Improved Protocol for hiPSC-derived Retinal Organoid Generation Increases Yield and Decreases Variability

Abstract

Purpose : Retinal organoids (RO) generated from human induced pluripotent stem cells (hiPSC) constitute important in-vitro tools for retinal disease modeling and therapeutic drug development. However, low organoid yield and variability between batches of RO cultures hinder the use of these models in translational applications that require a high level of reproducibility for quantitative analyses. Likewise, lengthy culture times make the model impractical for drug screening purposes. Therefore, we set out to optimize current protocols to overcome these limitations.

Methods : Two different wild-type hiPSC lines were used for RO generation using the well-established protocol by Zhong et al. (2014). Five different media supplementation regimens of were tested on RO cultures starting at 63 days of differentiation: 1µM retinoic acid (RA), 10µM RA, 1µM 9-cis retinal (9cisRAL), 10µM 9cisRAL, and no supplementation. RO yield and photoreceptor outer segment (OS) length were longitudinally assessed in live organoids from each condition over a 4-month period. Additionally, ROs were collected at various time points, fixed, and their composition was evaluated by immunofluorescent staining and confocal microscopy with image quantification using FIJI software.

Results : Compared to other supplementation conditions, RO cultures that were supplemented with 9cisRAL displayed higher yield, increased OS length, and accelerated photoreceptor differentiation.

Conclusions : In this work we evaluated the performance of 5 different media supplementation regimens on the yield and quality of RO generation, to identify the most robust conditions for quantitative translational applications. We found that even though all tested conditions were able to produce RO with all the different retinal cell types and with photoreceptors displaying outer segment structures, 9cisRAL produced better RO yield than other conditions. Moreover, this condition accelerated OS development by approximately 30 days and produced longer OS.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.