Drug-induced retinal toxicity remains a significant challenge in ocular drug development. Detecting adverse retinal effects early is essential for selecting safer therapeutic candidates and supporting informed decision-making throughout discovery and preclinical development. The retina is a highly organized neural tissue composed of multiple interacting cell types, making the assessment of retinal safety particularly complex.
A range of experimental models contribute to retinal research and drug development, including in vitro systems, animal models, and human tissue-based approaches. Human induced pluripotent stem cell (iPSC)-derived retinal organoids complement these established models by providing access to human retinal biology within a controlled and physiologically relevant in vitro system.
Modeling retinal toxicity in a physiologically relevant system
Human iPSC-derived retinal organoids recapitulate key features of retinal development and contain the major cell types found in the neural retina, including photoreceptors, retinal ganglion cells, bipolar cells, amacrine cells, horizontal cells, and Müller glia (Figure 1). These cells are organized within stratified retinal layers, providing a three-dimensional tissue architecture that closely resembles the native retina (Figure 2).
A key advantage of retinal organoids is their predictable developmental progression. Retinal ganglion cells emerge at early stages of differentiation, around day 60, while a laminated retinal architecture is established by approximately day 150. From day 180 onwards, photoreceptors and bipolar cells become increasingly mature (Figure 3), enabling researchers to select organoids at developmental stages most appropriate for a particular application, including retinal toxicity screening.
Figure 1. Phenotypic characterization: expression of retinal organoid markers. iPSC-derived retinal organoids contain the major retinal cell types organized in a laminar structure, as shown by the expression of Recoverin (photoreceptors), Rhodopsin (rods), Opsin LW/MW and SW (cones), PKC-α (bipolar cells), AP-2α (amacrine cells), SNCG (ganglion cells), CRALBP (Müller glial cells), and PROX1 (horizontal cells).
Figure 2. Mature retinal organoids exhibit a characteristic phase-bright outer rim and elongated outer segment-like structures (white arrow).
Figure 3. Developmental progression of human iPSC-derived organoids along with the abundance of retinal cell types throughout the differentiation timeline.
This cellular complexity is particularly important for toxicity assessment, as retinal toxicants often affect specific retinal cell populations rather than the entire tissue. By incorporating multiple retinal cell types within a structured microenvironment, retinal organoids provide a physiologically relevant context for studying compound-induced retinal injury.
Retinal toxicity assay readouts
Retinal toxicity rarely manifests as a single endpoint. Exposure to a compound may induce early metabolic dysfunction, alter tissue morphology, trigger apoptosis, or affect specific retinal cell populations.
For this reason, retinal organoid studies often combine multiple complementary assays to generate a more comprehensive toxicity profile.
Common retinal toxicity readouts include:
- ATP-based cell viability assays
- LDH cytotoxicity assays
- Brightfield morphological assessment
- Immunofluorescence analysis of retinal markers
- TUNEL staining for apoptosis
- Flow cytometry-based viability measurements
- Gene expression analysis
- Short- and long-term treatment studies
Combining these approaches enables researchers to characterise both the extent of toxicity and the biological mechanisms underlying retinal damage.
Retinal organoids respond to known retinal toxicants
The value of any toxicology model depends on its ability to distinguish between toxic compounds and those that do not induce retinal injury. Validation studies have demonstrated that human retinal organoids respond reproducibly to compounds with established retinal toxicity profiles, supporting their use in preclinical safety evaluation.
Figure 4. Retinal toxicology assessment using human iPSC-derived retinal organoids from Axol Bioscience. Known retinal toxicants, including thioridazine and 4-hydroxytamoxifen, induced dose-dependent reductions in organoid viability, while ketorolac tromethamine and triamcinolone acetonide had no measurable effect on viability under the tested conditions. Viability was assessed using an ATP-based luminescence assay. Cytotoxic compounds also induced morphological alterations, including organoid darkening and irregular edges, observable by brightfield imaging.
As shown in Figure 4, treatment with thioridazine and 4-hydroxytamoxifen resulted in clear dose-dependent reductions in retinal organoid viability. Morphological alterations accompanied these changes, consistent with compound-induced toxicity. In contrast, ketorolac tromethamine and triamcinolone acetonide did not reduce viability across the concentrations tested. Together, these findings demonstrate the ability of retinal organoids to discriminate between compounds with different retinal safety profiles.
Retinal organoids also demonstrate permeability to small molecules. Studies using the naturally fluorescent compound doxorubicin have demonstrated its progressive penetration throughout the organoid structure over time. This characteristic enables compounds to reach internal retinal cell populations and supports the use of retinal organoids as a practical platform for retinal toxicity screening.
Advancing preclinical retinal safety testing
As ophthalmic drug development increasingly focuses on advanced therapeutics, including biologics, gene therapies, and precision medicines, there is a growing need for predictive model systems that can generate physiologically relevant data earlier in development.
Human iPSC-derived retinal organoids provide a scalable platform that combines multicellular retinal architecture with compatibility across a broad range of molecular, imaging, and functional assays. Their ability to recapitulate human retinal responses makes them valuable tools for investigating safety liabilities, understanding mechanisms of toxicity, and supporting compound selection during preclinical development.
At Axol Bioscience, retinal organoid models support mechanistic studies, toxicity assessment, and therapeutic evaluation across a range of ocular research applications. By generating physiologically relevant safety data earlier in the discovery process, these models are helping researchers gain deeper insight into compound safety and advance the development of safer retinal therapies.
References
- Dorgau B, Georgiou M, Chaudhary A, Moya-Molina M, Collin J, Queen R, Hilgen G, Davey T, Hewitt P, Schmitt M, et al. Human retinal organoids provide a suitable tool for toxicological investigations: a comprehensive validation using drugs and compounds affecting the retina. Stem Cells Translational Medicine. 2022;11(2):159-177. doi:10.1093/stcltm/szab010. [academic.oup.com], [pubmed.ncb…lm.nih.gov]
- Hallam D, Hilgen G, Dorgau B, Zhu L, Yu M, Bojic S, Hewitt P, Schmitt M, Kustermann S, Steel D, et al. Human-induced pluripotent stem cells generate light responsive retinal organoids with variable and nutrient-dependent efficiency. Stem Cells. 2018;36(10):1535-1551. doi:10.1002/stem.2883. [ncbi.nlm.nih.gov]
- McClements ME, Steward H, Atkin W, Goode EA, Gándara C, Chichagova V, MacLaren RE. Tropism of AAV vectors in photoreceptor-like cells of human iPSC-derived retinal organoids. Translational Vision Science & Technology. 2022;11(4):3. doi:10.1167/tvst.11.4.3. [tvst.arvoj…urnals.org], [newcellsbi…tech.co.uk]
- Ashworth KE, Weisbrod J, Ballios BG. Inherited retinal diseases and retinal organoids as preclinical cell models for inherited retinal disease research. Genes. 2024;15(6):705. doi:10.3390/genes15060705. [pubmed.ncb…lm.nih.gov], [mdpi.com]
- Lee YJ, Jo DH. Retinal organoids from induced pluripotent stem cells of patients with inherited retinal diseases: a systematic review. Stem Cell Reviews and Reports. 2025;21(1):167-197. doi:10.1007/s12015-024-10802-7.
