Long term expandable mouse and human induced nephron progenitor cells enable kidney organoid maturation and modeling of plasticity and disease

SUMMARY:

Nephron progenitor cells (NPCs) self-renew and differentiate into nephrons, the functional units of the kidney. Here, manipulation of p38 and YAP activity allowed for long-term clonal expansion of primary mouse and human NPCs, and induced NPCs (iNPCs) from human pluripotent stem cells. Molecular analyses demonstrated cultured iNPCs resemble closely primary human NPCs. iNPCs generated nephron organoids with minimal off-target cell types and enhanced maturation of podocytes relative to published human kidney organoid protocols. Surprisingly, NPC culture medium uncovered plasticity in human podocyte programs, enabling podocyte reprogramming to an NPC-like state. Scalability and ease of genome-editing facilitated genome-wide CRISPR screening in NPC culture, uncovering genes associated with kidney development and disease. Further, NPC-directed modeling of autosomal-dominant polycystic kidney disease (ADPKD) identified a small molecule inhibitor of cystogenesis. These findings highlight a broad application for the reported iNPC platform in the study of kidney development, disease, plasticity, and regeneration.

Graphical Abstract

eTOC Blub:

Nephron progenitor cells (NPCs) generate nephrons, the functional units of the kidney. Li and colleagues develop systems for long-term expansion of mouse and human NPCs which can further develop into mature nephron organoids. These in vitro systems generate novel insights into kidney development and disease, and reveal human nephron plasticity.

INTRODUCTION

SIX2⁺ nephron progenitor cells (NPCs) play a central role in kidney organogenesis, generating nephrons, the functional units of the kidney.^1–3 Dysregulation of NPC fates underlies a number of congenital kidney diseases⁴ while uncontrolled proliferation of NPCs in Wilms tumor is the most prevalent pediatric kidney cancer.⁵ Thus, a deeper insight into NPC biology is central to improving an understanding of kidney development, congenital disease and cancer, and to applying developmental insight to regenerating kidney functions.

We and others have developed systems to generate NPCs de novo from pluripotent stem cells or expand NPCs from primary NPCs isolated from embryonic/fetal kidneys such that: 1) mouse and human NPC-like cells can now be generated transiently using step-wise directed differentiation protocols from mouse and human pluripotent stem cells (hPSCs);^6–9 2) primary mouse and human NPCs can be isolated and expanded for a short period of time in two-dimensional (2D) culture format^10,11 or expanded long term over a few months in our previously reported three-dimensional (3D) culture format.¹² These systems have advanced an understanding of NPC biology and facilated the modeling of kidney development and diseases.¹³

While acknowledging progress, significant limitations remain. First, current hPSC-derived nephron organoids fail to generate mature, functional kidney cell types,^13–16 likely reflecting the quality of hPSC-derived NPC-like cells.¹⁶ Second, compared to 2D culture, the currently available NPC 3D culture system is tedious and less compatible with functional genomics tools, such as CRISPR screens,^17–19 hindering genomic scale study of NPC biology. Third, despite attempts from us and others,^10,12,20 it has not been possible to expand NPCs derived from hPSCs, the desired cell source for kidney regeneration and disease modeling, over the long term.

Here, we report the development and characterization of a chemically-defined 2D culture system supporting the stable clonal expansion and long-term culture of primary mouse and human NPCs, and hPSC-derived induced NPCs (iNPCs), and utilize human iNPC-derived kidney culture to explore cellular platicity, and demonstrate the facility of the platform for genome-wide genetic screening, and disease modeling and drug discovery.

RESULTS

p38 inhibition allows the derivation of clonal expandable NPC lines from any mouse strain.

We previously developed a 3D culture system that can expand mouse NPCs (mNPCs) as clusters of cells in a chemically-defined culture medium, mNPSR.¹² However, mNPSR did not support mNPC expansion in a regular monolayer (2D) culture setting. To solve this problem, we screened for small molecules and growth factors that support the derivation and long-term clonal expansion of mNPC lines from multiple mouse strains resulting in mNPSR-v2 (Fig. 1A, Fig. S1A and Methods S1).

Figure 1. p38 inhibition allows the derivation of NPC lines from any mouse strain.

(A) Schematic of mNPC line derivation and applications.

(B) Quantification of SIX2⁺/PAX2⁺ cell percentages in cultured in various conditions. Scale bars, 50 μm.

(c) Immunostaining of E13.5 mouse kidney section for markers as indicated. Scale bar, 50 μm.

(D) Morphology of mNPCs (day 28). Scale bar, 50 μm.

(E) Growth curve of mNPCs starting from 5,000 cells.

(F and G) Immunofluorescence images (F) and quantification (G) of mNPCs (day 28). Scale bars, 100 μm.

(H) Bright-field image showing co-culture of spinal cord (SP) with aggregated mNPCs. Scale bar, 200 μm.

(I) Whole-mount immunofluorescence analysis of mNPC-derived nephron structures in (H). Scale bar, 100 μm.

(J) Time-course images showing mNPC clonal expansion. Scale bars, 50 μm.

(K) Bright-field (BF) and immunofluorescence images of a single cell mNPC clone. Scale bars, 50 μm.

(L) Growth curve of a single mNPC.

(M) Cloning efficiency of mNPC lines.

(N) Whole-mount immunofluorescence analyses of a clonal NPC line-derived organoid. Scale bars, 100 μm.

(O) PCA of bulk RNA-seq datasets.

(P) Heatmap showing selected marker gene expression. Data are presented as mean ± SD. Each column represents counts from three biological replicates (n=3).

See also Figures S1, S2 and Methods S1, and Table S1.

Compared to mNPSR, mNPSR-v2 has four additional small molecules: SB202190 (inhibitor of p38 MAPK), DAPT (inhibitor of Notch signaling), A83–01 (inhibitor of TGF-β signaling) and LDN193189 (inhibitor of BMP signaling), and required a different concentration (1.5 μM) of CHIR99021 (inhibitor of GSK3). Of these additional components, we have previously found that adding A83–01 and LDN193189 to mNPSR can enable the expansion of mNPCs at a lower seeding density of mNPCs in a 3D culture setting.²¹ LDN193189¹⁰ and DAPT¹¹ have also been used to support short-term expansion of mNPCs. Consistently, we noticed addition of DAPT prevented spontaneous differentiation of 2D cultured mNPCs (Fig. S1C and D). The p38 MAPK inhibitor SB202190 has not been previously reported to support mNPC self-renewal, but appears to have the most significant effect in sustaining the percentage of SIX2⁺/PAX2⁺ mNPCs in culture (Fig. 1B and S1B). In addition, intrinsic p38 MAPK activity was found to be low in the self-renewing mNPCs in vivo (Fig. 1C). mNPCs expanded in mNPSR-v2 stably proliferate with highly homogeneous morphology (Fig. 1D and E), show uniform NPC marker gene expression (Fig. 1F and G), at similar levels to those observed in primary NPCs (Fig. S1E). Upon withdrawing each individual medium component from mNPSR-v2, we confirmed that all components are essential (Fig. S1F and G).

To examine the nephrogenic potential of the cultured mNPCs in vitro and in vivo, we examined first the inductive response in the classic spinal cord induction assay.^6,12 We observed the formation of numerous tubule-like structures after 7 days (Fig. 1H), with PODXL⁺ glomeruli, LTL⁺ proximal tubule, and CDH1⁺ distal tubule structures (Fig. 1I). Nephron organoids were also formed from cultured mNPCs using our chemically-defined media,¹² generating multiple segments of the nephron (Fig. S1H). When we reconstructed an engineered kidney from cultured mNPCs and cultured ureteric bud (UB),²² mNPCs induced dramatic branching morphogenesis from the UB (Fig. S1I), while the UB induced nephron formation from the mNPCs (Fig. S1J). When mNPCs were transplanted onto the chicken chorioallantoic membrane (CAM) in vivo,²³ mNPCs differentiated into nephrons and chick vasculature infiltrated the transplant (Fig. S1K–M, and Video S1).

2D-cultured mNPCs were efficiently expanded from single cells with a high cloning efficiency of 60–70% (Fig. 1J–N, S2A and B). To compare the global gene expression of cultured and primary mNPCs,^12,24 we performed bulk RNA-seq (Table S1). Based on principal component analysis (PCA), cultured NPCs were clustered tightly together (Fig. 1O and S2C), largely overlapped with primary E11.5, E12.5, and E13.5 mNPCs in PC1 and PC3 axes (Fig. S2D) as well as in the heatmap, using a selective group of typical NPC and nephron marker genes (Fig. 1P). Comparative analysis indicates cultured mNPCs resemble early-stage (E11.5 to E13.5) mNPCs. The robustness of mNPSR-v2 allowed us to derive mNPC lines from isolated E11.5 metanephric mesenchyme (Fig. S2E and F) or from whole kidney cells of an early embryonic kidney (Fig. S2G–K), enabling the derivation of mNPC lines from all mouse strains tested (Fig. 1A and Methods S1).

Plasticity of developing mouse nephron cells with mNPSR-v2 medium.

While developing the mNPSR-v2 medium, we noticed a significant portion (10–20%) of Six2-GFP negative (Six2-GFP⁻) cells isolated from dissociated kidneys at all stages (E12.5, E14.5, E16.5 and P0 kidneys) adopted a SIX2⁺/SALL1⁺ phenotype within 4 days of culture (Fig. 2A and B). This finding suggests the possibility of phenotypic plasticity on the part of non-NPC type(s) in mNPSR-v2 medium. To exclude the possibility of potential contamination of a small number of Six2-GFP⁺ mNPCs during FACS, we cultured cells from kidneys at postnatal day 3 (P3), P4, P5 and P7, stages that do not have Six2-GFP⁺ cells (Fig. S3A), in mNPSR-v2 for 4 days. Approximately 17% of the P3 kidney cells and 7% of the P4 kidney cells were scored as SIX2⁺/SALL1⁺, rare SIX2⁺/SALL1⁺ cells were observed also within P5 and P7 cultures (Fig. 2C and D, and S3B and C). This happens most efficiently in mNPSR-v2 (18%), less efficiently in mNPSR (6%), and does not happen in basal medium with 10% FBS (0%) (Fig. S3D and E). To further characterize the induced Six2⁺ cells, we developed an assay to purify these cells using an NPC lineage tracing reporter system (Fig. 2E).³ Similar to immunostaining results, after 4 days of culture, around 20% of the cultured P3 kidney cells became Six2-tdTomato (Six2-tdT)⁺ as shown by FACS (Fig. S3F and G). On culture day 8, Six2-tdT⁺ cells showed a more complete NPC profile expressing Six2, Pax2, Sall1, Wt1, Gdnf, Hoxd11, and Eya1, coinciding with loss of expression of nephron marker genes Slc12a1, Slc12a3, and Aqp1 (Fig. 2F).

Figure 2. Plasticity of developing mouse nephron cells with mNPSR-v2 medium.

(A - D) Immunofluorescence images and quantification of the expression of SIX2 and SALL1, from cultured Six2-GFP⁻ cells (A and B) and from cultured postnatal kidney cells (C and D). Note that only fluorescence signals in the nucleus were true SIX2 signals. Membrane-bound signals were non-specific. Scale bars, 100 μm.

(E) Schematic showing genetic labeling and FACS isolation of the induced Six2-tdT⁺ cells.

(F) qRT-PCR analysis of samples as indicated.

(G) Schematic showing genetic labeling, FACS isolation, and culturing of Wnt4-tdT⁺ cells.

(H) Flow cytometry gating plot showing the purification of Wnt4-tdT⁺ cells.

(I and J) Immunofluorescence images (I) and quantification (J) of the expression of SIX2 and SALL1 in samples as indicated. Scale bars, 100 μm.

Data are presented as mean ± SD. Each column represents counts from three biological replicates (n=3). The significance was determined by two-tailed unpaired Student’s t tests; ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001.

See also Figure S3 and Methods S1.

Wnt4+ cells, the immediate differentiated descendants of NPCs, can migrate back to the cap mesenchyme niche, where NPCs reside, and revert to an NPC state in vivo.²⁵ To investigate whether Wnt4+ cells contribute to the plasticity we observed in vitro, we injected tamoxifen into P2 Wnt4-GFP-Cre-ERT2; Rosa26-tdTomato (Wnt4-tdT for short) mice to permanently label Wnt4⁺ cells and their progeny. Kidneys were isolated 24hrs later on P3, and tdTomato⁺ cells FACS purified and cultured in mNPSR-v2 (Fig. 2G and H). After 4 days of culture, more than 70% of the cultured Wnt4-tdT+ cells stained positive for SIX2/SALL1, as compared to 20% starting from P3 whole kidneys (Fig. 2I and J). These results suggest that, as in vivo, the Wnt4+ cells represent a highly plastic cell population that can be efficiently reversed to an NPC state in vitro in the presence of mNPSR-v2 mirroring niche maintenance conditions.

Genome-wide CRISPR screen in NPC lines offers resources for studying kidney development and disease.

To understand NPC biology from the genome-wide perspective, we introduced a genome-wide CRISPR knockout library²⁶ into cultured NPCs to screen for genes essential for NPC fitness (Fig. 3A) comparing the relative abundance of sgRNAs in the cultured NPCs by next-generation sequencing 3 weeks after introduction of the CRISPR library (Table S2). With the MAGeCK-VISPR tool,²⁷ sgRNA abundance changes from 4 different sgRNAs targeting the same gene in the library are statistically integrated as beta scores (Table S2). Positively-selected (i.e. increased sgRNA abundance) genes with more dramatic sgRNA abundance increase will have higher positive beta scores; while negatively-selected genes with more dramatic sgRNA abundance decrease appear as lower negative beta scores.

Figure 3. Genome-wide CRISPR screen in NPC lines.

(A) Schematic illustrates the workflow of genome-wide CRISPR screen.

(B) Box plot showing the distribution of beta scores. Boxes, 25th to 75th percentiles; whiskers, 1st to 99th percentiles.

(C) MAGeCKFlute scatterplots of beta scores showing common genes identified from replicates.

(D) Top 11 enriched IPA Canonical Pathways from CRISPR screen replicate #1.

(E–I) MAGeCKFlute scatterplots of beta scores from two CRISPR screen replicates.

(J and K) Normalized read counts of individual sgRNAs at the start and the end of the screen.

(L and M) Immunofluorescence images (L) and quantification (M) of NPC marker gene expression upon inhibitor treatment. Scale bars, 50 μm.

Data are presented as mean ± SD. Each column represents counts from three biological replicates (n=3). The significance was determined by two-tailed unpaired Student’s t tests; ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001.

See also Figure S4 and Table S2.

