RPE cells generated from human iPS cells do not fully-mature in vitro and continue to express several fetal markers. This precludes their use as effective disease models for adult-onset diseases and also reduces their capacity for an effective cell therapy for degenerative eye diseases. We have developed methods to fully-polarize and mature differentiated RPE generated from human iPS cells. Using mouse models with defective ciliogenesis, we demonstrated that primary cilium is necessary for inducing polarization phenotype in RPE cells. Furthermore, we showed that this phenomenon in mouse RPE cells is regulated by turning-off of canonical WNT signaling. In mouse mutants with ciliogenesis defects canonical WNT signaling continues to stay active at least until postnatal day one. We used three different drugs to regulate ciliogenesis in iPSC-RPE cells; aphidicolin a cell cycle blocker that induces two primary cilia per cell; PGE2 that enhances ciliogenesis by increasing ciliary protein trafficking; and HPI-4 a dynein ATPase inhibitor that blocks ciliary protein trafficking thus inhibiting cilia function. Our results show that treatment of aphidicolin and PGE2 significantly improve formation of apical processes on RPE cells, improves epithelial shape, and increases the expression of RPE-specific markers as compared to untreated or HPI-4 treated cells. Gene expression analysis suggests that aphidicolin treatment significantly shifts the gene expression pattern of cells towards adult-like phenotype. We confirm that, similar to mouse models with ciliogenesis defect, these changes in iPSC-RPE are also regulated by canonical WNT signaling. Further analysis of aphidicolin and PGE2 treated samples suggested that although ciliogenesis significantly improved in the entire monolayer, the number of cells with double cilia stayed below 5%. Aphidicolin only affects dividing cells and we confirmed by Ki67 labeling that there are less than 5% dividing cells in cultures. To understand how the effect of aphidicolin spreads to the entire monolayer, we first confirmed using electrophysiology of an intact RPE monolayer that the entire monolayer is significantly more polarized as compared to untreated or HPI-4 treated cells. Members of non-canonical WNT pathways mediate polarity in epithelial tissue. Using antibodies specific for PKC-delta an important mediator of non-canonical WNT signaling and phospho-MLC a downstream target of non-canonical WNT signaling, we confirmed that non-canonical WNT pathway was specifically activated in RPE cells by primary cilium induction. We used two different specific inhibitors of PKC-delta to confirm that primary cilium induced RPE monolayer polarization is mediated by members of non-canonical WNT pathway. This work provides insight into developmental pathways required for complete polarization of RPE monolayers. Furthermore, it provides fully-mature and functional RPE cells for developing effective cell-based therapies and in vitro disease modeling. This manuscript has been prepared and is ready for submission. Furthermore, RPE cells derived from iPS cells are mixed between peripheral and macular type. We are using a high throughput screening to make pure peripheral or macular RPE cell types from iPS cells. Using a machine learning based algorithm we can determine RPE shape metrics allowing us to analyze different types of cells made by the above-mentioned protocols. By comparing these cells to native RPE cells, we determine their type and then manipulate developmental pathways to guide them towards specific RPE cell type lineage.
Showing the most recent 10 out of 12 publications
