Generation of CiPSCs by defined small-molecule compounds
In order to effectively trace the generation of CiPSCs, Oct4 promoter-driven green fluorescent protein (GFP) reporter MEFs (OG-MEFs), a reporting system for pluripotent marker, were employed. According to the previously developed method, it takes about 40 days to induce MEFs into CiPSCs by small-molecule compounds13 (Fig. 1A). As early as day 5 (D5, same abbreviation in below), there were a number of epithelial colonies formed. These colonies rapidly expanded and some of them expressed GFP at day 40, indicating that the cells have acquired pluripotency (Fig. 1B). At this point, these clones have not yet shown typical ESC-like morphology (domed, phase-bright, homogeneous with clear-cut edges) (Fig. 1B). Subsequently, these GFP+ colonies were picked up and seeded on feeder cells to establish CiPSC lines for further characterization and differentiation and defined as P1. These P1 clones maintained GFP expression and ESC-like morphology throughout passages (Fig. 1C, Supplementary Fig. 1A).
A Schematic diagram of the induction of Oct4-GFP-CiPSCs from MEFs. B Morphological changes during the induction of CiPSCs. The inset picture in D40 showed the expression of green fluorescence from Oct4-GFP cells. Scale bars, 400 μm. C Representative bright field and fluorescence images of the CiPSCs. Scale bar, 200 μm. D Karyotypic analysis of the CiPSCs colonies. E Immunostaining of pluripotency markers Oct4, Nanog, Sox2 and SSEA4 in CiPSC colonies. Scale bars, 50 μm. F The qRT-PCR analysis of mRNAs of pluripotency markers in MEFs, mESCs, miPSCs and two independent CiPSCs. Mean ± SEM, N = 3 independent experiments. G H&E staining of CiPSC-derived teratomas shows the presence of various structures from three germ layers. The black arrowhead pointed to the representative tissues of the three germ layers. See also Supplementary Fig. 1.
First, CiPSCs were identified to have normal karyotypes (Fig. 1D). Next, the pluripotent cell markers of CiPSCs were determined at mRNA and protein levels. The pluripotency-associated proteins Oct4, Nanog, Sox2, and SSEA4 could be identified by immunostaining of CiPSCs colonies (Fig. 1E). Meanwhile, the CiPSCs expressed pluripotency-associated transcripts such as Oct4, Nanog, Sox2, Lin28a, and Sall4 and their expression levels were similar to those of mESCs (Fig. 1F). Last, the differentiation potentials of CiPSCs were further examined. In vitro differentiation test showed that CiPSCs could differentiate into different cell types of three germ layers (Supplementary Fig. 1B). As previously described, teratoma formation experiment is the gold standard for verifying whether the cells possess pluripotency, and HE staining is a commonly used method for identifying the germ layers of teratoma12,13. CiPSCs were able to form teratoma after injected into immune-deficient SCID mice, and we could identify representative tissues of the endoderm, mesoderm, and ectoderm in the generated teratoma by HE staining, such as the endoderm: pseudostratified ciliated columnar epithelioid respiratory epithelium, and stratified squamous epithelioid gastrointestinal epithelium, the mesoderm: the large, round and empty adipose tissue, and cylindrical, multinucleated, striated muscle tissue, the ectoderm: blackened melanin tissue, and rich cytoplasm, acidophilic keratinized beads, which confirmed that the injected cells have the ability to differentiate into three germ layers, demonstrating their pluripotency (Fig. 1G). These findings demonstrated that CiPSCs were fully pluripotent as other typical iPSC lines.
Generation of polarized retinal organoids from CiPSCs
We optimized previously reported methods to differentiate 3D ROs (Fig. 2A)4,18. During the formation of embryoid bodies (EBs), the majority of cell aggregates developed neuroepithelium at D5, and a few optic vesicle-like structures appeared at D7 (Fig. 2B). For ROs differentiation, the D7 EBs were transferred into retinal maturation medium, and more optic cup-like structures formed at D10 (Fig. 2B). The optic cups contained retinal progenitors (Rax+, Pax6+, Vsx2+) and retinal pigment epithelium (RPE)progenitors (Otx2+, Rax−) (Fig. 2C). Furthermore, the neural retinae were polarized with the inner side to be basal (Laminin+) and the outer side apical (ZO-1+) (Fig. 2D). When tracing the differentiation of EBs from Oct4-GFP+ CiPSCs, we observed that the GFP+ signal gradually decayed and completely disappeared at D8 (Supplementary Fig. 1C).