After performing quality control analyses for the similarity of biological replicates (Fig. S4A), global changes of sgRNA abundance in the screen (Fig. S4B and S4C), and negative selection for “pan-essential genes”²⁸ (Fig. 3B), we identified 410 positively-selected genes (beta scores >1.5) and 696 negatively-selected genes (beta scores <−1.5) shared between replicates (Fig. 3C). Ingenuity Pathway Analysis (IPA) confirmed the enrichment of fundamental cellular pathways necessary for cell survival, and identified mTOR signaling, microRNA biogenesis, and oxidative phosphorylation pathways, which play critical roles in NPC self-renewal,^29–33 among the top enriched pathways (Fig. 3D, S4D and E, and Table S2). Consistent with an essential role for FGF, WNT, and LIF signaling in NPC self-renewal in vitro (Fig. S1F and G), genes encoding critical receptor, signaling mediator, and effector proteins of these pathways, were identified in the screen (Fig. 3E, S4F and G, and Table S2).

Based on CRISPR screen results and NPC-specific gene expression profiles,^12,24 a short-list of 64 potential NPC self-renewal genes were obtained (Table S2 and Methods). Interestingly, 25 out of the 64 genes encode proteins that localize to the nucleus (Fig. 3F), including genes linked to NPC programs: Sall1,^34–36 Wt1,³⁷ Foxc1,³⁸ and Myc.^39,40 Amongst genes with no previous NPC association we identified Sobp, a co-factor for Six1, which encodes a key transcriptional determinant of the NPC state.⁴¹ SOBP is known to interfere in the transcriptional activation of SIX1/EYA1 target genes during craniofacial development, likely leading to Branchio-oto-renal syndrome (BOR).⁴² These observations, together with our CRISPR screen result, support a potential role for Sobp in the regulation of NPC fates.

Currently, around 30 Wilms tumor-related genes have been well established.⁵ Of note, 11 out of the 30 genes (36.7%) were identified by our CRISPR screen (Fig. 3G, and Table S2). Out of the 330 CAKUT associated genes,⁴ 25 (7.6%), were identified in our CRISPR screen, consistent with the notion that dysregulation of NPC fates represents a significant source for kidney malformation (Fig. 3H and Table S2). Confirmation of the large numbers of known Wilms tumor and CAKUT related genes suggests that our CRISPR screen datasets might identify other genes with previously unknown functions in these diseases (Table S2).

Epigenetic mechanisms have been found to play critical roles in NPC self-renewal in vivo.⁴³ Our CRISPR screen has identified the majority of these reported epigenetic factors, including Hdac1,⁴⁴ Chd4,⁴⁵ Ezh2,⁴⁶ and Smarca4.⁴⁷ In addition, Kmt2a (Mll1) and Kat6a, were found to be among the top negatively-selected genes (Fig. 3I–K). After confirming the expression of Kmt2a and Kat6a in the cultured NPCs and primary NPCs (Table S1), we employed two small molecule inhibitors to KMT2A, MLL1 (inh), and WDR5 degrader,⁴⁸ and two small molecule inhibitors to KAT6A, WM-1119, and MOZ-IN-3, and experimentally validated KMT2A and KAT6A activities are essential for NPC self-renewal (Fig. 3L and M, and S4H and I). Considering recent whole-exome sequencing in families with CAKUT identified dominant monogenic point mutations in human KMT2D and KAT6B genes associated with syndromic CAKUT,⁴⁹ these findings suggest mutations in KMT2D and KAT6B might dysregulate NPC programs in human CAKUT.

Taken together, our CRISPR screen datasets (Tables S1) provide valuable genome-scale resources for future studies of kidney development, Wilms tumor, and CAKUT.

YAP activation derives long-term expandable human NPC lines.

p38 MAPK activity, TGF-β and BMP signaling, are intrinsically low in the primary hNPCs in vivo (Fig. 4A–C and S5A). Thus, we included p38, TGF-β, and BMP inhibitors in the culture medium to expand human NPCs as we did in mNPSR-v2. By replacing mouse LIF with human LIF in the mNPSR-v2 medium, we prepared hNPSR-v1 medium. Relying on our engineered H1 human embryonic stem cells (hESCs) with dual reporter system for SIX2-GFP and PAX2-mCherry, following a well-established 10-day hPSC-to-NPC differentiation protocol⁸ (Fig. S5B), we purified the SIX2⁺/PAX2⁺ hPSC-induced NPC (iNPC) population by FACS (Fig. S5C) and cultured them in hNPSR-v1. However, after 2 weeks iNPC cell proliferation slowed, NPC marker gene expression was down-regulated and iNPCs lost YAP expression in the nucleus (Fig. 4D and E). Considering the reported role of YAP activity in the self-renewal of NPCs,⁵⁰ we supplemented small molecule YAP agonist TRULI⁵¹ to the medium to make hNPSR-v2 (Methods S1).

Figure 4. YAP activation derives long-term expandable human NPC lines.

(A–C) Immunofluorescence images of human fetal kidney sections (11.4wk). Scale bars, 100 μm.

(D and E) Immunofluorescence images (E) and quantification (F) of YAP expression in cultured iNPCs. Scale bars, 50 μm.

(F) Schematic showing the derivation of iNPC lines and applications.

(G) Bright-field image of iNPCs (day 87). Scale bar, 50 μm.

(H) Growth curve of iNPCs starting from 5,000 cells.

(I and J) Immunofluorescence images (I) and quantification (J) of iNPCs (day 21). Scale bars, 50 μm.

(K) Bright-field images showing clonal expansion of iNPCs. Scale bars, 50 μm.

(L) Immunofluorescence images of a single cell iNPC clone. Scale bars, 50 μm.

(M and N) 3D (M) and 2D (N) PCA plots of bulk RNA-seq data.

(O) Heatmap showing gene expression of selected marker genes. “D0-iNPC-SIX2” and “D0-iNPC-SIX2/PAX2” are FACS-purified SIX2⁺ and SIX2⁺/PAX2⁺ iNPCs without further culture; “Pri-SIX2-Neg” are primary SIX2-negative non-NPCs from human fetal kidneys.

(P and Q) Bright field (BF) and fluorescence images (P) and quantification (Q) of mCherry expression in iNPCs upon lentiviral overexpression of mCherry (lentiviral OE), or knock-in of mCherry-expressing cassette into AAVS1 allele (CRISPR KI). Scale bars, 50 μm.

(R-U) Whole-mount immunofluorescence analyses of nephron organoids generated from iNPCs (day 42). Scale bars, 50 μm.

Data are presented as mean ± SD. Each column represents counts from three biological replicates (n=3). The significance was determined by two-tailed unpaired Student’s t tests; ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001.

See also Figures S5, S6, and Methods S1 and Table S1.

iNPCs grew stably in hNPSR-v2 (Fig. 4F–H), and after 3 weeks of culture, more than 95% of cells maintained expression of SIX2/PAX2/WT1/SALL1/ITGA8 (Fig. 4I and J). After long-term culture of more than 3 months, all examined NPC marker genes were still retained, except for a gradual decrease of PAX2 starting after 1 month of culture (Fig. S5D–F). iNPCs showed a clonal efficiency of 58–70% in reseeding experiments (Figs. 4K and L, S5G and H). Importantly, the robustness of hNPSR-v2 enabled the direct derivation of iNPC lines without prior FACS enrichment of SIX2⁺/PAX2⁺ iNPCs (Fig. 4F, S5I–K and Methods). Primary human NPCs were stably expanded in hNPSR-v2 for 100 days while retaining NPC gene expression (Fig. S6A–J). Bulk RNA-seq analysis was performed to compare iNPCs without further culture in hNPSR-v2 (day 0), cultured human NPCs (day 15 to day 80) from different sources (iNPC or primary NPCs) (Table S1) and primary human NPCs from human fetal kidneys.²⁴ PCA analysis (Fig. 4M, N, and S5L) and heatmap of NPC marker genes and nephron segment anchor genes (Fig. 4O) both showed that, compared to iNPCs without further culture, or cultured in the v1 medium, NPCs cultured in the v2 medium are more similar to primary NPCs. Of note, iNPCs without further culture (iNPC D0) showed significantly lower expression of SIX2, HOXA11, and HOXD11 genes (Fig. 4O), which are induced after culture in hNPSR-v2. Furthermore, we showed that genetic manipulations, including lentiviral gene overexpression and CRISPR-Cas9 based gene knock-in, can be efficiently conducted in the cultured iNPCs, opening avenues of applications for this system (Fig. 4P and Q).

We next applied a pulse of Wnt activation to generate nephron organoids from cultured NPCs, a standard approach in protocols driving hPSC kidney organoid development that mirrors Wnt induction of the mesenchymal-to-epithelial transition of NPCs in vivo (Fig. S6K, and Methods). We observed a clear mesenchymal-to-epithelial transition in the first 3 days after Wnt activation, with numerous PAX2-mCherry reporter positive renal tubule-like structures formed over a course of 14 days of culture (Fig. S5M and N). Immunostaining further confirmed the formaton of renal vesicle (RV) or S-shaped body (SSB) stage early nephrons 8 days after differentiation (Fig. S5O), and the formation of various nephron segments after 2 weeks of differentiation, including PODXL⁺/NPHS1⁺ glomeruli, LTL⁺/HNF4A⁺/SLC34A1⁺/SLC27A2⁺ proximal tubule, and PAX2⁺/SLC12A1⁺ loop of Henle, with primary cilia observed exclusively on the apical side of the renal tubules (Fig. 4R–U, S5P and Q). Current kidney organoids generate a significant amount of off-target cell populations such as neurons and myocytes.^14,15,52,53 In contrast, in the cultured iNPC-derived nephron organoids, MAP2 and NeuN, two neuron genes frequently expressed in current kidney organoids, were not detected (Fig. S5R). Similar results were observed in the nephron organoids that were derived from long-term cultured primary human NPCs (Fig. S6L–N).

Transcriptome analyses reveal in vivo-like nephrogenesis, mature podocyte formation, and minimal off-target cells from iNPC-derived nephron organoid.

To examine progressive transcriptional changes during in vitro differentiation in our organoid model, we performed time-course bulk RNA-seq at 0, 2, 4, 7, 10, 14, and 21 days post induction with two replicates (Fig. 5A). Unsupervised clustering by principal component analysis (PCA) grouped the samples in a temporal order from undifferentiated NPCs (D0) to differentiated nephron organoid (D21) (Fig. 5B). Transcriptional signatures sequentially progress from NPC (D0) to pretubular aggregate (PTA), renal vesicle (RV), and S-shaped body (SSB) stages (D2-D7), and then exhibit profiles resembling maturing nephrons containing podocytes and renal tubules (D10-D21), recapitulating in vivo-like staged nephrogenesis (Fig. 5C). For example, the time-course gene expression profile includes the in vivo podocyte development trajectory: podocytes at the early period of the development (“early podocyte”, E-Pod) genes (e.g., SLC16A1, OLFM3, and PCDH9) at D7 and D10; mature podocytes (“late podocyte”, L-Pod) genes (e.g., NPHS1, PODXL, and PLA2R1) detected at D10 and gradually increased until D21 (Fig. 5C). Importantly, the late podocyte genes COL4A3 and COL4A4, encoding essential glomerulus basement membrane components associated with Alport syndrome,⁵⁴ were expressed abundantly in the iNPC-derived nephron organoids, in contrast to the previous reports of other hPSC-derived kidney organoid models.^8,9,55

Figure 5. Single cell multiome analyses of iNPC-derived nephron organoids.

(A) Schematic of the experimental design of multimodal analyses, and the major conclusions.

(B) PCA of bulk RNA-seq datasets as indicated.

(C) Heatmap showing expression of selected markers from bulk RNA-seq datasets.

(D) UMAP projection of iNPC-derived nephron organoid snRNA-seq dataset.

(E) UMAP projection of signature gene expression in iNPC-derived nephron organoid.

(F) Dot plot of cluster-enriched gene expression in iNPC-derived nephron organoid.

(G) UMAP projection of integrated single-cell datasets of iNPC-derived organoids and two other published hPSC-derived kidney organoids.

(H) Proportions of cell types identified in (G) in different kidney organoids. Morizane protocol: TT.r1 and TT.r2;⁵⁷ Takasato protocol: AN1.1, BJFF, and H9.⁵⁸

(I and J) NPHS1 expression in the integrated dataset (G) is shown through a feature plot (I), and a violin plot (J).

(K-M) Dot plots of marker gene expression in the proximal tubule population (K), distal tubule population (L), and podocyte population (M) from the integrated dataset (G).

(N) Genome browser views of snATAC-seq open chromatin regions of selected genes in iNPC-derived podocyte (COL4A3, COL4A4, and NPHS1) and distal tubule (SLC12A1), as compared to adult kidney’s podocyte and distal tubule.

See also Figure S7, Tables S1 and S3.

We then employed single-nuclear multiome analysis on D21 nephron organoids with the 10X Chromium Single Cell Multiome ATAC + Gene Expression platform. After applying quality-control metrics and eliminating potential sequencing artifacts through the Seurat V4 R package⁵⁶ (Fig. S7A and B), we obtained 21,404 nuclei from two biological replicates of organoid samples, with a median of 2,888 genes per nucleus from the snRNA-seq dataset, and 8,711 fragments per nucleus from the snATAC-seq dataset. Consistent with our bulk RNA-seq and immunostaining results, unsupervised UMAP clustering of the snRNA-seq dataset identified major compartments of the nephron including podocyte, proximal tubule and distal tubule. In addition, a COL1A1⁺ interstitium population and a minor population of off-target cells (0.67%) were also identified (Fig. 5D–F, S7C and D, and Table S3).

To directly compare our iNPC-derived organoids with existing kidney organoids generated from hPSCs, we integrated our snRNA-seq data (day 21) with two published datasets (days 26 and 28) representing two prevailing kidney organoid protocols from hPSCs: 1) our previously published kidney organoid dataset following the Morizane protocol⁵⁷ (samples TT.r1 and TT.r2); and 2) a recent publication following the Takasato protocol from the Humphreys lab generated using the same single-nuclear multiome approach⁵⁸ (samples AN1.1, BJFF.6, and H9) (Fig. 5G, S7E and F, and Table S3). The relative proportions of cell-types generated in each kidney organoid model was quantified. Compared to hPSC-derived kidney organoids, iNPC-derived kidney organoids have three unique features: 1) higher proportion of podocytes (Fig. 5H–J), 2) few off-target cells (Fig. 5H), and 3) no residual undifferentiated SIX2⁺ or EYA1⁺ NPCs (Fig. 5H). The observation of fewer off-target cells is consistent with our immunostaining results (Fig. S5R) and is in line with expectations considering the high purity of the iNPCs and their restricted developmental potential to generate nephrons.

