In the present study, we described in detail a procedure to generate kidney organoids from human iPSCs obtained from the blood erythroid progenitor cells of patients with ADPKD. Many somatic cell types have been successfully reprogrammed into iPSCs, including blood cells8. Here, we opted to use blood erythroid progenitors as the starting cell type for a number of reasons, such as their genomic integrity, epigenetic memory and efficient reprogramming12, 18, 19. The efficiency by which erythroid progenitors had been converted into iPSCs was as much as two orders of magnitude better (7%-28%) than the efficiency of differentiated blood cell types (0.02%-0.60%) or fibroblasts (0.74%)5. Despite the low abundance of these cells in the peripheral blood, they can easily be isolated and expanded in vitro to produce a sufficient number for reprogramming. One of the most important challenges in establishing suitable iPSCs is their genotype and phenotypic variability7, but there are some technical strategies proposed to circumvent or reduce such variability. In the present experiments, we used a non-integrating episomal system for reprogramming that does not require genome integration for the reprogramming genes to be expressed11, allowing less immunogenicity and rendering more genetically stable cells20, 21. Moreover, all samples were derived from the same source, which helps minimize epigenetic variability13, assuming that differentiation of the starting cell might influence the efficiency of the end results. Analysis of the pluripotency markers and the ability of iPSCs to differentiate into all three germ layers were compared side-by-side, and they were also compared to the H9 embryonic cell line, which was used as a positive control. All these comparisons showed no significant differences between ADPKD patients, HCs and H9 cells.
Several protocols have been developed for inducing the differentiation of iPSCs into kidney organoids22, 23. Based on protocols described by Cruz et al 7, we obtained structures with the characteristics of proximal tubules, including the expression of NHE3 and AQP1. In addition, the organoids expressed NPHS2, indicating the presence of podocyte-like cells. Conversely, AQP2 was not expressed in these tubular organoids. This finding is in agreement with previous studies pointing to several limitations in developing structures derived from the ureteric bud (UB) in vitro, with the development of UB requiring specific protocols, since the timing window for the addition of the specific inductor is very limited7, 24, 25.
In the process of generating kidney organoids from iPSCs, all of renal embryogenesis is recapitulated in vitro, initially involving the formation of cavitated spheroids from single cells 15. Indeed, we found that undifferentiated iPSCs from both ADPKD patients and healthy control formed epiblast-like spheroid structures that were morphologically similar. Although DNA sequencing of the PKD1 and PKD2 genes was not performed in the present study, the current findings indicate that the type of mutation carried by the patients had no influence on typical epiblast morphogenesis. Further studies are still needed to more properly evaluate the expression of pivotal molecules involved in this phase of embryonic development.
According to the protocol developed by Freedman et al15, we observed that spheroids did differentiate into tubular organoids 26 days after induction of differentiation, and the timing was similar for both ADPKD patients and HC (Fig. 3). Nevertheless, we noticed some variability regarding the timetable of expression and the levels of differentiation markers. On the other hand, the morphology of the tubular structures was similar and comparable among samples from the ADPKD patients and HC. In contrast, using iPSCs originating from renal biopsies of ADPKD patients, Freedman’s group found a higher variability in the ability of these cells to form organoids7. In addition to the huge diversity of mutations inherent to this disease, these differences in the efficiency of obtaining organoids may also be explained by the heterogeneous cell types present in biopsy samples, in contrast to the erythroid progenitors, which constitute a more homogeneous cell population. Taken together, these results indicate that despite the variability and differences in the efficiency of different iPSC clones to form kidney organoids, obtaining the latter from ADPKD patients may constitute a unique model for studying the multiple forms of phenotypical manifestations of this disease. Moreover, it is important to note that if there is a considerable variability of mutations in the ADPKD-affected population, the patient-specific iPSCs, being totally immune compatible, they might be used for individual regenerative therapies in the future, in addition to allowing the development of an in vitro model to verify whether the very early phenotypical manifestations are also as diverse as the clinical presentation.
Unstimulated organoids derived from HC or ADPKD patients did not spontaneously form cysts. The typical morphology of cysts and the expression of cystogenesis markers were not detectable in any sample, even extending the organoid differentiation in culture until completing 35 days. It has been demonstrated that cell proliferation alone does not produce cysts26. Moreover, a third independent event (known as third-hit) is required to accelerate cyst formation and growth in vivo 25. Conversely, cyst formation induced by forskolin did occur in the organoids from ADPKD patients but not in those from HC. Of note, although present, cysts were observed at a low percentage in the ADPKD samples even in the presence of forskolin, as has also been observed by Freedman et al15 in organoids formed from the CRISPR/Cas9 model of ADPKD disease. However, when adherent organoids were transferred to a 3D suspension in their model, cystogenesis increased 10-fold7, indicating that the microenvironment is directly implicated in cystogenesis. The low percentage of cysts in the ADPKD organoids in the current study could be accounted for by the use of adherent organoids.
Aside from the role of the microenvironment in enabling efficient cyst formation in vitro, the experimental protocol employed in the present study formed proximal tubule organoids but not more distal portions of the nephron25. Notably, it is well established that in vivo, the proximal tubule does not represent the main site of cyst formation, corresponding to approximately 1.8% 27, whereas the distal nephron (particularly the collecting duct) is the main site of cyst origin.