1 mL mTeSR medium was added and the cells were gently pipetted up and down to generate a single cell suspension for cell counting. 5 mL of mTeSR medium was added, the cells were pelleted for 5 minutes at 200 g, resuspended in the appropriate working volume and cells were seeded at 150 to 3,000 cells per microwell (each microwell array mold contains 81 microwells) in 150 μl mTeSR containing 10 μM ROCK inhibitor Y-27632 (STEMCELL Technologies, #72304) into the top of the agarose mold. The plates were placed in the incubator for 30 minutes to allow the cells to settle into the microwells before gently filling up the well with 1.5 mL mTeSR medium containing 10 μM ROCK inhibitor to completely cover the agarose mold and incubating the forming embryoid bodies at 37°C, 5% CO2. The medium was replaced with NIM on the following schedule: on day one, one third was replaced; on day two, one half was replaced; on day three, all of the medium was replaced. From day four to six embryoid bodies were fed daily with 1.5 mL of NIM and on day 7 embryoid bodies from one well were transferred to a 6-well plate coated with Matrigel (Corning, #356230). Traditionally, the control of embryoid body size has been achieved by seeding many thousand dissociated iPSCs per well in low attachment 96-well plates (Kuwahara et al., 2015; Nakano et al., 2012). However, we achieved optimal embryoid body size by seeding less than 1,000 iPSCs per microwell. Seeding this low number of iPSCs into 96-well plates does not efficiently generate embryoid bodies. Interestingly, we found that the efficacy of organoid production was more sensitive to changes in embryoid body size using IMR90.4 iPSCs than F49B7 iPSCs, suggesting that embryoid body size should be optimized for each line.
For embryoid bodies from both generation methods, on day 16, NIM was exchanged for ‘3:1 medium’ containing 3 parts DMEM (GIBCO, #10569‒010) per 1 part F12 medium (GIBCO, #31765‒027), supplemented with 2% B27 without vitamin A (GIBCO, #12587-010), 1% NEAA Solution, 1% penicillin / streptomycin (GIBCO, #15140‒122). On day 28 – 32 retinal structures were detached from the Matrigel plate by either microdissection (Zhong et al., 2014) or by checkerboard scraping. Microdissection of neuroretinal structures was performed with needles (VWR, #613-3834) held by a syringe (VWR, #613-2997) under an inverted microscope (EVOS XL Core, Thermo Fisher Scientific, #AMEX1000) in a cell culture hood. Retinal structures were cultured in low attachment 3.5-cm dishes (Milian, #351008) in 3 mL medium. For checkerboard scraping, first, to break the tissue sheets into smaller pieces, a 1 to 2 mm2 grid was scratched through the cells on the culture plate with a 10-μl or 200-μl pipette tip. Second, the entire contents of the culture plate were washed off the plate with a 1,000-μl pipette tip to generate numerous retinal aggregates (Video S1) and small uncharacterized debris. During the manual microdissection step in traditional organoid production, many retinal structures could not be identified by the observer and were consequently left behind, hence checkerboard scraping increased the yield of retinal structures that could be harvested. Retinal structures were typically not broken by the pipette passing through the dish, but rather came off the plate as large tissue pieces. The aggregates contained both regions of neural retina and retinal pigment epithelium. To remove debris and single cells, the aggregates were washed 3 × in a 15 mL tube by sedimentation in 3:1 medium and then maintained in suspension on sterile Petri dishes (VWR, #391‒2016) in 10 – 15 mL 3:1 medium with media changes every other day. Aggregates without phase-bright, stratified neuroepithelium indicative of retina formation (Figure S1) were sorted out one week after checkerboard scraping to leave behind only high-quality retinal organoids. Even before sorting, up to 80% of aggregates within the plate contained phase-bright, stratified neuroepithelium. Organoids were not further trimmed to remove non-retinal structures, because the presence of non-retinal tissue did not prevent growth and maturation of retinal tissue within the organoid.
From day 42, aggregates were cultured in 3:1 medium supplemented with an additional 10% heat-inactivated FBS (Millipore, #es‒009‒b) and 100 μM taurine (Sigma, #T0625‒25G) with media changes every other day. At week 10, the culture medium was supplemented with 1 μM retinoic acid (Sigma, #R2625). From week 14, the B27 supplement in 3:1 media was replaced by N2 supplement (GIBCO, #17502‒048) and retinoic acid was reduced to 0.5 μM.
Full paper
Login or
join for free to view the full paper.