A Schematic diagram of the methods for ROs differentiation from CiPSCs. B Morphological changes in early differentiation stages of ROs from CiPSCs. Arrowheads point to the optic vesicle and/or optic cup structures. Scale bars, 200 μm. C Retinal progenitor cells were shown by immunostaining with Rax (red), Pax6 (green), Vsx2 (green) and Otx2 (green). RPE progenitors were shown by Otx2+; Rax– immunostaining. Arrowheads indicate relevant colocalized cells. Scale bar, 50 μm. D Immunostaining of ROs with the apical (ZO-1) and basal (Laminin) markers on D24. Arrowheads show the apical side (ZO-1+) and the basal side (Laminin+) of organoids. Scale bar, 25 μm. E A gradual decline of cell proliferation at distinct timepoints during differentiation. Arrowheads indicate relevant PH3+ (green) and Ki67+ (red) cells. Images were captured using the same confocal settings. Scale bar, 50 μm. F, G Quantitative analysis of PH3+ and Ki67+ cells. Data was represented as means ± SEM. n > 8, **p < 0.01, ****p < 0.0001. See also Supplementary Fig. 1.
To further examine the cell proliferation during differentiation, we performed the immunostaining of organoids for PH3 and Ki67 expression (Fig. 2E). Both PH3+ and Ki67+ cells were gradually decreased from D14 to D21. The proportion of positive cells at D14, 18 and 21 is 9.00 ± 0.73%, 6.56 ± 0.41% and 4.40 ± 0.40% for PH3, and 67.67 ± 3.00%, 53.67 ± 3.01% and 47.13 ± 2.18% for Ki67 (means ± SEM. N > 8), respectively (Fig. 2F, G). Taken together, these results demonstrated that the cells in organoids gradually exited the cell cycles and differentiated into various retinal cell types.
Morphological and molecular characterization of CiPSC-derived retinal organoids
Consistent with previously reported results4,19, in this research, retinal organoids gradually formed big epithelial structures with clear stratification during differentiation, and the thickness of the photoreceptor layer became thinner with time (Fig. 3A). At the late stage of differentiation, outer segment (OS)-like structures were observed at the outer surface of organoids (Fig. 3A), which corresponded to the photoreceptor outer segments in vivo20.
A Morphological changes of ROs at different stages of differentiation. The amplified neural retina of the corresponding days was displayed in the bottom panel. The curves indicate the boundaries of photoreceptors. The polarized neuroretina displayed significant layered structure and the thickness of the photoreceptor became thinner with time. The white arrowhead pointed to the neuroretina structures. The black arrowhead pointed to the outer segment (OS)-like structures. Scale bar, 400 μm (top), 100 μm (bottom). B Immunostaining of Recoverin (photoreceptor precursor cell marker, in green) and Rhodopsin (rod photoreceptor cell marker, in red) in various stages of ROs. Indicated rectangle regions were amplified as inset pictures in the bottom left. Arrowheads indicate Recoverin and Rhodopsin double-positive cells which represent rod photoreceptors. Scale bars, 25 μm (left, for insets), 50 μm (right). Images were captured using the same confocal settings. C Quantitative analysis of Recoverin+ and Rhodopsin+ cells. Data was represented as means ± SEM. N ≥ 8, *p < 0.05, **p < 0.01, ****p < 0.0001. D The qRT-PCR analysis of the dynamic expression of photoreceptor cell-related genes Rcvrn and Rho during organoid differentiation. The average expression value for each gene was further normalized to that of D14 to yield relative expression values (log2 scale). Data was represented as means ± SEM. See also Supplementary Fig. 2.
To investigate the morphology of photoreceptors, we immunostained ROs for photoreceptor markers Recoverin and Rhodopsin. It showed that Recoverin+ and Rhodopsin+ cells were detectable at D21, and both Recoverin+ and Rhodopsin+ cells increased as differentiation proceeded (Fig. 3B). Interestingly, Recoverin and Rhodopsin were located at the apical side of the neural retinal (Fig. 3B), consistent with photoreceptor morphogenesis in vivo. The proportion of positive cells at D18, 21, 25 and 28 was 0.0 ± 0.0%, 14.70 ± 0.73%, 33.56 ± 2.33% and 45.38 ± 1.98% for Recoverin, and 0.0 ± 0.0%, 5.90 ± 0.64%, 8.56 ± 0.73% and 14.22 ± 1.45% (means ± SEM. N > 8) for Rhodopsin, respectively (Fig. 3C), indicating the gradual maturation of the photoreceptors during differentiation. At the transcriptomic level, qRT-PCR analysis of photoreceptor cell-related genes Rcvrn, Rho and Nrl showed a continuous increase of their expression during differentiation (Fig. 3D, Supplementary Fig. 2A, B), which was also in concordance with the developing mouse retina in vivo (Supplementary Fig. 2A, B).