For the renal tubule compartment, the proximal tubule cells in iNPC-derived organoid showed high expression levels of CUBN and LRP2 (Fig. 5K). The majority of distal tubule cells showed high expression levels of SLC12A1 (Fig. 5L), consistent with their strong expression and apical localization shown by immunostaining (Fig. 4S and T). Consistent with our bulk RNA-seq results (Fig. 5C), the multiome analysis confirmed that the podocytes in the iNPC-derived organoid showed lower expression of early podocyte genes and higher expression of late podocyte genes compared to those forming in hPSC-derived organoids, suggesting improved podocyte differentiation in our model (Fig. 5M). Supporting this, in the snATAC-seq dataset, the open chromatin accessibility of COL4A3, COL4A4, and NPHS1 from iNPC-derived podocytes mimic the podocytes from adult kidneys⁵⁹ (Fig. 5N).

Reprogramming from podocyte to NPC by hNPSR-v2 reveals human podocyte plasticity.

To examine the potential for cell plasticity in human nephron development, we differentiated SIX2-GFP; PAX2-mCherry dual-reporter iNPCs to nephron organoids. 7 days after nephron induction, we FACS-isolated SIX2-GFP⁻ non-NPCs and cultured cells in hNPSR-v2 (Fig. 6A and B). After 9 days in hNPSR-v2, 27.7% of cells in the SIX2-GFP⁻ seeded cultures were SIX2-GFP⁺. SIX2-GFP⁺ cells were FACS-purified (using GFP) and stably expanded in hNPSR-v2 assuming a morphology indistinguishable from iNPCs (Fig. 6C) while maintaining consistent expression of NPC marker genes (Fig. S8A and B). Further, these SIX2-GFP⁻ descendant cells underwent similar organoid differention to primary SIX2-GFP⁺ iNPCs (Fig. S8C and D). Similar plasticity was observed from SIX2-GFP⁻ cells isolated from organoids 3, 5, and 8 days after nephron induction (Fig. S8E–I). Transcriptome analyses showed the transition from SIX2-GFP⁻ non-NPCs to SIX2-GFP⁺ NPCs, inseparatable from iNPCs on the basis of PCA of mRNA profiles (Fig. 6D, and S8J) and expression of NPC marker genes (Fig. S8K). Given the emergence of SIX2-GFP⁺ NPC-like cells from a SIX2-GFP⁻ population we termed this cell type “reprogrammed” NPC (rNPC).

Figure 6. hNPSR-v2 reveals human podocyte plasticity.

(A) Schematic of the reprogramming process.

(B) Flow cytometry gating plots showing isolation of SIX2-GFP⁻ cells from day 7 nephron organoids (left) and purification of SIX2-GFP⁺ rNPCs upon culture (right).

(C) Bright field image of rNPCs (day 13) derived from (B). Scale bar, 50 μm.

(D) PCA of bulk RNA-seq data.

(E) Flow cytometry gating plots showing isolation of SIX2-GFP⁻/PODXL⁺ podocytes from day 7 nephron organoids (left) and purification of rNPCs upon culture (right).

(F) Bright field images of SIX2-GFP⁻/PODXL⁺ podocytes from (E) and the derivative rNPCs. Scale bars, 50 μm.

(G-J) Immunofluorescence images and quantification of SIX2-GFP⁻/PODXL⁺ podocytes (G and I) and the derivative rNPCs (H and J). Scale bars, 50 μm.

(K) Bright-field (BF) and immunofluorescence images of MAFB-GFP nephron organoid (day 8). Scale bars, left, 200 μm; right, 100 μm.

(L) Flow cytometry gating plot showing the enrichment of rNPCs with ITGA8.

(M) Bright-field image of rNPC line derived from MAFB-GFP⁺ podocytes. Scale bar, 50 μm.

(N and O) Whole-mount immunofluorescence images of rNPC line (from MAFB-GFP⁺ podocytes)-derived organoids. Scale bars, left, 200 μm; right, 50 μm.

(P) Flow cytometry gating plots showing isolation of PODXL⁺ primary podocytes from 17.4 week human fetal kidney (left) and enrichment of rNPCs upon culture with ITGA8 (right).

(Q) Immunofluorescence images of primary podocyte-derived rNPCs (day 31). Scale bars, 50 μm.

(R) Whole-mount immunofluorescence images of rNPC (from primary podocyte)-derived nephron organoid. Scale bars, left, 200 μm; right, 50 μm.

(S) Heatmap showing marker gene expression during podocyte-to-NPC reprogramming.

(T) Schematic of the model.

See also Figures S8, S9, and Table S1.

Podocyte progenitors are recruited late in renal vesicle formation but specified early developing podocytes share molecular features of NPCs.^2,60 We thus investigated whether specified podocytes can be reprogrammed to rNPCs. PODXL⁺/SIX2-GFP⁻ podocyte population (32.2%) were sorted out by FACS from nephron organoids, which were futher verified to be SIX2⁻/SALL1⁻/MAFB⁺/NPHS1⁺/WT1⁺/PODXL⁺ (Fig. 6G and I). After 7 days of culture in hNPSR-v2, podocytes underwent a mesenchymal transition (Fig. 6F and S8L). Remarkably, by day 8 of culture, 44.6% of original PODXL⁺/SIX2-GFP⁻ cells were SIX2-GFP⁺ (Fig. 6E). Stable rNPC lines derived after this timepoint showed uniform NPC marker gene expression (Fig. 6H and J) and a nephrogenic potential similar to that of other rNPC and iNPC lines (Fig. S8M). As a secondary validation, we employed our previously described MAFB-GFP knockin reporter hPSC line derived from the H9 hESC background.⁶¹ iNPC lines were generated from MAFB-GFP hPSC line. MAFB-GFP⁺ podocytes were isolated by FACS from day 8 iNPC-derived nephron organoids and cultured in hNPS-Rv2 (Fig. 6K and S8N). rNPCs were isolated from these cultures by FACS sorting on surface marker ITGA8⁺, which is highly enriched in NPCs, and continued culture in hNPSR-v2 (Fig. 6L and M). Analysis of NPC gene expression (Fig. S8O and P) and nephrogenic potential (Fig. 6N and O) supported a reprogramming of MAFB-GFP⁺ podocytes to NPCs. Importantly, PODXL⁺ primary podocytes, purified from 17.4 week (17.4wk) and 11.4wk human fetal kidneys and cultured in hNPSR-v2, successfully derived rNPC lines with nephrogenic potential (Fig. 6P–R, S9A–H, and Methods).

To examine progressive transcriptional changes during podocyte-to-NPC reprogramming, we performed time-course bulk RNA-seq (Methods). PCA and Pearson correlation heatmap clearly segregated D0 (podocytes) samples from samples of all other time points, with D4/D6 samples and D9/D16/D24 samples forming two clusters (Fig. S9I and J). By examining marker genes for human nephrogenesis, podocyte genes were found to be highly expressed in the D0 samples but nearly completely depleted from D2 (Fig. 6S). On the contrary, expression of genes representing PTA/RV stage (e.g. PAX8, WNT4, CRYM) was induced from D2, reached the peak on D4, and decreased on D6. From D9 to D24, the samples no longer express any podocyte or PTA/RV signature genes, but show strong NPC signature gene expression. Between day 0 and 4 of culture, podocyte signature gene significantly down-regulated while PTA/RV genes were significantly up-regulated (Fig. S9K and M, and Table S1). Comparing D24 to D4 showed a marked increase in expression of NPC signature genes (Fig. S9L and N, and Table S1). These data support a model in which podocytes cultured in hNPSR-v2 undergo a reversal of the differentiation program through a PTA/RV intermediate before stabilizing as rNPCs (Fig. 6T).

Rapid PKD modeling and small molecule screening from genome-edited NPCs.

Autosomal-dominant polycystic kidney disease (ADPKD) is the most prevalent inherited kidney disease.⁶² Studies have successfully modeled ADPKD through the genetic removal of PKD1 or PKD2 in hPSCs, and hPSC differentiation to cyst-forming kidney organoids.^61,63–66 Starting from genome-edited mouse and human NPC lines, we have developed rapid, efficient, and scalable ADPKD modeling and small molecule sreening platforms (Fig. 7A and B).

Figure 7. PKD modeling and small molecule screening from genome-edited NPCs.

(A and B) Schematic of the experimental protocols.

(C) Bright-field and GFP images of mini mNPC aggregates in Aggrewell. Scale bars, 500 μm.

(D) Quantification of pHH3⁺ nuclei in LTL⁺/CDH⁺ cells in mini nephron organoids.

(E) Heatmap of selected gene expression as determined by qRT-PCR.

(F) Metabolic analyses of mini nephron organoids using Seahorse assays at baseline (blank boxes) and at stressed levels (filled boxes).

(G) Heatmaps showing the quantification of screen results in cyst formation efficiency (left) and cyst diameter (right). Identified HDAC inhibitors (green), and BRD4 inhibitors (blue).

(H) Venn diagram showing the common small molecules identified in (G).

(I) Western blot analysis of PC2 expression in candidate PKD2^−/− single cell hPSC clones.

(J) Schematic of genotyping results in PKD2^−/− clones.

(K) Bright field images showing cyst formation. Scale bars, 500 μm and 200 μm (enlarged pictures, right panels).

(L) Bright-field images of organoids upon various small molecule treatment. Scale bars, 200 μm.

(M) Quantification of the percentages of cystic organoids shown in (L).

(N) Immunofluorescence images of samples in (L) for TUNEL assay. Scale bars, 50μm.

Data are presented as mean ± SD. Each column represents counts from three biological replicates (n=3). The significance was determined by two-tailed unpaired Student’s t tests; ns, not significant; *, p<0.05; **, p<0.01.

See also Figures S10–S15.

We employed a modified one-vector multiplexed CRISPR-Cas9 system⁶⁷ for one-step gene knockout directly in Cas9-expressing mouse NPCs (Fig. S10A–C, and Methods S1). We then generated Pkd1 and Pkd2 knockout NPCs using this system (Fig. S10D and E), isolated single cell clones (Fig. S10F) and validated successful knockout (Fig. S10G–J). To examine cyst formation, nephron organoids were manually dissected into smaller pieces which were transferred to shaking culture⁶³ (Fig. S11A). Almost all Pkd1^−/− and Pkd2^−/− NPC-derived nephron organoids formed cysts of different sizes, whereas no cysts were formed in EGFP^−/− control organoids (Fig. S11B and C). Cysts were formed following a freeze-and-thaw cycle (Fig. S11D–F). Treatment with CFTRinh172,⁶⁸ metformin,^69,70 and AZ505,^71,72 previously reported to suppress cyst growth, decreased the percentage of organoids undergoing cyst formation and cyst diameter within the organoids (Fig. S11G–I). As expected, these AVPR2 negative nephron organoids did not respond to the AVPR2 inhibitor tolvaptan.^73,74 To scale up the model, we used AggreWell plates to mass produce mini 3D NPC aggregates. After aggregation, the mini NPC aggregates were transferred into shaking culture, the setting for nephron induction and cyst formation (Fig. 7A and C). Under the optimized culture condition (Fig. S12A–F, and Methods), Pkd1^−/− or Pkd2^−/− NPC aggregates differentiated into cystic nephron organoids as early as 4 days after shaking culture (Fig. S11J–M).

Cultured NPC-derived PKD organoid models recapitulate key molecular, cellular, and metabolic features of ADPKD pathogenesis. Pkd2^−/− cystic organoids showed co-expression of LTL and CDH1 in the majority of the cyst-lining cells (Fig. S13A), increased cellular proliferation (Fig. 7D, S13A and B), enhanced expression of cell cycle genes, and increased activity of the mTOR pathway^75,76 and MYC^77,78 (Fig. 7E and S13C and D). Consistent with the observations in vivo that PKD cyst-lining cells use significantly more energy from glycolysis to support the high metabolic needs for cyst growth,^79,80 SeahorseXFp^™ analysis revealed that both the oxygen consumption rate (OCR) and the extracellular acidification rate (ECAR) were elevated in Pkd2^−/− organoids (Fig. 7F, S13F–I, N, and P). Importantly, the metabolic phenotypes following Pkd2 removal were specific to differentiated epithelium-containing organoids but were not observed in undifferentiated NPCs (Fig. S13E, J–M, O, and P).

As a proof-of-concept for small molecule screening using this PKD organoid model (Fig. 7A), we tested a small molecule library comprising 148 compounds targeting major pathways regulating the epigenome (Fig. 7G, S14A and Methods). Fourteen compounds showed significant inhibititory effects on cyst formation in at least one biological replicate (Fig. 7G and H), while metfomin dose-dependently inhibited cyst formation and tolvaptan had no effect as controls (Fig. S14B). Of the 14 hits, 12 replicated in a second screen (Fig. 7H). Of these, 8 were annotated as HDAC inhibitors and 3 as BRD4 inhibitors, both previously identified as targets for ADPKD.^81,82 The final hit, PTC-209, a specific inhibitor of BMI-1, has not previously been linked to cysts suppression in ADPKD but showed a dose-dependent inhibition in our assay (Fig. S14C–E).

To replicate screening in human PKD2 mutant organoids (Fig. 7B), biallelic frame-shift mutations were generated in PKD2 in the SIX2-GFP hPSC line (Fig. 7I and J). Mini NPC aggregates were generated from these followed by shaking culture and nephron induction (Fig. 7B, S15A, and Methods). After 8 days of shaking culture, cysts were observed in PKD2^−/− NPC-derived nephron organoids, but not in wild-type control organoids. These cysts continued to grow larger with the majority of the cyst-lining cells expressing LTL and CDH1 (Fig. 7K, S15B–D). These cystic organoids respond to CFTRinh172, metformin, AZ505, and tubacin,⁸³ but not tolvaptan (Fig. S15E–G). Incubating with PTC-209 resulted in a dose-dependent cyst inhibitory effect without obvious cellular toxicity (Fig. 7L–N, S15H–J).

DISCUSSION

The in vitro NPC culture system will facilitate future mechanistic studies on the specific actions of p38 MAPK and YAP activity in expansion of mouse and human NPCs. Surprisingly, hNPSR-v2 supported the dedifferention of committed nephron progenitors (WNT4⁺) and cells undergoing podocyte development (MAFB⁺ and NPHS1⁺) to NPC-like cells. These observations support in vivo observations of a developmental plasticity within induced Wnt4⁺ NPCs in the developing mouse kidney.²⁵ How this plasticity operates at the molecular level, to which point in the podocye developmental program cells retain plasticity, and how patient-specific models of podocytopathies might be derived from reprogrammed adult podocytes, are interesting future questions.

