Retinal organoids derived from rhesus macaque iPSCs undergo accelerated differentiation compared to human stem cells

Abstract Purpose To compare the timing and efficiency of the development of Macaca mulatta, a nonhuman primate (NHP), induced pluripotent stem cell (rhiPSC) derived retinal organoids to those derived from human embryonic stem cells (hESCs). Results Generation of retinal organoids was achieved from both human and several NHP pluripotent stem cell lines. All rhiPSC lines resulted in retinal differentiation with the formation of optic vesicle‐like structures similar to what has been observed in hESC retinal organoids. NHP retinal organoids had laminated structure and were composed of mature retinal cell types including cone and rod photoreceptors. Single‐cell RNA sequencing was conducted at two time points; this allowed identification of cell types and developmental trajectory characterization of the developing organoids. Important differences between rhesus and human cells were measured regarding the timing and efficiency of retinal organoid differentiation. While the culture of NHP‐derived iPSCs is relatively difficult compared to that of human stem cells, the generation of retinal organoids from NHP iPSCs is feasible and may be less time‐consuming due to an intrinsically faster timing of retinal differentiation. Conclusions Retinal organoids produced from rhesus monkey iPSCs using established protocols differentiate through the stages of organoid development faster than those derived from human stem cells. The production of NHP retinal organoids may be advantageous to reduce experimental time for basic biology studies in retinogenesis as well as for preclinical trials in NHPs studying retinal allograft transplantation.


| INTRODUC TI ON
High-acuity vision in primates is attributed to the development of the macula lutea (macula)-a region with a high density of cone photoreceptors and specialized circuitry. The center of the macula is the fovea centralis (fovea), which is devoid of retinal vasculature and has a characteristic depression, the foveal pit, which is com- have demonstrated promise in rodents, [3][4][5] felines, 6 and nonhuman primates. [7][8][9] However, transplantation experiments to the mammalian retina with a macula are relatively few, and successful vision restoration has been limited. [10][11][12] The translation to clinically relevant treatments relies on the ability to accurately model these therapeutic approaches in a context similar to human patients. While rodents, felids, and canids provide important data on safety and efficacy, they lack true foveal architecture, and therefore, they are not a complete model for the complexity of human macular diseases. It is important to demonstrate the safety and efficacy of these treatments in preclinical trials. Perhaps the best models, we have for preclinical macular studies are nonhuman primates. Nonhuman primates (NHPs) are highly genetically, anatomically, physiologically, and behaviorally similar to humans, and they represent excellent models for translational research when other animal models do not recapitulate the disease of interest.
Rhesus macaques (Macaca mulatta) are one of the most commonly used NHPs in biomedical research. We have recently described rhesus macaques with inherited retinal disease, which may benefit from transplantation of retinal tissue to restore macular function. 13 Others have also described rhesus macaques with macular abnormalities. [14][15][16] As the number of NHP models of IRDs expand, these resources can be useful to study cell transplantation in the context of pre-existing retinal disease. Examples of allogeneic cell transplantation are limited. 17 Exploration of transplantation of allogeneic retinal cells is necessary to overcome the barriers to human clinical studies. Therefore, we have sought to differentiate induced pluripotent stem cells (iPSCs) derived from rhesus monkeys to generate three-dimensional retinal organoids for transplantation into animals with IRDs.
Similar to human stem cells, rhesus macaque iPSCs (rhiPSC) are characteristically indistinguishable from human iPSCs. A previous study using cynomolgus macaque (Macaca fascicularis) derived embryonic stem cells demonstrated that the production of retinal cell types from NHP iPSCs is possible using a two-dimensional differentiation protocol. 18 However, while rhiPSCs have been used for the derivation of blood products 19 and cardiac cells, [20][21][22] little is known about their ability to produce three-dimensional retinal organoids. In this study, we demonstrate the ability of rhiPSCs to produce 3D retinal organoids using a composite established retinal differentiation protocol. [23][24][25] We characterize the development of rhiPSC-derived 3D retinal organoids and their composition with immunocytochemistry and single-cell next-generation sequencing. To determine the reproducibility of the differentiation protocol, we used three rhesus iPSC lines: two established rhiPSC89 26 and rhiPSC90 27 ; and a novel line (from the lab of James Thomson), rhiPSC2431, that is in a defined and xeno-free culture system. Our results demonstrate that rhiPSC-derived retinal organoids follow an abbreviated, but similar development to human iPSC-derived retinal organoids that is consistent with the shorter gestational period of rhesus macaques.

| rhiPSC2431 derivation and stem cell culture
Rhesus macaque fibroblasts were generated from skin punches.