Overall, these findings suggested that CiPSCs could differentiate into retinal organoids with polarized photoreceptor layers, which highly resembled in vivo photoreceptor genesis.
Development and maturation of retinal organoids
Retinal neurons form stratified structures called inner and outer plexiform layers which are essential for visual signal processing and transferring. On D18, Pou4f1, a marker for RGCs, was detected at the basal side of organoids (Fig. 4A). However, RGCs could not survive throughout the later stages of differentiation (Fig. 4A), probably due to the lack of axon formation21 and/or the insufficiency of oxygen and nutrients22 during differentiation, which also happened in other 3D ROs differentiation systems19.
A Retinal ganglion cells (RGCs) indicated by Pou4f1 (red) at D18 and D28 in ROs. Arrowheads indicate relevant immunostaining with Pou4f1. Scale bar, 50 μm. Images were captured using the same confocal settings. B Immunostaining of ROs for Prox1 (horizontal/amacrine cell marker, in green) and Calretinin (amacrine/ganglion cell marker, in red). Arrowheads indicate Prox1+, Calretinin+ double-positive cells representing amacrine cells. Scale bars, 25 μm (left, for insets), 50 μm (right). C Immunostaining of ROs for Prkca (rod bipolar cell marker, in green) and Rhodopsin (rod photoreceptor cell marker, in red). Arrowheads indicate relevant immunostaining with Prkca and Rhodopsin. Scale bars, 25 μm (left, for insets), 50 μm (right). D Müller glial cells indicated by Rlbp1 (red) in the ROs. Recoverin (green) is a marker for photoreceptor precursors. Arrowheads indicate relevant immunostaining with Rlbp1. Scale bar, 50 μm. Images were captured using the same confocal settings. E Immunostaining of organoids with connecting cilium marker Arl13b (green) and rod photoreceptor marker Rhodopsin (red). Arrowheads indicate relevant immunostaining with Arl13b and Rhodopsin. Scale bars, 5μm (left, for insets), 10 μm (right). F Immunostaining of organoids with synaptic vesicle marker Synaptophysin (red). Arrowheads indicate relevant immunostaining with Synaptophysin. Scale bars, 50 μm. G Ctbp2 (green) and Recoverin (red) are markers for synapse and photoreceptor precursors, respectively. Arrowheads indicate relevant immunostaining with Ctbp2 and Recoverin. Scale bars, 50 μm (left), 12.5 μm (right, for insets). H Transmission electron microscopy (TEM) imaging of ROs showed outer limiting membrane (*), inner segments (arrows) mitochondria (M), connecting cilia (CC), basal bodies (BB), and OS-like structures. Scale bars, 1 μm (left), 0.5 μm (middle, right). See also Supplementary Fig. 2.
Prox1 (a marker for horizontal and amacrine cells) and Calretinin (a marker for amacrine and ganglion cells) were detected on D25. Both signals were polarized at the basal side of organoids, and the Calretinin+ cells were located at the inner side of Prox1+ cells, which was the same as the distribution of horizontal cells, amacrine cells, and RGCs in the mouse retina (Fig. 4B). At the same time, the Prox1+, Calretinin+ double-positive cells indicated the generation of amacrine cells (Fig. 4B). As the bipolar cells are late-born cell type during mouse retinal development20, Prkca, a marker of rod bipolar cells, can only be detected at the late stage of differentiation (Fig. 4C). Moreover, most Prkca+ signals were located at the basal side of the neural retina and the inner side of Rhodopsin+ cells, reflecting the correct distribution of bipolar cells and rod photoreceptors in ROs (Fig. 4C).
Müller glia cells also have essential tasks in maintaining retinal homeostasis and photoreceptor function. On D21, Rlbp1, a marker for Müller glia cells, was detected at the basal side of organoids. The positive signals spanned the whole neural retina and formed the outer limiting membrane (OLM) at the apical side by D28 (Fig. 4D), which was similar to in vivo retinogenesis. Rhodopsin+ rod photoreceptors grew beyond the OLM (Fig. 4D), indicating the formation of the cilia and outer segments.