Our observation that hPSC-derived iNPCs, without further culture in hNPSR-v2, are not fully programmed to the NPC stage, agrees with a recent report where an improvement in NPC specification led to improved proximal tubule specification.¹⁶ We observed iNPCs, with transcriptome much closer to primary NPCs, generate nephron organoids with more mature podocytes expressing COL4A3⁺/COL4A4⁺ hitherto only reported in human kidney organoids following implantation into the mouse kidney.^57,84,85 Given that podocytes are the kidney cell type most associated with recessive genetic disease,⁸⁶ including deficiencies in COL4A3/4/5 associated with Alport syndrome, the culture system offers opportunites for disease modeling.

Limitations of the Study

Genetic dependencies revealed by CRISPR screening in vitro may include “synthetic” dependencies of in vitro culture in addition to genes with similar in vivo actions. While the iNPC-derived nephron organoids showed enhanced podocyte maturation over kidney organoid systems, there is much room for improvement in our (and other) organoid differentiation systems to generate a full range of nephron cell types, most notably for the loops of Henle, distal convoluted tubules and varied cell types of distal-most connecting segment. Here, starting with a readily manipulable, controlled cell base of iNPC cells offers advantages over “conventional” hPSC-directed organoid systems with a complexity of target cell types in screening for appropriate conditions to maximize iNPC developmental outcomes.

STAR METHODS

RESOUCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Zhongwei Li (zhongwei.li@med.usc.edu).

Materials availability

Cell lines, plasmids, and other unique resources generated in this study are available from the lead contact, Zhongwei Li (zhongwei.li@med.usc.edu), with a completed Materials Transfer

Agreement.

Data and code availability

Bulk RNA-seq and CRISPR screen data have been deposited at GEO (GSE230707). Single Cell Multiome data have been deposited at GEO (GSE251862). This study does not report original code, any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Human tissues

11.4 week and 17.4 week human fetal kidney samples used in this study were collected under Institutional Review Board approval (USC-HS-13–0399 and CHLA-14–2211). Following the patient decision for pregnancy termination, the patient was offered the option of donation of the products of conception for research purposes, and those that agreed signed an informed consent. This did not alter the choice of termination procedure, and the products of conception from those that declined participation were disposed of in a standard fashion. The only information collected was gestational age and whether there were any known genetic or structural abnormalities.

Mice

All experiments on animals were performed in accordance with institutional guidelines and IACUC protocol (USC IACUC Protocol # 20829) approved by University of Southern California. 6–8 weeks old female Swiss Webster mice were purchased from Taconic Biosciences (Model # SW-F). 2–8 months old male heterozygous Six2^{tm3(EGFP/cre/ERT2)Amc} mice (Six2^GCE, JAX # 009600), and Gt(ROSA)26Sor^{tm1.1(CAG-cas9*,-EGFP)Fezh} mice (Cas9-GFP, JAX # 026179), were crossed with Swiss Webster mice (6–8 weeks old, female) to obtain timed-pregnant embryos to derive Six2-GFP and Cas9-GFP NPC lines, respectively. 2–8 months old male heterozygous Tg(Hoxb7-Venus*)17Cos mice (Hoxb7-Venus, JAX # 016252) were crossed with Swiss Webster mice (6–8 weeks old, female) to obtain timed-pregnant E11.5 embryos to derive UB organoids. Six2-tdTomato (Six2-tdT) mice were generated through crossing 2–8 months old male heterozygous Six2^GCE mice with 6–8 weeks old female homozygous Gt(ROSA)26Sor^{tm14(CAG-tdTomato)Hze} mice (tdTomato reporter mice, JAX # 007908). Wnt4-tdTomato (Wnt4-tdT) mice were generated through crossing 2–8 months old male heterozygous Wnt4^{tm2(EGFP/cre/ERT2)Amc} mice (Wnt4^GCE, JAX # 032489) with 6–8 weeks old female homozygous tdTomato reporter mice. All the mice were maintained on a 12 hours: 12 hours light-dark cycle with food and water ad libitum.

hPSC lines

Experiments using hPSCs were approved by the Stem Cell Oversight Committee (SCRO) of University of Southern California under protocol # 2018–2. Human pluripotent stem cells are routinely cultured in mTeSR1 (STEMCELL Technologies #85850) or mTeSR1 Plus (STEMCELL Technologies #100–0276) medium in monolayer culture format coated with Matrigel and passaged using dispase as previously described,⁸⁷ or using Versene Solution (Thermo Fisher # 15040066) following manufacturer’s protocols.

METHOD DETAILS

Deriving clonal expandable NPC lines from any mouse strain with mNPSR-v2 medium

Deriving NPC lines from Six2-GFP mice

Timed pregnant E12.5-E18.5 kidneys were isolated from Six2-GFP (Six2^GCE, JAX # 009600) mouse embryos, and then the kidneys were minced into small pieces. The minced kidney pieces were transferred into 1.5mL Eppendorf tubes and spun down at 300 g for 3 minutes. (Note that we aliquoted the kidney pieces into multiple tubes to ensure the volume of the tissue pellet after centrifugation was less than 100 μl per tube to ensure best dissociation). The dissection medium was then carefully aspirated. The kidney pieces were then washed once with sterile PBS. 500 μL of pre-warmed Accumax Cell Dissociation Solution (Innovative Cell Technologies, Cat. No. AM-105) was added to the tube to resuspend the kidney pieces. The tube was incubated in 37°C incubator for 20~22 minutes. 500 μL 10% FBS medium (10% FBS in DMEM) was added to the tube, and then GENTLY pipetted up and down 20 to 25 times to further dissociate the kidney pieces. The tube was spun down and the supernatant was then removed. FACS medium (cold PBS with 2% FBS) was added to resuspend the cell pellet and the cell suspension was filtered through 40 μm cell strainer (Greiner bio-one, Cat. No. 542040) to remove cell clumps before FACS to sort out Six2-GFP⁺ NPCs. Purified Six2-GFP⁺ NPCs were counted with TC20^™ Automated Cell Counter (Bio-Rad, Cat. No. 1450102), and then cultured in mNPSR-v2 medium (Methods S1) to establish NPC lines. Take 96-well plate for example, we counted and seeded 5,000 FACS-purified NPCs into one Matrigel-coated well with 100 μl mNPSR-v2 medium. The medium was refreshed 2 days after cell seeding. When NPCs grew to 80–90% confluent (on day 3 or day 4), the culture medium was removed, and the cells were washed once with 100 μl PBS, and then dissociated with 50 μl pre-warmed Accumax. The cells were incubated with Accumax for 8 mins, and then 150 μl 10% FBS medium was added to neutralize Accumax. The medium was pipetted up and down GENTLY for 5–7 times to make single cell suspension, which was then seeded at 1:20–1:30 passage ratio to a new well in 96-well plate. Change medium 2 days after seeding. On day 3 or day 4, the cells grew to 80–90% confluent and can be passaged again using the same protocol described above. (Note, for coating with Matrigel (R&D Systems, # 3433-010-01) in one well of 96-well plate, dissolve 1 mg Matrigel into 25 ml cold DMEM/F12, and then aliquot 100 μl medium into each well. The plate is then incubated in 37°C for at least 1–2 hours before the coating medium is aspirated followed by cell seeding.)

Deriving NPC lines from E11.5 metanephric mesenchyme (MM)

E11.5 kidneys were isolated and MM was manually dissected out from the E11.5 kidneys following our previously described protocol to isolate E11.5 UB and MM (Zeng et al., 2021). 20 isolated MM were pooled together and dissociated in 500 μl Accumax for 8–10 minutes (scale it down if less MM were isolated) before 500 μl 10% FBS medium was added to neutralize Accumax. The medium was then pipetted up and down GENTLY for 7–10 times to dissociate the MM into single cells. 5,000 cells were then seeded into one well in a Matrigel-coated 96-well plate (scale up or down based on the surface area, if different culture format was used) in mNPSR-v2 medium to derive NPC line, using similar protocols described above starting from Six2-GFP⁺ NPCs.

Deriving NPC lines from whole kidneys

E12.5-E15.5 kidneys were isolated, minced into small pieces, and dissociated in Accumax as described above for dissociating E12.5-E18.5 Six2-GFP kidneys. 5,000 cells were then seeded into one well in a Matrigel-coated 96-well plate (scale up or down if different culture format was used) in mNPSR-v2 medium to derive NPC line, using similar protocols described above starting from Six2-GFP⁺ NPCs. Note that, different from NPC line derivation from Six2-GFP⁺ NPCs, or from isolated E11.5 MM, whole kidney cells need to go through 2–3 passages to enrich the NPC population in the culture before a stable NPC line can be established with 90–95% purity.

Deriving single cell clonal mouse NPC lines

NPC lines can be derived from FACS-purified Six2-GFP+ NPCs, isolated E11.5 MM, or whole kidney cells as described above. When NPC lines are stably established, single cell clonal NPC lines can be generated. For that, when the NPC lines grew to around 80% confluency, the cells were dissociated into single cells following the protocol described above for passaging NPCs. The cells were counted and seeded into Matrigel-coated 96-wells at the density 0.5 cell per well with 100 μl mNPSR-v2 medium so that most wells would have either one single cell or no cell (day 0). On day 3, 50 μl used medium was removed, and 100 μl fresh medium was added. On day 6, the used medium was completely removed and 100 μl fresh medium was added. On day 9, NPC clones were clearly observed under the microscope in about 30% of the wells seeded. Cells in these wells were dissociated following our protocol described above, and all the cells were seeded into another 96-well. On day 11, the cells reached around 80–90% confluency and were passaged routinely thereafter every 3 days as described above. To determine cloning efficiency, 60 wells of a 96-well plate were seeded with single cell NPC following the method mentioned above (D0). On day 3, the wells containing clusters of cells from a single cell were labeled and counted, and the wells without any cell were discontinued. On day 9, the number of wells grown at least 50% confluency was counted. Cloning efficiency was calculated by using the number of wells recorded on day 9 divided by the number of wells recorded on day3, from three independent experiments.

Generation of mouse nephron organoids from mouse NPC lines

mNPCs cultured in mNPSR-v2 medium were dissociated into single cells using Accumax as described above. 30,000 cells were seeded into one well of a U-bottom 96-well plate (Thermo Fisher Scientific, # 174929) and cultured with 100 μl mNPSR-v2 medium overnight for cells to aggregate. Nephron organoids were formed thereafter following the protocol we described previously.⁸⁷ Briefly, on the next day (Day 0), 3D NPC aggregates were transferred onto 6-well format transwell membrane (Corning, # 3450) with 1.2 mL KR5-CF medium at the bottom chamber and cultured for 2 days. On Day 2, change medium to 1.2 mL KR5 medium and change medium every other day. Samples were harvested on Day 7 for various assays.

KR5 medium

Basal medium: DMEM/F12 (1:1) (1 X), Invitrogen, Cat. No. 11330–032.

Supplements:

Reagent Name	Company	Cat. No.	Final Concentration
GlutaMAX-I (100 X)	Invitrogen	35050–079	1 X
MEM NEAA (100 X)	Invitrogen	11140–050	1 X
2-Mercaptoethanol (55 mM)	Invitrogen	21985–023	0.11 mM
Pen Strep (100 X)	Invitrogen	15140–122	1 X
KSR (KnockOut™ Serum Replacement)	Invitrogen	10828–028	5 %

KR5-CF medium

KR5 medium with 4.5 μM CHIR99021 (C) and 200 ng/mL FGF2 (F).

Spinal cord induction assay

After mNPSR-v2 cultured mNPCs reached 80–90% confluency in 2D culture, cells were dissociated into single cells using Accumax Cell Dissociation Solution. 30,000 cells were seeded into each well of U-bottom 96-well plate with 100 μl mNPSR-v2 medium and cultured overnight for re-aggregation. On the next day (Day 0), spinal cord was isolated from E12.5 embryo, and 3D mNPC aggregates were transferred to 6-well format transwell membrane tightly close to spinal cord dorsal part with 1.2mL KR5 in bottom well. Medium was changed every other day with fresh KR5 medium, and the samples were harvested for various assays on day 7.

Mouse engineered kidney generation from cultured NPCs and UB

Mouse UB (mUB) was cultured as we previously described.⁹⁸ The day before mouse kidney reconstruction, mNPCs were dissociated and 50,000 cells were seeded into one well of U-bottom 96-well low-attachment plate with 100μL mNPSR-v2 medium and cultured overnight in 37°C incubator to generate 3D mNPC aggregate. A small piece (with 6–10 branching tips) of day 5–10 cultured mUB organoid was manually dissected out using sterile needles and inserted into a microdissected hole in a 3D mNPC aggregate (sterile needle was used to pierce a hole in the center of mNPC aggregate sphere). This structure was then transferred into a well of a U-bottom 96-well low-attachment plate with 100 μl kidney reconstruction medium (APEL2 + 0.1 μM TTNPB) plus 10 μM Y27632, using a P200 pipette with the top 0.5–1 cm of the tip cut to widen the tip, and cultured in 37°C incubator (day 0). After 24 h (day 1), the reconstructed UB/NPC structure was then transferred onto a 12-well transwell insert membrane (Corning, Cat. No. 3460). 300 μl kidney reconstruction medium was added to the lower chamber of the transwell. Medium was changed every two days for a total of 7 days while the reconstructed kidney branching and maturation progressed.

Chicken chorioallantoic membrane (CAM) assay

CAM assay was conducted as a cost-effective method to evaluate the nephrogenic potential of cultured NPCs in vivo. Briefly, Cas9-GFP mNPC lines were dissociated into single cells, and then 30,000 cells were seeded into 96-well U-bottom low-attachment plate (Thermo Fisher Scientific, # 174929) in mNPSR-v2 medium to generate 3D NPC aggregate. On the next day (D0), change medium with 100ul KR5-CF medium and follow by continuous two days culture (D0-D2). On day 2, transfer 3D NPC aggregates into chicken chorioallantoic membrane. On day 5 and day 7, use microscope to take videos to record the vasculature system in implants. On day 7, harvest the samples for various assays.

Dissociation of postnatal P3-P7 mouse kidneys

P3-P7 pups were euthanized and the kidneys were dissected. Mince one kidney into small pieces using blade (Cincinnati Surgical, # 75870–580). Collect small amount of kidney pieces into an Eppendorf tube (after spinning down, the pellet volume should be less than 30 μl). Spin down at 300g for 5min, remove the supernatant and resuspend the pellet with 500 μl warmed-up FRESH 1X Collagenase IV (Thermo Fisher Scientific, # 17104019). Put the Eppendorf tube into rotating shaker (Eppendorf ThermoMixer^® F2.0) set at 37C with 800rpm shaking speed. Every 10min, take out the tube and pipet up and down the samples for 10 times. Dissociation in Col IV for a total of 30min. After 30min and the 3^rd pipetting, spin the cells down at 300g for 1min to collect the kidney tubules and glomerulus as the pellet. Resuspend the pellet with warmed-up 500ul Accumax. Put the Eppendorf tube into rotating shaker set at 37C with 800rpm shaking speed. Every 10min, take out the tube and pipet up and down the samples for 10 times. Dissociation in Accumax a total of 30min. After 30min, add 500 μl 10% FBS to neutralize the dissociation. Then start the 3rd pipetting for 20 times. Spin down the cells at 300g for 5min, remove supernatant, and resuspend the pellet with 200 μl ACK lysing buffer (Thermo Fisher Scientific, # A1049201), to remove the blood cells (room temperature for 3min). Add 1ml 10% FBS and spin the cells down at 300g for 5min. Remove the 1.2ml supernatant and resuspend the cells into 100 μl 10% FBS medium for cell counting (use Trypan Blue to determine the live cell percentage). Usually, the live cell percentage of dissociated cells was 50–75%.