Significance Statement
We demonstrate the generation of retinal organoids from rhesus macaque pluripotent stem cells develops faster than those derived from human stem cells. Usage of rhesus macaque retinal organoids can reduce experimental time for basic biology studies in retinogenesis as well as for preclinical trials in NHPs studying retinal allograft transplantation.

| Retinal differentiation
Spontaneously differentiated cells or iPSC colonies with unhealthy morphology were scraped off before the beginning of differentiation. iPSC colonies were briefly treated with dispase GlutaMAX (Thermo Fisher), and 2-mg/ml heparin (Sigma)) were added, and the iPSC colonies were scraped off and triturated.
The fragmented colonies were then plated in low adhesion plates (coated with polyHEMA (8 mg/cm 2 )), and corresponding iPSC media were added to achieve 3:1 PSC media/neural induction media (NIM). The embryoid bodies (EBs) were sequentially weaned into neural induction media with 50% media changes replaced by fresh NIM. On day 5, a 50% NIM change was done, and on day 6, the EBs were treated with 1.5-nM recombinant human BMP4 (PeproTech, 10-05ET) in fresh NIM. On day 7, the EBs were plated on reduced growth factor Matrigel (50 EBs/well of a 6-well plate).

| Immunocytochemistry
Retinal organoids were fixed with 4% PFA on ice for 20 min and then washed three times with DPBS. The retinal organoids were then equilibrated in 15% sucrose and 30% sucrose until they sunk.
The retinal organoids were then flash frozen in dry ice-ethanol bath in Tissue-Plus OCT compound. Subsequently, the blocks were sectioned at 10 µm and blocked in blocking solution (4% BSA and 0.5% Triton X-100 in PBS) for an hour at room temperature. Primary antibodies were then incubated at 4°C overnight. Excess primary antibody was washed off with three washes of PBS. Alexa Fluorconjugated secondary antibodies matching the primary antibody host were incubated for an hour followed by 5-min incubation in DAPI. The slides were then washed three times, one minute per wash. Finally, the slides were cover slipped with FluorSave Reagent (Millipore, 345789). A table of primary and secondary antibodies is included in Figure S2. The samples were then imaged using an Olympus FluoView FV1000 spectral confocal microscope.

| Single-cell sequencing and data analysis
Single-cell cDNA library preparation and sequencing were performed following manufacturer's protocols (10× genomics). Singlecell suspension at 1000 cells/µl in PBS were loaded on a chromium controller to generate single-cell GEMS (Gel Beads-In-EMulsions).
The scRNA-Seq library was prepared with chromium single-cell 3' reagent kit v3 (10× Genomics). Cell Ranger software v3.1 (https:// www.10xge nomics.com) with default settings was used for alignment, barcode assignment, and UMI counting of the raw sequencing data with genome reference hg19. After generating UMI count profile, we applied Seurat 4.0 (https://satij alab.org/seurat) for quality control and downstream analysis.
For quality control, we excluded genes detected in less than 3 cells, and cells were filtered out if UMI counts are less than bottom 3% and greater than top 1% of total quantile. We removed cellcycle effects by regressing out cell-cycle scores during data scaling using of all signals associated with cell cycle using "CellCycleScoring" function in Seurat. Next, a global-scaling normalization method 'LogNormalize' in Seurat was employed to normalize the gene expression measurements for each cell by the total expression, then the result is multiplied by a scale factor (10,000 by default) and log-transformed. We selected variable genes and computed prin-

| Rhesus iPSCs differentiate into 3D retinal organoids
In this study, we differentiated three different rhiPSC lines using a stepwise retinal differentiation protocol to generate retinal organoids ( Figure 1A). Undifferentiated stem cell colonies were cultured in nonadherent conditions to generate embryoid bodies (EBs), which were cultured in neural induction media (NIM) for 7 days. Next, the EBs were plated onto Matrigel-coated plates and grown in 2D from day 7 to 30 at which time, the regions of the culture displaying retinal morphology were selected, lifted, and grown in nonadherent 3D conditions ( Figure 1B). The morphology of the rhiPSC-derived cultures developing retinal tissue was very similar to that observed in H9 hESCs. Following the criteria proposed in Capowski et al. 28 we binned retinal organoids into three stages: stage 1, retinal organoids displaying phase bright appearance; stage 2, retinal organoids displaying phase dark appearance; and stage 3 retinal organoids displaying outer segment protrusions ( Figure 1C).
During the 2D-retinal differentiation protocol, we noticed that the self-organizing optic vesicle-like structures (OVs) developed earlier and were well formed as early as day 20 ( Figure 1B, Day 20).
OV formation was not mutually exclusive to formation of neural rosettes 3 or horseshoe-like structures described by others. 29 OVs continued to develop until day 30, at which point they were manually selected from the plate and resuspended in 3D-RDM media supplemented with retinoic acid (RA) until day 100. Thereafter, the retinal organoids were cultured in 3D-RDM until analysis.  Lower power images are shown in Figure S3.  Figure 4H-K). We did not observe BRN3-positive cells ( Figure 4I).