To examine the maturation of photoreceptors, we conducted immunostaining of organoids for cilium and synapse-related proteins. The connecting cilia marker Arl13b was detected at the apical side of the neural retina, co-labeled with Rhodopsin+ rod photoreceptors (Fig. 4E). Meanwhile, some Rhodopsin+ signals grew beyond the Arl13b+ cilia (Fig. 4E), indicating the formation of the outer segments. Synaptophysin is a marker for synaptic vesicles, normally labeling the outer plexiform layer (OPL) and inner plexiform layer (IPL) in the neural retina. The immunostaining result showed a stratified distribution of synaptophysin staining, representing the OPL and IPL (Fig. 4F) and reflecting the maturation of photoreceptors. In addition, Ctbp2 is a presynaptic protein that typically stains at the rod terminal in the ribbon synapse23. Ctbp2 staining was detected at the basal side of the neural retina on D32 and co-localization of Ctbp2 with Recoverin indicated the formation of synaptic connections between photoreceptors and bipolar cell types (Fig. 4G).
The ultrastructures of photoreceptors were further characterized by transmission electron microscopy (TEM)24. Numerous components of mature photoreceptors such as OLM, connecting cilia (CC), basal cilia (BC), mitochondria-rich inner segment (IS), and outer segment (OS)-like structure could be identified in the TEM images (Fig. 4H).
At the mRNA level, the qRT-PCR analysis of RGCs (Pou4f2, Atoh7), interneurons (Neurod1), and Müller glia (Rlbp1) related genes showed an increase of gene expression during differentiation, which conformed to the developmental mouse retina in vivo (Supplementary Fig. 2C–E).
These results suggested that CiPSC-derived 3D ROs contained all types of examined retina cell types, including photoreceptors, RGCs, horizontal, amacrine, bipolar and Müller glial cells, and the genesis and distribution of different cell types recapitulated the mouse retinogenesis in vivo. More importantly, the CiPSC-derived photoreceptors should have the ability to receive and transfer visual information by the successful formation of intact structures.
Dynamic transcriptomic profiling of retinal organoids
To characterize the transcriptome during retinal development, RNA-seq of CiPSC-derived ROs (D0, 14, 21, 28, and 32) and in vivo mouse retinae (embryonic day 18 (E18), postnatal day 1(P1), P7, P14, and P30) was performed (Fig. 5A–C and Supplementary Fig. 3). The principal component analysis (PCA) of these samples indicated that developmental timing is the most important factor, which, as the first principal component (PC1), can explain 72% and 67.5% of the variance in the two datasets (Fig. 5A, B).
A, B The principal component analysis (PCA) of the expressed CPM values for samples from day 14, 21, 28, and 32 of retinal organoids and embryonic day 18 (E18), postnatal day 1(P1), P7, P14, and P30 in vivo mouse retinae. Percentages indicate experimental variance assigned to each PC. C Heatmaps showing the expression of selected cell type-specific genes in retinal organoids and in vivo developing mouse retinae. The average expression value (counts per million [cpm]) at each timepoint is plotted in log2 scale. The color scale bars refer to gene expression. Three biological sample replicates were used at each timepoint. See also Supplementary Fig. 3.
The expression of selected cell-type-specific genes in ROs and in vivo mouse retinae was further analyzed (Fig. 5C, Supplementary Fig. 3). Based on IHC data, gene expression patterns during organoid differentiation showed a remarkable resemblance to those of the developing mouse retina in vivo (Fig. 5C, Supplementary Fig. 3A). Rod and cone photoreceptor genes showed a progressive increase with differentiation progress, which was consistent with in vivo retinogenesis. Changes in horizontal, amacrine, bipolar, and Müller glial cell-specific gene expression in organoids revealed a striking coincidence with in vivo retinae, suggesting that the CiPSC-derived ROs closely recapitulated in vivo development. Moreover, expression of RGC-specific genes was downregulated with differentiation and was hardly detectable in organoids after D21 (Fig. 5C, Supplementary Fig. 3A), consistent with their disappearance in organoids. Moreover, Pearson correlation analysis between retinal organoids and mouse retinae showed that the transcriptomes of D21 ROs correlated strongly with P1 and P7 mouse retinae (Supplementary Fig. 3B). Likewise, correlation was high among transcriptomes between D14 ROs and E18 mouse retinae, D28 ROs and P7 mouse retinae, D35 ROs and P14 mouse retinae, respectively (Supplementary Fig. 3B). These results suggested that the gene expression patterns of organoids were in concordance with the developing mouse retina in vivo.
Generation of Crx-tdTomato labeled photoreceptors from CiPSCs
The Crx plays an important role in specifying the fates of photoreceptor cells. Crx expression is first detected in photoreceptor precursors and then continues to express in all photoreceptors as the differentiation proceeds25. In order to track the development of photoreceptor cells, we constructed a Crx-reporter CiPSC line by inserting the coding sequence of the tdTomato fluorescent protein prior to the stop codon of Crx gene with CRISPR/Cas9 technology (Fig. 6A, Supplementary Fig. 4A–C).
