Retinal Organoid

Retinal Organoid is a self-organizing three-dimensional neuroectodermal microtissue differentiated in vitro from embryonic or induced pluripotent stem cells that recapitulates embryonic optic-vesicle evagination, bilayered optic-cup invagination, and chronological lamination of all major vertebrate retinal cell classes—culminating in outer-nuclear-layer rod and cone photoreceptors capable of opsin phototransduction and ribbon-synapse signaling. Because rodents lack a cone-rich macula and diverge from primates in photoreceptor splicing isoforms, human retinal organoids serve as the central experimental and preclinical platform for macular development, adeno-associated virus (AAV) vector screening, inherited retinal dystrophy modeling, and clinical photoreceptor transplantation.

Four-Phase Evagination and Invagination Without External Molds

For over a century after Hans Spemann's 1901 amphibian lens-induction studies, embryologists debated whether folding of the embryonic optic vesicle into a two-walled optic cup requires external mechanical pushing from the overlying lens ectoderm and periocular mesenchyme. In 2011, Mototsugu Eiraku, Yoshiki Sasai, and colleagues resolved this question in mouse embryonic stem cells—and extended it to human ESCs with Tokushige Nakano in 2012—showing that floating three-dimensional aggregates supplemented with low concentrations of laminin-entactin extracellular matrix (Matrigel, 1–2% v/v) spontaneously form two-walled optic cups without lens ectoderm, cornea, or external molds.

Live multiphoton microscopy and laser-ablation biophysics demonstrated that RAX⁺/SIX3⁺/PAX6⁺ diencephalic neuroepithelium executes a four-phase relaxation–expansion mechanical program:

1. Hemispherical Evagination: An eye-field domain (RAX⁺/LHX2⁺) buds outward as a single-layered optic vesicle bounded by basal lamina.
2. Distal Apical Myosin Relaxation: Proximal cells acquire MITF⁺/OTX2⁺ retinal pigment epithelium (RPE) identity under Wnt/BMP cues and retain high Rho-associated kinase (ROCK)-dependent phosphorylated myosin light chain stiffness apically, whereas distal VSX2 (CHX10)⁺/FGF8⁺ future neural retina downregulates apical ROCK–myosin II tension and flattens.
3. Boundary Hinge Apical Constriction: Cells at the circular RPE–neural retina junction undergo apical constriction into wedge-shaped hinge cells, bending the distal sheet inward.
4. Tangential Proliferation Buckling: Tangential division across the pliable VSX2⁺ neural retina pushes against the stiff MITF⁺ RPE rim, driving inward invagination into a goblet-shaped bilayered optic cup (~500–600 μm across in humans, roughly twice the mouse optic cup diameter).

Over 180 to 280 days in vitro—tracking human fetal gestation—multipotent RAX⁺/PAX6⁺/VSX2⁺ retinal progenitors derived from human ESCs or induced pluripotent stem cells (adapted to Matrigel-free floating neurospheres by Xiufeng Zhong, M. Valeria Canto-Soler, and colleagues in 2014) exit mitosis in two conserved waves to construct a five-layered retina:

• Early Neurogenic Wave (Weeks 5–14): Produces retinal ganglion cells (RGCs) (POU4F2⁺/ISL1⁺) firing Na_V-mediated action potentials in the inner ganglion cell layer, followed by ONECUT1/2⁺ horizontal interneurons, THRB⁺/RXRG⁺ cone photoreceptors (short-wavelength OPN1SW S-cones preceding OPN1LW/OPN1MW L/M-cones under thyroid hormone T₃ control), and TFAP2A⁺/GAD1⁺ amacrine interneurons.
• Late Neurogenic and Gliogenic Wave (Weeks 14–38): Driven by OTX2, CRX, and the rod factor NRL (NR2E3 co-activator), progenitors generate abundant rod photoreceptors (RHO⁺), VSX2⁺/PRKCA⁺/GRM6⁺ ON and OFF bipolar interneurons, and RLBP1⁺ Müller glia forming ZO-1⁺ adherens junctions at the outer limiting membrane (OLM).

Maturing retinal organoids supplemented with 9-cis-retinal chromophore, taurine, and docosahexaenoic acid project an ARL13B⁺/CEP290⁺ 9+0 connecting cilium budding stacked outer-segment discs across the exterior perimeter. In 2020, Cameron S. Cowan, Botond Roska, and colleagues benchmarked single-cell transcriptomes across 38 weeks of organoid development against adult human peripheral and foveal retinas, confirming cell-type convergence and functional synapses containing RIBEYE/CtBP2 synaptic ribbons apposed to GRM6 (mGluR6)–TRPM1 ON-bipolar dendrites.