| Rhesus retinal organoids express retinal celltype-specific markers
However, there was some TUJ1-positive signal ( Figure 4J,K) that we presumed to be neurites of degenerating RGCs that no longer labeled with BRN3 as the nuclei were probably apoptotic. Lower power images are shown in Figure S4.  Figure 5R) and have pedicles that were localized with CTBP2 ( Figure 5S).
The inner layer was composed of both CHX10-positive bipolar cells ( Figure 5N) and SOX9-positive Muller Glia ( Figure 5X). The inner layer was somewhat less stratified and organized than the outer layer of the organoid at this stage. As we observed in stage 2, there were no longer any retinal ganglion cells (BRN3/TUJ11 double-positive cells), but some TUJ1-positive neurites persisted ( Figure 5U-Y). Similarly, photoreceptors previously observed in the inner aspect of the organoid, likely representing newly born photoreceptors undergoing radial somal migration, were for the most part absent. However, occasionally, they arranged themselves into rosettes with outer segments pointing toward the center ( Figure 5P'-T'). Lower power images are shown in Figure S5.

| Single-cell RNA-seq showed stageappropriate retinal cell types in rhesus retinal organoids
In order to determine the cell types being produced in rhesus retinal organoids, we used single-cell RNA sequencing (scRNA-seq). organoids ( Figure S6) as well as violin and feature plots of known retinal cell-type-specific markers ( Figure S7) was also assessed.

| Rhesus stem cells differentiate precociously and less efficiently compared with human stem cells
We used light microscopy to systematically assign morphological staging levels to retinal organoids based on criteria defined by Capowski et al. 28 In general, stage 1 retinal organoids developed  Figure 7B). Similarly, the proportion of embryoid bodies, which adopted retinal morphology was 26.1% in human stem cells, but only 1.4%-4.7% in rhesus iPSCs ( Figure 7C).
We attempted to modify the differentiation protocol, in particular by abbreviating the early steps to better suit the intrinsic timing of rhesus macaque iPSC differentiation ( Figure S8). However, our attempts did not yield better results.

| DISCUSS ION
Retinal research in NHPs is of particular importance due to the presence of macular and foveal structure, which is required for high-  Rhesus iPSC lines are more difficult to culture and may require feeder layers of MEFs since they have a tendency to differentiate prematurely in their absence. The difficulty in maintaining them may outweigh the potential savings in the shorter retinal differentiation time. Optimization of a feeder-free system that is similar to that of human PSCs in an mTESR1 system would be advantageous.
Another consideration to take into account is that the differentiation protocol will have to be modified to increase the efficiency of retinal organoid production from rhesus iPSCs. Optimization of the retinal differentiation protocol for rhesus iPSCs is the subject of continued research.

ACK N OWLED G EM ENTS
The authors thank Amander Clark and James A. Thomson for their laboratory and sharing their iPSC lines, Deepak Lamba and Shereen Chew for help discerning details of retinal organoid differentiation, and David Gamm, Adam Miltner, and Robert J. Johnston for helpful discussions, and the technical assistance of Brad Shibata in the NEI Microscopy Core (P30EY12576) led by Dr. Paul FitzGerald.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest related to this study.

AUTH O R CO NTR I B UTI O N S
Antonio Jacobo Lopez was involved in conception and design, collection and/or assembly of data, data analysis and interpretation, and manuscript writing. Sangbae Kim was involved in collection and/ or assembly of data, data analysis and interpretation, and manuscript writing. Xinye Qian was involved in collection and/or assembly of data, data analysis and interpretation. Jeffrey Rogers, J. Timothy Stout, Sara M. Thomasy, and Anna La Torre were involved in final approval of the manuscript. Rui Chen was involved in conception and design, financial support, provision of study material, collection and/or assembly of data, data analysis and interpretation, and final approval of the manuscript. Ala Moshiri was involved in conception and design, financial support, provision of study material, collection and/or assembly of data, data analysis and interpretation, manuscript writing, and final approval of the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that supports the findings of this study are available in the supplementary material of this article.