A Schematic of the generation of Crx-tdTomato reporter CiPSCs and the transplantation of CiPSC-derived photoreceptors into rd10 mice. B Immunostaining of photoreceptor marker Crx in Crx-tdTomato ROs. Images were captured using the same confocal settings. Scale bar, 50 μm. C Representative bright field and fluorescence images of the Crx-tdTomato reporter CiPSC-ROs. Scale bar, 200 μm. D Representative infrared images of pupil areas measured in dark (a1) and light (a2) corresponding to dashed-line circles in 30 days of post-transplantation rd10 mice that received CiPSC-PR in one eye and a sham injection (DPBS) in the other at P24. E Pupillary response plots (a2/a1) against the different intensities of irradiance for D. N = 9 for both groups, ANOVA, mean ± SEM. F Integration and survival of Crx-tdTomato+ photoreceptors in the rd10 mouse retina at 2 months post-transplantation. Scale bars, 25 μm. G Orthogonal view of Recoverin (green) and Crx-tdTomato (red) double-positive cells (arrowhead). Scale bars, 25 μm (left), 10 μm (right). H Representative image to show that host bipolar cells (Prkca+, green, arrowhead) connected with Crx-tdTomato (red) photoreceptors at the tip of dendritic processes. Scale bars, 30 μm. I Representative image to display presynaptic formation at the host-graft interfaces (arrowheads). Scale bars, 25 μm (left), 10 μm (right). See also Supplementary Fig. 4.
From D14 onwards, both the number of tdTomato+ photoreceptors (Fig. 6B, Supplementary Fig. 4D) and the tdTomato intensity (Supplementary Fig. 4E) kept increased, especially at the apical side. The tdTomato+ signals colocalized with Crx+ cells by immunostaining (Fig. 6B), indicating that insertion of tdTomato coding sequence did not alter the expression pattern of Crx gene. Meanwhile, the percentage of tdTomato+ cells constituted around 85% of all cells in the ROs at D32 (Supplementary Fig. 4E), which was consistent with the mature mouse retina in vivo20. Additionally, qRT-PCR analysis of Crx also showed a continuous increase of Crx expression during differentiation, which was also in concordance with the developmental mouse retina in vivo (Supplementary Fig. 4F). Furthermore, almost all organoids strongly expressed Crx-tdTomato on D21, demonstrating the consistency in generating tdTomato+ photoreceptors (Fig. 6C). Therefore, the photoreceptors in retinal organoids could be reflected and recorded by tdTomato reporter during differentiation.
Transplantation of CiPSC-derived photoreceptors into rd mouse model
To determine whether CiPSC-derived photoreceptors (CiPSC-PR) can restore visual function, single Crx-tdTomato+ photoreceptor cell suspension was transplanted into the subretinal space of P24 rd10 mice, a mouse model of retinal degeneration (Fig. 6A). According to previous studies of our lab, cell sorting operations would lead to severe decline of cell states and affecting the survival and integration of transplanted cells in the host retina. Therefore, in this experiment, digested single cells were directly transplanted without sorting to ensure better cell survival and integration. For comparison, the P24 rd10 mice received Crx-tdTomato+ donor cells in one eye and sham injections (DPBS) in the other (N = 9). Four weeks after transplantation, the pupillary light reflex (PLR) test was conducted to measure photoreceptor function because the light-induced pupil constriction is a behavioral response that depends on photoreceptors’ functional connections with central brainstem targets26,27 (Fig. 6D). The results showed that the eyes received Crx-tdTomato+ cells had a more sensitive pupil reflex compared with the sham-injected eyes, especially at low light intensities (Fig. 6E).
Three months after CiPSC-PR transplantation, the eyes were examined by immunofluorescence for evidence of functional integration into the host retina. The tdTomato+ or tdTomato+ Recoverin+ cells were distributed in the ONL (Fig. 6F, G). The Prkca+ rod bipolar cells were in close proximity to tdTomato+ cells, indicating the potential formation of synaptic connections between graft photoreceptor cells and host rod bipolar cells (Fig. 6H). Moreover, CiPSC-PR were found to form synaptic terminals that express the synaptic vesicle protein synaptophysin, and synaptophysin+ synapses were consistently located in the area of the host-graft interface as well as the IPL within the host retina (Fig. 6I), which is essential for transmitting the light signals into the inner retina.
These results demonstrated that some of the transplanted CiPSC-PR integrated into the host retina, formed functional synaptic connections with downstream retinal neurons, and contributed to visual function.
- SEO Powered Content & PR Distribution. Get Amplified Today.
- PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
- PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
- PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
- PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
- Source: https://www.nature.com/articles/s41536-024-00387-7