In 2022, Akanksha Saha, David M. Gamm, Raunak Sinha, and colleagues obtained whole-cell patch-clamp recordings from cones in 200-to-260-day human retinal organoids. In darkness, cyclic GMP (cGMP) holds CNGA3/CNGB3 (cone) and CNGA1/CNGB1 (rod) channels open, sustaining a depolarizing inward Na⁺/Ca²⁺ "dark current" (V_m ≈ −40 mV). Illumination photoisomerizes retinal to activate transducin (G_t) and phosphodiesterase-6 (PDE6), hydrolyzing cGMP to close CNG channels and hyperpolarize cones by 15–25 mV with sub-second recovery, Weber–Fechner adaptation, and spectral tuning matching macaque foveal cones.

Ciliopathy Disease Modeling and Antisense / CRISPR Rescue

Because >60% of monogenic retinitis pigmentosa (RP) and Leber congenital amaurosis (LCA) mutations disrupt primate-specific photoreceptor splicing or ciliary transport, patient-derived retinal organoids provide decisive human rescue readouts:

• CEP290 LCA10 (c.2991+1655A>G): Childhood LCA10 introduces a human-specific cryptic splice donor in CEP290 intron 26 inserting a premature stop exon selectively inside human photoreceptors—absent in knock-in mice. Patient organoids validated both sepofarsen antisense oligonucleotides and dual-gRNA CRISPR–Cas9 gene editing restoring full-length CEP290 and ciliogenesis.
• RPGR X-Linked Retinitis Pigmentosa and USH2A Usher Syndrome: Patient organoids carrying RPGR ORF15 frameshifts display opsin mislocalization, shortened cilia, and photoreceptor loss rescued by AAV-RPGR augmentation or CRISPR repair.

AAV Capsid Engineering and Clinical Photoreceptor Sheet Grafting

Because AAV tropism across primate limiting membranes diverges from rodent retina, Roska and colleagues screen barcoded AAV capsid libraries (AAV7m8, AAV-fovea) and synthetic promoters directly in live human retinal organoids. Where host photoreceptors have perished—leaving inner bipolar and ganglion neurons as targets for biological repair or bioelectronic visual prostheses—clinical GMP human iPSC-derived retinal organoid sheets (Masayo Takahashi and Michiko Mandai) entered human trials for advanced retinitis pigmentosa, showing multi-year subretinal graft survival without tumorigenesis.

Why do photoreceptors face the exterior surface of most retinal organoids?

In floating suspension protocols optimized for long-term culture without continuous RPE envelopment, the outer limiting membrane orients outward toward oxygenated medium. This outward brush border enhances nutrient and 9-cis-retinal delivery to metabolically active inner and outer segments and allows direct patch-clamp electrode access to single cones and rods.

Why do retinal ganglion cells decline at late stages in isolated retinal organoids?

In vivo, retinal ganglion cell axons project down the optic nerve to the lateral geniculate nucleus and superior colliculus to receive target-derived BDNF. Lacking brain targets and deep vascular perfusion in vitro, organoid RGCs undergo trophic-deprivation apoptosis after 16–20 weeks unless rescued by assembly into retino-thalamo-cortical assembloids.

Why is exogenous 9-cis-retinal supplied before recording organoid light responses?

In the intact eye, photoisomerized all-trans-retinal is recycled back to 11-cis-retinal by RPE65⁺ retinal pigment epithelium and Müller glia. Because mature neural retinal organoids lack a continuous apposed RPE monolayer, supplying 9-cis-retinal or 11-cis-retinal regenerates light-sensitive opsin visual pigment for electrophysiology.

1. Eiraku, M., Takata, N., Ishibashi, H., Kawada, M., Sakakura, E., Okuda, S., ... & Sasai, Y. (2011). Self-organizing optic-cup morphogenesis in three-dimensional culture. Nature, 472(7341), 51–56. https://doi.org/10.1038/nature09941
2. Nakano, T., Ando, S., Takata, N., Kawada, M., Muguruma, K., Sekiguchi, K., ... & Sasai, Y. (2012). Self-formation of optic cups and storable stratified neural retina from human ESCs. Cell Stem Cell, 10(6), 771–785. https://doi.org/10.1016/j.stem.2012.05.009
3. Zhong, X., Gutierrez, C., Xue, T., Hampton, C., Vergara, M. N., Cao, L.-H., ... & Canto-Soler, M. V. (2014). Generation of three-dimensional retinal tissue with functional photoreceptors from human iPSCs. Nature Communications, 5, 4047. https://doi.org/10.1038/ncomms5047
4. Cowan, C. S., Renner, M., De Gennaro, M., Gross-Scherf, B., Goldblum, D., Hou, Y., ... & Roska, B. (2020). Cell types of the human retina and its organoids at single-cell resolution. Cell, 182(6), 1623–1640. https://doi.org/10.1016/j.cell.2020.08.013
5. Saha, A., Capowski, E., Fernandez Zepeda, M. A., Nelson, E. C., Gamm, D. M., & Sinha, R. (2022). Cone photoreceptors in human stem cell-derived retinal organoids demonstrate intrinsic light responses that mimic those of primate fovea. Cell Stem Cell, 29(3), 460–471. https://doi.org/10.1016/j.stem.2022.02.003