Generation of SIX2-GFP/PAX2-mCherry knock-in dual reporter hPSC line

CRISPR-Cas9 based genome editing was used to insert 2A-EGFP-FRT-PGK-Neo-FRT or 2A-mCherry-loxP-PGK-Neo-loxP cassette downstream of the stop codon (removed) of endogenous SIX2 or PAX2 gene, respectively. DNA sequences ~1 kb upstream and ~1 kb downstream of the stop codon for endogenous SIX2 (upstream F: CCGGAATTCTGCCCAGTTTGGAGCTACAG; up-stream R: TACGAGCTCGGAGCCCAGGTCCACGAGGTT; downstream F: CGCGTCGACAACCCATTTGCCTTGATGAG; downstream R: CCCAAGCTTCCCGAA-GAACATTCACATGAGG) or PAX2 (upstream F: GAAGTCGACTTTCCACCCATTAGGGGCCA; up-stream R: TATGCTAGCGTGGCGGTCATAGGCAGCGG; downstream F: TATAC-GCGTTTACCGCGGGGACCACATCA; downstream R: GACGGTACCAGTAACTGCTGGAG-GAAGAC) were cloned as homology arms upstream and downstream of 2A-EGFP-FRT-PGK-Neo-FRT (for SIX2-GFP) or 2A-mCherry-loxP-PGK-Neo-loxP cassette (for PAX2-mCherry), respectively, to facilitate homologous recombination. 2A-EGFP fragment was cloned from pCAS9_GFP (Addgene # 44719) and the FRT-PGK-Neo-FRT cassette was cloned from pZero-FRT-Neo3R (kindly provided by Dr. Keiichiro Suzuki). 2A-mCherry-loxP-Neo-loxP fragment was cloned from Nanog-2A-mCherry plasmid (Addgene # 59995). The different fragments were then cloned to a modified pUC19 plasmid with additional restriction sites inserted, to make the complete donor plasmids for both knock-in experiments. Oligos for making sgRNA-expressing plasmid for SIX2 knockin (F: CACCGGGGCTCCTAGAACCCATTTG; R: AAAC CAAATGGGTTCTAGGAGCCCC) or PAX2 knockin (F: CACCGATGACCGCCACTAGTTACCG; R: AAACCGGTAACTAGTGGCGGTCATC) were synthesized, annealed, and cloned into the pSpCas9(BB)-2A-Puro (PX459) V2.0 plasmid (Addgene # 62988). Both donor and sgRNA plasmids for SIX2 reporter knockin were transfected into the H1 hESCs using the Lipofectamine 3000 Transfection Reagent (Invitrogen, Cat. No. L3000015). Neomycin-resistant single cell colonies were picked up manually and genotyping was performed based on PCR. Clones with biallelic knock-in of SIX2-GFP were chosen for second round screen where plasmid encoding flippase was delivered via transfection to allow the transient expression of flippase, whose activities excise the FRT-flanked PGK-Neo cassette from the SIX2-GFP knock-in alleles. PCR was performed to identify single cell clones in which PGK-Neo cassettes were excised from both alleles. Then the same strategy was used to knock in PAX2 reporter based on the successful biallelic SIX2-GFP knock-in clones.

PKD2 knock-out in SIX2-GFP knock-in reporter hPSC line

SIX2-GFP knock-in reporter hPSC line was generated using the method described above. We further knocked out PKD2 gene in the reporter hPSC line using CRISPR/Cas9 based genome editing. For that, sgRNA was designed to target the first exon of human PKD2 gene with sgRNA targeting sequence: CCGCGATAACCCCGGCTTCG. sgRNA was inserted into lentiCRISPR v2 plasmid (Addgene # 52961) and lentivirus was produced and then infected into SIX2-GFP hPSC line. After puromycin selection followed by clonal expansion of hPSCs, 11 single cell clones were picked up and expanded. Proteins were extracted from the 11 single cell clones and Western blot was performed to identify the candidate PKD2^−/− clones #10 and #11. PCR-based genotyping from the genomic DNA, following by Sanger sequencing, further confirmed the generation of frame-shift mutations in both alleles of PKD2 gene from #10 and #11 clones.

Deriving iNPC lines from human pluripotent stem cells

Deriving iNPC lines from FACS-purified SIX2-EGFP/PAX2-mCherry iNPCs

Directed differentiation from SIX2-GFP/PAX2-mCherry knock-in dual reporter hPSCs into iNPCs was performed following a previously published protocol,⁸ with minor modifications. Briefly, hPSCs were dissociated into single cells with Accumax and 40,000 cells were seeded into one well in a 12-well plate with 1ml mTeSR medium plus 10 μM Y27632. Medium was changed daily with fresh 1ml mTeSR medium without Y27632 for another 2 days. At this time (day 0), hPSCs formed small colonies and were ready for directed differentiation. Phase 1, day 0 to day 4 (D0-D4), hPSCs were cultured with 1 ml Advanced RPMI 1640 Medium (Thermo Fisher Scientific, # 12633–012) supplemented with 8 μM CHIR99021 and 10nM LDN193189; medium was refreshed on D2 and D3. Phase 2 (D4-D7), medium was changed to Advanced RPMI 1640 Medium supplemented with 10 μM Y27632 and 10 ng/ml activin A; medium was refreshed daily. Phase 3 (D7-D10), medium was changed to Advanced RPMI 1640 Medium supplemented with 50 ng/ml FGF9; medium was changed daily till D10. On D10, cells were dissociated into single cells with pre-warmed Accumax, and SIX2-GFP/PAX2-mCherry iNPCs were sorted out through BD FACSAria^™ III Cell Sorter. Sorted iNPCs were counted and 10,000 cells were seeded into one well in a 96-well plated coated with Matrigel and cultured with 100 μl hNPSR-v2 medium (Methods S1). The medium was refreshed 2 days after cell seeding. When iNPCs grew to 80–90% confluent (on day 3 or day 4), the culture medium was removed, and the cells were washed once with 100 μl PBS, and then dissociated with 50 μl pre-warmed Accumax. The cells were incubated with Accumax for 8 mins, and then 150 μl 10% FBS medium was added to neutralize Accumax. The medium was pipetted up and down GENTLY for 5–7 times to make single cell suspension, which was then seeded at 1:10 passage ratio to a new well in 96-well plate. Change medium 2 days after seeding. On day 3 or day 4, the cells grew to 80–90% confluent and can be passaged again using the same protocol described above. (Note, Matrigel coating protocol for iNPC culture is the same as the one described for mNPC culture.)

Deriving iNPC lines without prior FACS-based purification of iNPCs

hPSCs were differentiated following the same protocol described above to generate iNPCs. On D10 of differentiation, instead of using FACS to purify the iNPCs based on SIX2-GFP; PAX2-mCherry dual reporter system, the dissociated whole cells were counted and then 10,000 cells were seeded into one well of 96-well coated with Matrigel and cultured with 100 μl hNPSR-v2 medium. Thereafter, after 2 weeks of continuous culture and passage in hNPSR-v2 medium following protocol described above, stable iNPC line was established with 90–95% purity of iNPCs.

Deriving single cell clonal iNPC lines

When iNPC lines are stably established, single cell clonal iNPC lines can be generated. For that, when the iNPC lines grew to around 80% confluency, the cells were dissociated into single cells following the protocol described above for passaging iNPCs. The cells were counted and seeded into Matrigel-coated 96-wells at the density 0.5 cell per well with 100 μl hNPSR-v2 medium so that most wells would have either one single cell or no cell (day 0). On day 3, 50 μl used medium was removed, and 100 μl fresh medium was added. On day 6, the used medium was completely removed and 100 μl fresh medium was added. On day 9, NPC clones were clearly observed under the microscope in about 30% of the wells seeded. Cells in these wells were dissociated following our protocol described above, and all the cells were seeded into another 96-well. On day 12–14, the cells reached around 80–90% confluency and were passaged routinely every 3 or 4 days at the ratio of 1:10 as described above for bulk iNPC culture. To determine cloning efficiency, 60 wells of a 96-well plate were seeded with single cell NPC following the method mentioned above (D0). On day 3, the wells containing clusters of cells from a single cell were labeled and counted, and the wells without any cell were discontinued. On day 9, the number of wells grown at least 40% confluency was counted. Cloning efficiency was calculated by using the number of wells recorded on day 9 divided by the number of wells recorded on day3, from three independent experiments.

Deriving hNPC lines from human fetal kidneys

Tweezers were utilized to dissect nephrogenic zones from 9 to 18 week-old human fetal kidneys. Nephrogenic zones were minced into small pieces, and transferred into 1.5 mL Eppendorf tubes, and spun down at 300 g for 3 minutes. (Note that we aliquoted the kidney pieces into multiple tubes to ensure the volume of the tissue pellet after centrifugation was less than 100 μl per tube to ensure best dissociation). The dissection medium was then carefully aspirated. The kidney pieces were then washed once with sterile PBS. 500 μL of pre-warmed Accumax was added to the tube to resuspend the kidney pieces. The tube was incubated in 37°C incubator for 20~22 minutes. 500 μL 10% FBS medium (10% FBS in DMEM) was added to the tube, and then GENTLY pipetted up and down 20 to 25 times to further dissociate the kidney pieces. The tube was spun down and the supernatant was then removed. FACS medium (cold PBS with 2% FBS) was added to resuspend the cell pellet and the cell suspension was filtered through 40 μm cell strainer (Greiner bio-one, Cat. No. 542040) to remove cell clumps. Cells were counted and diluted to 1 million cells per 100 μl in FACS medium, and ITGA8 antibody (R&D Systems, # AF4076) was added at the ratio of 1:200 for live cell staining. The incubation with ITGA8 antibody was performed on ice for 1 hour, followed by wash with 1 ml FACS medium and incubation for 1 hour with fluorescence protein-conjugated secondary antibody (Donkey anti-Goat IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor Plus 488 or 568, Thermo Scientific, # A11055 or A11057) protected from light. After that, another wash with 1 ml FACS medium was performed and the cell suspension was then subjected to FACS to sort out ITGA8⁺ hNPC through BD FACSAria^™ III Cell Sorter. Purified ITGA8⁺ hNPCs (70–90% SIX2⁺/PAX2⁺) were cultured with the same protocol as described for purified SIX2-GFP; PAX2-mCherry iNPCs. After 2 to 3 passages in hNPSR-v2 medium, stable hNPC line with 90–95% hNPCs were established.

Lentiviral expression of mCherry in iNPCs

30,000 SIX2-GFP⁺/PAX2-mCherry⁺ iNPCs were seeded into one well in 24-well plate. 24 hours later, iNPCs were infected with lentivirus expressing mCherry driven by EF1α promoter (EF1a_mCherry_P2A_Hygro, Addgene, # 135003) using spinfection method described above. 24 hours after infection, 150 μg/ml Hygromycin B (Thermo Scientific, Cat. No. 10687010) was added to the medium to select for iNPCs that have been successfully infected. More than 95% of the iNPCs showed bright mCherry expression after 1 week of Hygromycin B selection. (Note that exogenous mCherry is much brighter than endogenous mCherry from the PAX2-mCherry reporter, allowing the separation of these two mCherry signals.)

Targeted genome editing at the AAVS1 loci in iNPCs

30,000 SIX2-GFP⁺/PAX2-mCherry⁺ iNPCs were seeded into one well in 24-well plate. 24 hours later, iNPCs were transfected with a mixture of two plasmids that provide donor DNA for targeted knockin of CAG promoter-driven mCherry expression cassette at the AAVS1 loci (pAAVS1-P-CAG-mCherry, Addgene, # 80492), and that express Cas9 and sgRNA (pXAT2, Addgene, # 80494), at the ratio of 3:1, using Lipofectamine 3000 transfection reagent. 24 hours after transfection, 0.3 μg/ml puromycin was added to the medium to select for iNPCs that have been successfully gene edited. More than 97% of iNPCs showed bright mCherry expression after 1 week of puromycin selection. (Note that exogenous mCherry is much brighter than endogenous mCherry from the PAX2-mCherry reporter, allowing the separation of these two mCherry signals.)

Generation of human nephron organoids from human NPC lines

Nephron organoid protocol – v1:

iNPCs or hNPCs cultured in hNPSR-v2 medium were dissociated into single cells using Accumax as described above. 30,000 cells were seeded into one well of a U-bottom 96-well plate (Thermo Fisher Scientific, # 174929) and cultured with 100 μl hNPSR-v2 medium overnight for cells to aggregate. On the next day (Day 0), 3D NPC aggregates were transferred onto 6-well format transwell membrane (Corning, # 3450) with 6 μM CHIR99021 in 1.2 mL STEMdiff^™ APEL^™2 medium (STEMCELL Technologies, # 05270) at the bottom of the chamber for 1-hour. Then medium on the bottom chamber was removed as much as possible, and 1.2mL STEMdiff^™ APEL^™2 medium with 50 ng/mL FGF9 and 1 μg/mL heparin was added for continuous culture. This medium was refreshed every other day. On Day 5, medium was changed to 1.2mL STEMdiff^™ APEL^™2 medium without any other factors. The medium was refreshed every other day till Day 14, when the samples were harvested for various assays. Nephron organoids described in Fig. 4 and 7 were generated using this protocol (v1). During manuscript revision, we slightly changed the protocol for better nephron formation. Nephron organoids described in Fig. 5 and 6 were generated following this improved protocol below (v2).

Nephron organoid protocol – v2:

When iNPCs or rNPCs reached 80–90% confluency in hNPSR-v2 culture, they were dissociated into single cells with Accumax as detailed above. 30,000 dissociated iNPCs or rNPCs were seed into one well of a U-bottom 96-well plate (Thermo Fisher Scientific, # 174929) and cultured with 100 μl hNPSR-v2 medium overnight for cells to aggregate. On the next day (Day 0), 3D NPC aggregates were transferred onto 6-well format Transwell plate (Corning, # 3450) with 6 μM CHIR99021 in 1.2 mL STEMdiff^™ APEL^™2 medium (STEMCELL Technologies, # 05270) (Stage I) at the bottom of the chamber for 1 hour. The medium at the bottom chamber was then completely removed, and 1.2mL STEMdiff^™ APEL^™2 medium with 50 ng/mL FGF9 and 1 μg/mL heparin (Stage II) was added for continuous culture. This medium was refreshed every other day. On Day 5, medium was changed to 1.2mL Advanced RPMI 1640 Medium (Thermo Fisher Scientific, #12633–012) with 1X B27 and 200 nM A83–01 (Stage III). The medium was refreshed every other day till Day 14 or Day 21, when the samples were harvested for various assays.

Growth curve analysis of NPCs cultured in mNPSR-v2 or hNPSR-v2 media

For the growth curve analysis, 5,000 mNPCs, iNPCs, or hNPCs were seeded into each well of a 96-well plate coated with Matrigel and cultured in their corresponding mNPSR-v2 or hNPSR-v2 media. For each experiment, 16 wells were prepared for each line tested. Every 24 hours, 4 wells were dissociated, and the cell number was counted till 4 days after initial cell seeding, to determine the cell growth from day 0 to day 4. Three independent experiments were performed for each NPC line, serving as three biological replicates, to calculate the average cell numbers and the error bars at each time point.

Bulk RNA sequencing

Cultured mouse and human NPC samples, time-course iNPC-derived nephron organoid samples, rNPC samples, time-course podocyte-to-rNPC reprogramming samples, were collected and lysed in TRIzol reagent or DNA/RNA Shield and stored at −80°C. Total RNA was extracted using Direct-zol RNA MicroPrep Kit (Zymo) or Quick-RNA Microprep Kit (Zymo). Bulk RNA-Sequencing was then performed through the Molecular Pathology Genomics Core of Children’s Hospital Los Angeles (CHLA), or through Novogene USA Inc. For sequencing through CHLA, cDNA library was prepared using KAPA Stranded mRNA-Seq Kit (KAPA Biosystems) and sequenced using Illumina HiSeq 2500. For sequencing through Novogene, library prep and sequencing were performed using standard Novogene bulk RNA-seq pipeline.

Bulk RNA-seq data analysis

RNA sequencing data was analyzed using Partek Flow Genomic Analysis Software. In addition to the new data generated described in this study, legendary bulk RNA-seq data of primary mouse and human NPCs were used as positive controls, and primary kidney cells that are not NPCs were used as negative controls. These include E12.5 Six2-negative primary mouse non-NPCs, E11.5, E12.5, E13.5, E16.5 and P0 primary mouse NPCs,⁸⁷ SIX2-negative primary human non-NPCs, SIX2+ human NPCs, SIX2+/MEIS1+ human NPCs,⁹⁹ and ITGA8+ human NPCs.⁸⁸ FASTQ files were trimmed from both ends based on a minimum read length of 25 bps and an end minimum quality score (Phred) of 20 or higher. For mouse samples, reads were aligned to mm39 using STAR 2.7.8a. For human samples, reads were aligned to hg38 using STAR 2.7.8a. Aligned reads were quantified to the Partek E/M annotation model. Gene counts were normalized using DESeq2 Median ratio. 972 mouse NPC signature genes were identified by comparing primary Six2⁺ E12.5 NPCs and primary Six2⁻ E12.5 non-NPCs⁸⁷, with cutoff of fold change > 1.5 or < −1.5, and p-value < 0.01 using DESeq2. 1058 human NPC signature genes were identified by comparing 5 datasets of primary ITGA8⁺, SIX2⁺ or SIX2⁺/MEIS1⁺ NPCs with 2 datasets of primary SIX2⁻ non-NPCs,^24,88 with cutoff of fold change > 1.5 or < −1.5, and FDR < 0.05 using DESeq2. Principal component analysis (PCA) for Figure 5B was performed using human nephron lineage signature genes. All other PCA were performed using the mouse or human NPC signature genes identified. Hierarchical clustering of selected genes was produced based on the genes’ DESeq2 Median ratio values, clustering samples and features with average linkage cluster distance and Euclidean point distance. Volcano plots (Figure S9 K and L) were generated using the DEG (differentially expressed genes) between podocytes (D0) and podocytes cultured with hNPSR-v2 for 4 days (K), and between podocytes cultured with hNPSR-v2 for 4 days and 24 days (L), with cutoff of fold change > 2 or < −2, and FDR < 0.05. Gene ontology (GO) analysis (Figure S9 M and N) were performed using Metascape (https://metascape.org) based on the DEG (differentially expressed genes) between podocytes (D0) and podocytes cultured with hNPSR-v2 for 4 days (M), and between podocytes cultured with hNPSR-v2 for 4 days and 24 days (N) with cutoff fold change > 5 or < −5 for D4 vs. D0 (M), or > 2 or < −2 for D24 vs. D4 (N), and p value < 0.01. All processed bulk RNA-seq data are summarized in Table S1.

Single nuclei isolation from iNPC-derived nephron organoids

Expanded iNPCs (D25-D45) were used to generate nephron organoids following the nephron organoid induction protocol (v2, described above). Total 10 organoids were harvested at the Day 21 time point for 10x Genomics Single Cell Multiome (snRNA/snATAC-seq) sequencing. These organoids were first minced into small pieces, transferred to 1.5mL Eppendorf tubes, and washed with 1x DPBS. These organoid pieces were then dissociated into single cells using 10 mg/mL Bacillus licheniformis cold active protease (Sigma P5380) mixed with 2.5 mg/mL collagenase type 4 (Worthington, #LS004188) and 125 U/mL DNase I (Worthington, #LS002058) in 1x DPBS at 12°C. The digestion mix was agitated twenty times every 5 min with wide bore P-200 pipet tip. The dissociation reaction was terminated by quenching with fetal bovine serum (total end concentration 10%) when there were mostly single cells (after around 45 min). The tube was then centrifuged at 300g for 5 mins, supernatant was removed, and the cell pellets were washed twice with DPBS, centrifuged in between, and resuspended in prechilled nuclear lysis buffer as outlined in 10X protocol (CG000365 RevC). After 5 mins of nuclear lysis, the single cell suspension viability went from 99% to less than 3% as assayed with Trypan Blue staining and visualization with a Countess III Cell Counter. The now nuclear pellet was washed and centrifuged three times to remove debris and passed through a 40 μM filter (Pluriselect 43–50040) upon last wash. The final nuclear pellet was resuspended in prechilled 1X Nuclei Buffer (10X Genomics, #2000207).

snRNA-seq and snATAC-seq library construction and sequencing

Nuclei concentration was assessed and a target of ~13,000 × 2 nuclei (in two technical replicates) were loaded into two separate Tn5 transposition reactions. Loading onto Chip J, GEM Generation and Library Preparation followed the manufacturer’s protocol (10X Genomics, CG000338) with 7 preAmp, 7 ATAC Library, 8 cDNA, and 14 SI GEX PCR cycles used. All cleanups were performed using suggested SPRI-Select bead ratios. Final library concentration and quality were assessed by BioAnalyzer. Complementary snRNA (GEX) and snATAC libraries were then sequenced on the Illumina-S4 Novaseq 6000 platform at Novogene.

Data analysis for snRNA-seq/snATAC-seq datasets

Data for integration.

Other organoid transcriptomic scRNA-seq⁵⁷ at differentiation day 28 and snRNA-seq/snATAC-seq⁵⁸ at differentiation day 26 datasets were downloaded and loaded into Seurat version 4.3.0.1 and Signac version 1.10.0. Generation of Seurat objects included quality-control metrics for genes per cell (250–12,000), mRNA transcripts per cell (250–75,000), and maximum mitochondrial percentage (35%) to eliminate low quality cells.

Data integration.

These objects were then prepped for integration with DietSeurat and integrated into one unified snRNA-seq data object through standard Seurat data integration pipeline (found on https://satijalab.org/seurat/archive/v4.3/integration_introduction) using the following functions: SplitObject, NormalizeData, FindVariableFeatures, and SelectionIntegrationFeatures, FindIntegrationAnchors, and IntegrateData. Linear dimension reduction was performed on this integrated object with a total number of 75 PCs during principal component analysis.

Identifying gene markers and calculating cell identity proportions.

Cluster characterization and kidney cell type labeling was conducted by running differential gene expression analysis with FindAllMarkers on the integrated and labeled object, and the top differentially expressed genes were used to identify cell types. Cell count per sample was calculated by first re-naming clusters to cell identities with levels. The integrated and labeled Seurat object was split into individual organoid datasets using SplitObject by original identity. These individual objects were then converted to a dataframe with dplyr version 1.1.2 function bind_rows and base R function as.data.frame. Finally, cell proportions for each organoid identity were converted to percentages based on each individual dataset – e.g. all percentages of the cell identities are unique to each identity. Specific cell type populations were analyzed by sample by subsetting the integrated and labeled Seurat object based on assigned kidney cell type, and gene expression was analyzed with DotPlot grouped by original identity, and/or VlnPlot split by original identity. All processed single-cell analysis data are summarized in Table S3.

Genome-wide CRISPR screen

The Brie genome-wide CRISPR knockout library²⁶ (Addgene # 73632) was introduced to cultured NPCs via lentiviral infection in two different NPC lines as two biological replicates. This library contains 4 sgRNAs for each of the 19,674 protein-coding genes of the mouse genome, and 1,000 non-targeting control sgRNAs, totaling 78,637 unique sgRNAs (Table S2-Tab1). The infection was carried out at a low multiplicity of infection (MOI) of 0.3, to ensure the majority of the NPCs express only one sgRNA. For each experiment, twenty-five million NPCs were used for the initial infection so that at least 100 cells would carry the same sgRNA to protect against random loss of sgRNA if lower cell number was used. 0.3 μg/ml puromycin was added 48 hours after lentiviral infection, to select for the successfully infected cells, which were further cultured continuously for a total of 3 weeks since lentiviral infection. Genomic DNA was extracted after 3 weeks of culture, and targeted PCR was performed to amplify the sgRNA integrated into the genome for next-generation sequencing, following Sequencing Protocol provided by Addgene (“Broad Institute PCR of sgRNAs for Illumina sequencing”). Next-generation sequencing was performed from the Molecular Pathology Genomics Core of Children’s Hospital Los Angeles using Illumina HighSeq 2500.

Genome-wide CRISPR screen data analysis

Normalized read counts for each individual sgRNA in the plasmid library, before and after CRISPR screen, CRISPR screen beta scores (Table S2-Tab2), and the scatterplots of beta scores, were generated using MAGeCKFlute.¹⁰⁰ 1798 genes from CRISPR screen replicate #1, and 1627 genes from CRISPR screen replicate #2, with beta scores > 1.5 or < −1.5, and p-values < 0.05, were identified as potential hits (Table S2-Tab3, 4) for further analyses using Canonical Pathway Analysis tools of the Ingenuity Pathway Analysis (IPA) platform (Table S2-Tab5–7). Full list of reference FGF, Wnt, and LIF signaling pathway-related genes, and the genes identified in the screens, were summarized in Table S2-Tab8–11. To identify potential NPC self-renewal genes in the whole genome, we removed from the CRISPR screen hits essential genes and genes with low gene expression levels in primary NPCs.^12,24 We then overlapped the gene list with two NPC-enriched gene lists—E12.5 Six2⁺ NPC vs. Six2⁻ non-NPC¹² and E16.5 NPC vs. IPC (interstitial progenitor cell).²⁴ With this stringent data filtration method, a short-list of 64 genes were obtained. Full list of NPC-specific genes identified by comparisons between primary E12.5 NPC and primary E12.5 non-NPC,⁸⁷ or between primary E16.5 NPC and primary E16.5 IPC,²⁴ and the 64 highly confident NPC self-renewal hits, were summarized in Table S2-Tab12–14. Full list of reference CAKUT or Wilms tumor-related genes, and the genes identified in the screens, were summarized in Table S2-Tab15, 16.

KMT2A and KAT6A inhibitor treatment experiments

20,000 mNPCs were seeded into each Matrigel-coated well of 96-well plates, and cultured in (1) mNPSR-v2 + DMSO, (2) mNPSR-v2 + 100 nM WDR5 degrader, (3) mNPSR-v2 + 1 μM MLL1 (inh), (4) mNPSR-v2 + 5 μM MOZ-IN-3, or (5) mNPSR-v2 + 5 μM WM-1119. Medium was changed every two days and samples were harvested for immunofluorescence staining and qRT-PCR assay on day 8. Three independent experiments were conducted for each group as biological replicates.

Establishment of one-step multiplexed CRISPR/Cas9 knockout lentiviral plasmids

Three different sgRNAs (sgRNA-A, sgRNA-B, and sgRNA-C) were designed to target EGFP coding sequences,⁶⁷ or the first exons of Pkd1, or Pkd2 genes. These sgRNAs were inserted individually into lentiGuide-Puro plasmid (Addgene # 52963) following the cloning protocols provided from the plasmid depositor. Then, for each target gene, the three sgRNA expression cassettes were subcloned one by one from the three lentiGuide-Puro plasmids into a modified pLKO.1-TRC plasmid (additional multiple cloning site BclI-EsrGI-MluI-NheI-PstI-SalI-XbaI-XmaI was inserted between the original PpuMI and EcoRI sites in the pLKO.1-TRC plasmid (Addgene # 10878)) to make tandem sgRNA expression cassettes in the same lentiviral vector.

sgRNA targeting sequences:

Gene	sgRNA-A	sgRNA-B	sgRNA-C
EGFP	AAGGGCGAGGAGCTGTTCAC	CTGAAGTTCATCTGCACCAC	GGAGCGCACCATCTTCTTCA
Pkd1	GCTGCGCTGACGATGCCGCT	CTGGCCGGAGACCCTGGGCG	AGCGGCCGGAGCAATTGACG
Pkd2	CGAGATGGAGCGCATCCGGC	TCGCCCGCGCCGCGAGCGTC	AGTGGCGCCCGGGCAGTCGG

One-step gene knockout in mouse NPCs with multiplexed CRISPR/Cas9 KO system

Multiplexed CRISPR/Cas9 KO plasmids targeting EGFP, Pkd1, or Pkd2 were generated as described above. Lentivirus was produced from these plasmids and used to infect mNPCs. For that, lentivirus was first packaged following protocols we described previously⁸⁷ and was then concentrated 100x using Lenti-X Concentrator kit (Takara, # 631231). Concentrated lentivirus was aliquoted and stored in −80°C before use. The lentivirus was used at 1x final concentration together with 10 μg/ml polybrene (Sigma-Aldrich, Cat. No. TR-1003-G) diluted in mNPSR-v2. Lentiviral infection was conducted in mNPC cultured in wells of 96-well plate. For gene editing in Cas9-EGFP mNPC line, used culture medium was removed from each well and 100 μl lentivirus-polybrene-mNPSR-v2 mixture was added to the well. The plate was then centrifuged at 800 g for 15 minutes at room temperature. After the spinfection, the lentivirus-polybrene-mNPSR-v2 mixture was removed and the infected mNPCs were washed three times gently with 150–200 μl pre-warmed PBS, then cultured in 100 μl fresh mNPSR-v2 medium. 24 hours after infection, 0.3 μg/ml puromycin was added to the medium to select for NPCs that have been successfully infected. Once bulk Pkd1^−/− or Pkd2^−/− mNPC lines were generated, single cell clonal Pkd1^−/− or Pkd2^−/− mNPC lines were generated following the same protocols we described above for generating clonal mNPC lines from mNPCs.

Genotyping of Pkd1^−/− or Pkd2^−/− single cell clonal mNPC lines

Genomic DNA of Pkd1^−/− or Pkd2^−/− single cell clonal mNPC lines were extracted using QuickExtract^™ DNA Extraction Solution (Lucigen, # QE09050). PCR primers were designed to flank the sgRNA targeting sites of Pkd1 or Pkd2 genes (listed below). PCR were performed and PCR amplicons were used to conduct gel running and purified by DNA Clean & Concentrator kit (Zymo Research, # D4031). A-tailing were then performed on the purified PCR products following the NEB protocol: https://www.neb.com/protocols/2013/11/01/a-tailing-with-taq-polymerase. A-tailed PCR products were ligated with digested linear pGEM^®-T Vector (Promega, Cat. No. A3600), followed by transformation with One Shot^™ TOP10 Chemically Competent E. coli (Thermo Fisher Scientific, # C404010). Transformed E. coli were evenly daubed on IPTG/X-gal/Amp agarose plates and incubated at 37°C overnight. White bacterial clones were then picked up, inoculated, followed by mini-prep plasmid extraction (Zymo Research, # D4020). Extracted plasmids were sent to for Sanger sequencing. SnapGene software was utilized to analyze sequencing data.

Primers for PCR-based Pkd1^−/− or Pkd2^−/− genotyping:

Gene	Forward	Reverse
Pkd1	CCCTCCTGAACTGCGGCT	GGACCCAGTCATGATGCTCTA
Pkd2	GCAAGCTACCCCGTAGGAATG	GCGCAGGCAGTTGTCAAGC

Optimization of mini Pkd2^−/− cystic organoid generation

Clonal Pkd2^−/− mNPC lines and GFP^−/− mNPC lines were used to optimize mini organoids model. For the first round of optimization, after mNPSR-v2 cultured NPCs reached 80–90% confluency in 2D culture, cells were dissociated into single cells using Accumax Cell Dissociation Solution. 500,000 cells were seeded into each well of AggreWell^™800 24-well Plate (STEMCELL Technologies, Cat. No. 34850) with 2 mL mNPSR-v2 medium and cultured overnight. On the next day (Day 0), mini 3D NPC aggregates formed, and they were transferred into 12-well plates with around 30 mini aggregates/well in 1mL medium. mini 3D NPC aggregates were treated with different concentrations of CHIR99021 at 3.0 μM, 4.5 μM or 6.0 μM in KR5 or hBI medium under shaking culture at 120 rpm (VWR Orbital Shaker Model 1000) for 2 days (D0-D2), and then followed by culture with KR5 or hBI for 5 days (D2-D7). For the second round of optimization, mini 3D NPC aggregates were transferred into 12-well plates or ultra-low attachment plates under shaking culture or suspension culture, and treated with CHIR99021 at 4.5 μM in KR5 medium with 1% MTG or without MTG for 2 days (D0-D2), and then followed by culture in KR5 with 1% MTG or without MTG for 5 days (D2-D7). Cystic organoid formation efficiency was quantified based on 3–4 independent experiments and cyst diameters were measured by the measure tools of Olympus Cellsens Standard software. See also Figure S12 for more information.

hBI medium

Basal medium: DMEM/F12 (1:1) (1 X), Invitrogen, Cat. No. 11330–032.

Supplements:

Reagent Name	Company	Cat. No.	Final Concentration
GlutaMAX-I (100 X)	Invitrogen	35050–079	1 X
MEM NEAA (100 X)	Invitrogen	11140–050	1 X
2-Mercaptoethanol (55 mM)	Invitrogen	21985–023	0.11 mM
Pen Strep (100 X)	Invitrogen	15140–122	1 X
B-27 Supplement (50 X), minus vitamin A	Invitrogen	12587–010	1 X
ITS Liquid Media Supplement (100 X)	Sigma	I3146–5ML	1 X

Generation of PKD organoid models from mouse and human NPC lines

Traditional PKD organoid models:

Clonal Pkd2^−/− mNPC lines or PKD2^−/− iNPC lines were used to generate corresponding mouse or human ADPKD models. The protocols for generation of mouse and human nephron organoids on transwell plates from NPC lines were described above. On day 7 (mouse) or day 14 (human) of organoid differentiation, each organoid was cut into 6 pieces using needles and subjected to shaking culture at 120 rpm (VWR Orbital Shaker Model 1000) in KR5 medium to develop cysts. Cystic organoid formation efficiency was quantified based on 3–4 independent experiments and cyst diameters were measured by the measure tools of Olympus Cellsens Standard software. For small molecule testing using this traditional PKD organoid model, for each small molecule treatment group, 30 small organoid pieces from fully developed mouse or human nephron organoids were pooled together in one well of a 12-well plate, cultured in KR5 medium with shaking, and supplemented with the small molecule to be tested. Small molecules were freshly added to the medium and were refreshed with KR5 medium every two days. 4 days after small molecule treatment, cystic organoids were recorded through imaging and cyst formation efficiency and cyst diameter were quantified. Concentrations of small molecules used in this study were determined based on previous publications.^61,101–103 Cystic organoid formation efficiency was quantified based on 3–4 independent experiments and cyst diameters were measured by the measure tools of Olympus Cellsens Standard software.

Scalable mini PKD organoid models:

Clonal Pkd2^−/− mNPC lines or PKD2^−/− iNPC lines were used to generate corresponding mouse or human ADPKD mini organoid models. For that, 500,000 mNPCs or iNPCs were seeded into each well of AggreWell^™800 24-well Plate (STEMCELL Technologies, # 34850) with 2 mL mNPSR-v2 medium (mouse) or hNPSR-v2 medium (human), and cultured overnight to form mini aggregates (~300 mini aggregates per well, or ~7,200 mini aggregates per plate). On the next day (Day 0), mini 3D NPC aggregates from one well were transferred into 2 wells of 6-well plate with 2.5 mL KR5-CF medium in each well with shaking at 120 rpm. These aggregates were treated with KR5-CF medium for 2 days (mouse) or 3 days (human), then the medium was changed to 2.5 mL KR5 medium with medium refreshed every other day till harvested. Obvious PKD cysts typically emerged 4 days (mouse) or 8 days (human) after shaking culture. Cystic organoid formation efficiency was quantified based on 3–4 independent experiments and cyst diameters were measured by the measure tools of Olympus Cellsens Standard software.

Small molecule screening with scalable mini PKD organoid models

Commercially available small molecule library (Cayman, # 11076) targeting major epigenetic processes was used for small molecule screening in the scalable mini PKD organoid models. Following protocols described above, thousands of mini NPC aggregates were generated using a full plate of AggreWell^™800 24-well Plate starting from 12 million Pkd2^−/− clonal NPC line #4, or #5, as two biological replicates. These mini NPC aggregates were seeded into 12-well plates with around 30 mini NPC aggregates per well (Day 0). These plates were then subjected to shaking culture with 1 mL KR5-CF medium per well for 2 days. Medium was changed to 1 mL KR5 medium per well after 2 days (Day 2). On Day 3, small molecules from the epigenetic library were added individually into each well at the concentration of 1 μM, with controls that have DMSO, or different concentrations of metformin or tolvaptan. Cystic organoid percentages and cyst diameters were quantified on Day 5 as described above. Similar protocols were used when individual hits were further validated rather than from the initial screen. For small molecule testing in the human mini PKD organoid models, the KR5-CF step was extended to 3 days (Day 0 to Day 3), followed by KR5 step with medium refreshed every two days. Small molecule candidates were added on Day 6, and data were analyzed on Day 8. Cystic organoid formation efficiency was quantified based on 3–4 independent experiments and cyst diameters were measured by the measure tools of Olympus Cellsens Standard software.

Seahorse assays on GFP^−/− or Pkd2^−/− NPCs and GFP^−/− or Pkd2^−/− nephron organoids

The XFp Extracellular Flux Analyzer (Agilent) was used for extracellular flux measurements. For GFP^−/− mNPCs (GFP^−/− mNPCs #1 and GFP^−/− mNPCs #2) vs Pkd2^−/− mNPCs (Pkd2^−/− mNPCs #4 and Pkd2^−/− mNPCs #5), 20,000 cells per well were seeded in iMatrix-511 (Nacalai USA, # 892021) coated Seahorse miniplates 18–20 hours before the assay. For GFP^−/− nephron organoids (Organoids were derived from GFP^−/− mNPCs #1 or GFP^−/− mNPCs #2, respectively) vs Pkd2^−/− nephron organoids (Organoids were derived from Pkd2^−/− mNPCs #4 or Pkd2^−/− mNPCs #5, respectively), on Day 4 of shaking culture following the mini PKD organoid protocol, one single nephron organoid was seeded into each well of poly-l-lysine coated Seahorse miniplate 1 hour before the assay (3 organoids/group). The XF sensor cartridges were hydrated overnight at 37°C without carbon dioxide in the Seahorse XF calibrant solution, as recommended by the manufacturer protocol. For the cell energy phenotype test, cell culture medium was replaced with XF assay medium (unbuffered DMEM [pH7.4] with 17.5 mM glucose, 0.5 mM pyruvate, and 4.5 mM glutamine). Microplate with cells was placed in a 37°C incubator without carbon dioxide for one hour. Oligomycin and FCCP were added to the ports at the final concentration of 1 μM. Standard XFp cell energy phenotype test (3 cycles of baseline measurements and 5 cycles of Oligomycin + FCCP with 3 minutes mixing and 3 minutes measuring) were performed for 1hour. Measurements were recorded from three to six independent experiments with three technical replicates per group in each independent experiment. Data were either normalized to cell numbers for NPCs (stained and counted with Hoechst 33342), or normalized to total genomic DNA amounts for organoids (CyQUANT^® Cell Proliferation Assay, Invitrogen, C7026). Data were analyzed using Wave software. In sum, data were collected from Seahorse experiments from mNPC and organoids derived from two different Pkd2^−/− NPC lines (Pkd2^−/− mNPCs #4 and Pkd2^−/− mNPCs #5) and two different GFP^−/− NPC lines (GFP^−/− mNPCs #1 and GFP^−/− mNPCs #2), with 3 individual organoids from each NPC line in one experiment, and the experiment was performed twice as two biological replicates.

Isolation of SIX2-GFP⁻ cells or SIX2-GFP⁻/PODXL⁺ cells from SIX2-GFP/PAX2-RFP iNPC-derived nephron organoids

SIX2-GFP/PAX2-RFP iNPC-derived nephron organoids were generated as detailed above. Nephron organoids at Day 3 (D3), Day 5 (D5), Day 7 (D7), or Day 8 (D8) time points were collected into 1.5 mL Eppendorf tubes containing 10% FBS medium. After removing 10% FBS medium as much as possible, 500 μl pre-warmed fresh 1X Collagenase IV (Thermo Fisher Scientific, # 17104019) was added to the tube. The tube was then transferred onto a rotating shaker (Eppendorf ThermoMixer^® F2.0) set at 800 rpm shaking speed in 37°C cell culture incubator for a total dissociation time of 30 minutes. Every 10 min, the tube was taken out and the reaction mix was pipetted up and down for 10 times. After 30min and the 3^rd pipetting, cells were spun down at 300g for 3min to collect the organoid pieces at the bottom of the tube. The pellet was then resuspended with 500 μl pre-warmed Accumax, incubated at 37°C with 800 rpm shaking speed for 10min. Next, 500 μl 10% FBS was added to the tube to neutralize, and the reaction mix was pipetted up and down for 10 times followed by spinning down the tube at 300g for 5min. After the centrifugation, supernatant was removed and the cell pellet was resuspended with FACS medium (cold PBS with 2% FBS) for FACS sorting. For the isolation of SIX2-GFP⁻/PODXL⁺ cells, after spinning down the cells and removing supernatant, the pellet was resuspended with FACS medium containing PODXL antibody (R&D Systems, MAB1658) at the ratio of 1:200 for live cell staining. Cells were incubated with PODXL antibody for 1 hour on ice, followed by washing with 1 ml FACS medium once and incubating in FACS medium containing fluorescence protein-conjugated secondary antibody (Donkey anti-Mouse IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor 647, Thermo Scientific, A-31571) for 1 hour on ice in dark. Then, cells were washed again and resuspended with FACS medium for FACS to sort out SIX2-GFP⁻/PODXL⁺ cells using a BD SORP FACSYMPHONY S6 cell sorter. Cultured SIX2-GFP/PAX2-RFP iNPCs were used as control for gating SIX2-GFP⁻ cells or SIX2-GFP⁻/PODXL⁺ cells every time.

Isolation of MAFB-GFP⁺ cells from MAFB-P2A-eGFP iNPC-derived nephron organoids

MAFB-P2A-eGFP (MAFB-GFP) iNPCs were derived from the parental hPSC line following the protocol detailed above. Nephron organoids were generated from MAFB-GFP iNPCs following the protocol described above. D8 nephron organoids were harvested for MAFB-GFP⁺ cell isolation following the dissociation and FACS protocol as detailed above.

Isolation of PODXL⁺ primary podocytes from human fetal kidney

Kidney nephrogenic zone pieces were dissected out from 11.4-week or 17.4-week human fetal kidneys using tweezers. These pieces were minced into smaller pieces and transferred into 1.5 mL Eppendorf tubes. After spun down the tubes at 300 g for 3 minutes, supernatant was carefully aspirated and the pieces were washed once with sterile PBS. Then, 500 μL of pre-warmed Accumax was added to each tube to resuspend the pieces and the tubes was incubated in 37°C incubator for 20~22 minutes. After the incubation, 500 μL 10% FBS medium (10% FBS in DMEM) was added to each tube, and the reaction mix was pipetted up and down very gently 20 to 25 times to further dissociate the pieces into single cells. The tubes were then spun down and the supernatant was removed, and the cells pellet was resuspended with FACS medium for PODXL staining followed by FACS for PODXL⁺ cells as described above.

Chemical reprogramming from non-NPCs and podocytes to rNPCs with hNPSR-v2

Bulk SIX2-GFP⁻ non-NPCs, and SIX2-GFP⁻; PODXL⁺ podocytes, from SIX2-GFP; PAX2-mCherry dual-reporter iNPC-derived nephron organoids, MAFB-GFP⁺ podocytes from MAFB-GFP reporter iNPC-derived nephron organoids, and PODXL⁺ primary podocytes from human fetal kidneys, were isolated as described above and cultured in hNPSR-v2 medium. The cells were continuously passaged following iNPC culture protocol described above upon grown into confluency. After culture for 7–30 days, rNPCs were purified by FACS based on SIX2-GFP expression (from SIX2-GFP; PAX2-mCherry dual-reporter background), or enriched based on ITGA8 expression (from MAFB-GFP reporter background or from primary podocytes) as described in the manuscript (for ITGA8 staining, follow protocol described in section “Deriving hNPC lines from human fetal kidneys”). The purified/enriched rNPCs were then continuously cultured following iNPC culture protocol. For human fetal kidney-derived primary podocyte culture, after 22 days of culture in hNPSR-v2, more than half of the cultured primary podocytes from 17.4wk human fetal kidney were positive for SIX2 and SALL1 (Fig. S9D and E). After 30 days of culture, ITGA8⁺ cells (68.5%) were isolated by FACS and cultured (Fig. 6P); cultured cells displayed a typical NPC morphology (Fig. S9F), consistent expression of NPC marker genes (Fig. 6Q and S9G) and nephrogenic potential on differentiation (Fig. 6R and S9H), consistent with reprogramming. An rNPC line was also derived from primary podocytes isolated from a 11.4wk hFK sample (Fig. S9A–C), confirming plasticity of primary human podocytes. For the described time-course bulk RNA-seq, FACS based on SIX2-GFP expression was performed 7 days after SIX2-GFP⁻; PODXL⁺ podocytes were cultured in hNPSR-v2 medium to purify the rNPCs.

Toxicity analysis of PTC-209 on mini PKD2^−/− organoids

PKD2^−/− iNPCs were expanded with hNPSR-v2 following the protocol as described above. iNPCs were dissociated into single cells using Accumax when reached 80–90% confluency in culture. 500,000 cells were seeded into one well of an AggreWell^™800 24-well Plate (STEMCELL Technologies, # 34850) with 1.5 mL hNPSR-v2 medium and cultured overnight to form mini aggregates (~300 mini aggregates per well). On the next day (Day 0), after mini aggregates formed, old medium was gently removed as much as possible, and 1 mL STEMdiff^™ APEL^™2 medium containing 6 μM CHIR99021 (Stage I) were added to the well for a 1-hour CHIR99021 treatment. After the 1-hour treatment, old medium was removed and mini aggregates were transferred into 1 well of a 6-well plate with 2.5mL STEMdiff^™ APEL^™2 medium containing 50 ng/mL FGF9 and 1 μg/mL heparin (Stage II) in the well. The plate was placed on a shaker set at 120 rpm for continuous shaking culture until sample harvesting. On day 5, the medium was changed to 2.5 mL Advanced RPMI 1640 Medium (Thermo Fisher Scientific, #12633–012) with 1X B27 and 200nM A83–01 (Stage III). On day 7, these PKD2^−/− organoids were transferred and evenly divided into 6 groups/wells in a 12-well plate with 1 mL Stage III medium containing DMSO, Tolvaptan_10 μM, PTC-209_10 nM, PTC-209_100 nM, PTC-209_1 μM, or staurosporine_0.1μM in each well (~50X organoids/group), under shaking culture. After 2 days of culture (On Day 9), pictures were taken to record the cyst formation efficiency and cyst growth for each group. Organoids were then harvested for TUNEL assay as described below.

Immunofluorescence staining

Whole-mount staining:

Samples were fixed in 4% PFA (4% Paraformaldehyde Aqueous Solution, Electron Microscopy Sciences, #157–4) for 15 minutes (kidney organoids, cystic organoids, or engineered kidneys) in Eppendorf tubes or tissue culture plates at room temperature. They were then washed four times in 1X PBS (Corning, Cat. No. 21–040-CV) for total 30 minutes. After the washes, samples were blocked in blocking solution (0.3% PBST containing 3% BSA) for 30 minutes at room temperature or 4°C overnight followed by primary antibody staining (primary antibodies were diluted in blocking solution) at 4°C overnight. On the second day, samples were washed four times with 0.3% PBST for total 60 minutes at room temperature. Secondary antibodies diluted in blocking solution were added and samples were incubated at 4°C overnight. On the third day, samples were washed four times with 0.3% PBST for total 60 minutes at room temperature then mounted for imaging. Confocal Microscope Zeiss LSM 800: AxioObserver.M2 was used for imaging recording in this study.

Cryo-section staining:

Samples were fixed in 4% PFA for 30 minutes (human fetal kidneys, mouse embryonic kidneys, or mouse postnatal kidneys) in Eppendorf tubes or tissue culture plates at room temperature and then washed four times in 1X PBS for total 30 minutes. Fixed kidneys were transferred and incubated in 30% sucrose overnight at 4°C. After these kidneys sunk to the bottom in the sucrose on the next day, they were then transferred into a plastic mold and embedded in OCT Compound (Scigen, Cat. No. 4586K1) and froze in −80°C for 24 hours to make a cryo-block. The cryo-blocks were sectioned using Leica CM1800 Cryostat. For staining, these sectioned slides were blocked with blocking solution for 30 minutes followed by 2 hours of primary antibodies staining, all at room temperature. After primary staining, wash the slides four times with 0.3% PBST for a total of 15 minutes, then secondary staining for one hour at room temperature. After secondary staining, the slides were washed four times with 0.3% PBST for a total of 15 minutes and mounted with mounting medium (Southern Biotech, Fluoromount-G^® Mounting Medium, #0100–01). For the Cryo-section staining of human fetal kidneys and mouse embryonic kidneys with p-p38, p-Smad2 or p-Smad1/5/9 antibodies, we used TSA-based immunocytofluorescence staining with TSA Plus Cyanine 3 Evaluation Kit (PerkinElmer, # NEL744E001KT) to enhance the phosphorylation signals. Slides were first incubated with 3% H₂O₂ reagent at room temperature for 10 minutes, rinsed with PBS, then blocked for 30 mins with blocking buffer. After blocking, these slides were incubated with diluted primary antibody overnight at 4°C (p-Smad2 dilution ratio-1:4000, p-Smad1/5/9 dilution ratio-1:4000, p-p38 dilution ratio-1:8000, in blocking buffer). On the second day, slides were rinsed with 0.3% PBST, incubated with anti-rabbit or anti-mouse HRP (1:2,000 in blocking buffer) for 2 hours at room temperature, rinsed again with 0.3% PBST, then incubated with TSA working solution (1:100 dilution) for 15 minutes at room temperature followed by a final rinse with 0.3% PBST. After the completion of TSA-based staining, proceed to standard IF staining.

Immunofluorescence staining in 96-well plates:

NPCs cultured in 96-well plates were stained directly in the plates to determine various NPC marker gene expression. For that, the used culture medium was first removed from each well and 50 μl 4% PFA was added to fix the samples in the plates for 10 minutes at room temperature. Fixed samples were then gently washed three times in 1X PBS (Corning, Cat. No. 21–040-CV) 3 times for total 15 minutes, blocked in 100 μl blocking solution (0.1% PBST containing 3% BSA) for 30 minutes at room temperature, then followed by primary antibody staining at room temperature for 2 hours. Then, samples were gently washed two times with PBST for 10 minutes and secondary staining was conducted for one hour at room temperature. After the secondary staining, samples were gently washed three times with PBST for 15 minutes, and the PBST from the last wash was kept in the well to prevent samples from drying out. Samples were then ready for observation and recording. Note: NPCs cultured on Matrigel-coated 96-well plates can easily detach from the plate during staining, it is important to add/remove reagents gently in the process of staining. Validated primary antibodies in this study can be found in “Key Resources Table”.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit Monoclonal anti-SIX1	Cell Signaling Technology	Cat# 12891, RRID: AB_2753209
Rabbit Polyclonal anti-SIX2	Proteintech	Cat# 11562-1-AP, RRID: AB_2189084
Mouse Monoclonal anti-SIX2	Abnova	Cat# H00010736-M01, RRID: AB_436993
Mouse Monoclonal anti-SALL1	R&D Systems	Cat# PP-K9814-00, RRID: AB_2183228
Rabbit Monoclonal anti-WT1	Abcam	Cat# ab89901, RRID: AB_2043201
Rabbit Polyclonal anti-PAX2	BioLegend	Cat# 901001, RRID: AB_2565001
Rabbit Polyclonal anti-PAX8	Proteintech	Cat# 10336-1-AP, RRID: AB_2236705
Mouse Monoclonal anti-MEIS1/2/3	Active Motif	Cat# 39795, RRID: AB_2750570
Rat Monoclonal anti-PODXL (mouse)	R&D Systems	Cat# MAB1556, RRID: AB_2166010
Mouse Monoclonal anti-PODXL (human)	R&D Systems	Cat# MAB1658, RRID: AB_2165984
Biotinylated Lotus Tetragonolobus Lectin (LTL)	Vector laboratories	Cat# B-1325, RRID: AB_2336558
Mouse Monoclonal anti-LRP2	MyBioSource	Cat# MBS690201
Biotinylated Dolichos Biflorus Agglutinin (DBA)	Vector laboratories	Cat# B-1035, RRID: AB_2314288
Rabbit Monoclonal anti-CDH1	Cell Signaling Technology	Cat# 3195, RRID: AB_2291471
Mouse Monoclonal anti-CDH1	BD Biosciences	Cat# 610182, RRID: AB_397581
Mouse Monoclonal anti-AQP1 (mouse)	Santa Cruz Biotechnology	Cat# sc-25287, RRID: AB_626694
Rabbit Monoclonal anti-AQP1 (human)	Abcam	Cat# ab168387, RRID: AB_2810992
Mouse Monoclonal anti-HNF4	R&D Systems	Cat# MAB4605
Goat Polyclonal anti-SLC12A1	Abcam	Cat# ab240542, RRID: AB_2910116
Rabbit Polyclonal anti-SLC12A3	Sigma-Aldrich	Cat# HPA028748-100UL, RRID: AB_10603886
Mouse Monoclonal anti-SLC27A2	Novus Biologicals	Cat# NBP2-37738
Rabbit polyclonal anti-SLC34A1 antibody	Novus Biologicals	Cat# NBP2-13328
Goat Polyclonal anti-ITGA8	R&D Systems	Cat# AF4076, RRID: AB_2296280
Rabbit Monoclonal anti-p-p38	Cell Signaling Technology	Cat# 4511, RRID: AB_2139682
Rabbit Monoclonal anti-p-SMAD2	Cell Signaling Technology	Cat# 18338, RRID: AB_2798798
Rabbit Monoclonal anti-p-Smad1/5/9	Cell Signaling Technology	Cat# 13820, RRID: AB_2493181
Rat Monoclonal anti-KRT8	DSHB	Cat# TROMA-I, RRID: AB_531826
Chicken Polyclonal anti-MAP2	Abcam	Cat# ab5392, RRID: AB_2138153
Rabbit polyclonal anti-NeuN	GeneTex	Cat# GTX16208
Mouse Monoclonal anti-Phospho-Histone H3 (Ser10) pHH3	Cell Signaling Technology	Cat# 9706, RRID: AB_331748
Mouse Monoclonal anti-YAP	Santa Cruz Biotechnology	Cat# sc-101199, RRID: AB_1131430
Rabbit polyclonal anti-PC2	Baltimore PKD Core	Cat# Rabbit mAB 3374 CT-14/4
Mouse Monoclonal anti-β-Actin	Cell Signaling Technology	Cat# 3700, RRID: AB_2242334
Goat polyclonal anti-JAG1	R&D Systems	Cat# AF599, RRID: AB_2128257
Rabbit polyclonal anti-POU3F3	Thermo Fisher Scientific	Cat# PA5-64311, RRID: AB_2645790
Sheep polyclonal anti-Nephrin	R&D Systems	Cat# AF4269, RRID: AB_2154851
Rabbit polyclonal anti-NPHS2	Abcam	Cat# ab50339, RRID: AB_882097
Mouse Monoclonal anti-acetyl-alpha tubulin	Sigma-Aldrich	Cat# MABT868, RRID: AB_2819178
Mouse Monoclonal anti-MAFB	R&D Systems	Cat# MAB3810, RRID: AB_2137675
DAPI	Thermo Fisher Scientific	Cat# D1306, RRID: AB_2629482
Biological samples
Human fetal kidney tissues (9–18 weeks)	USC and CHLA	N/A
Chemicals, peptides, and recombinant proteins
Activin A	Stemgent	Cat# 03-0001
SB431542	Reagents Direct	Cat# 21-A94
A83-01	Stemgent	Cat# 04-0014
BMP4	Stemgent	Cat# 03-0007
BMP7	R&D Systems	Cat# 354-BP-010
LDN-193189	Reagents Direct	Cat# 36-F52
CHIR99021	Reagents Direct	Cat# 27-H76
IWR-1	Sigma-Aldrich	Cat# 10161-5MG
Y-27632	Selleck Chemicals	Cat# S1049
PD0325901	Reagents Direct	Cat# 39-C68
FGF1	Peprotech	Cat# AF-100-17A
FGF2	Peprotech	Cat# AF-450-33
FGF7	Peprotech	Cat# 450-60
FGF8	Peprotech	Cat# AF-100-25
FGF9	Peprotech	Cat# 100-23
FGF10	Peprotech	Cat# AF-100-26
FGF20	Peprotech	Cat# 100-41
TNF-alpha	R&D Systems	Cat# 210-TA-020
VEGF	R&D Systems	Cat# 293-VE-010
HGF	Peprotech	Cat# 315-23
EGF	R&D Systems	Cat# 236-EG-200
JAK Inhibitor I	STEMCELL Technologies	Cat# 74022
Sphingosine-1-phosphate (S1P)	Sigma-Aldrich	Cat# S9666-1MG
Lysophosphatidic acid (LPA)	Sigma-Aldrich	Cat# L7260-1MG
SCF	R&D Systems	Cat# 255-SC-010
IGF-1	Sigma-Aldrich	Cat# I1271-.1MG
IGF-2	Peprotech	Cat# AF-100-12
Mouse LIF	Millipore	Cat# ESG1107
Human LIF	Millipore	Cat# LIF1050
All trans-Retinoic Acid	Santa Cruz Biotechnology	Cat# sc-200898
TTNPB	TOCRIS	Cat# 0761
SP600125	TOCRIS	Cat# 1496
SB202190	Axon Medchem	Cat# 1364
DAPT	Sigma-Aldrich	Cat# D5942-5MG
XMU-MP-1	Selleck Chemicals	Cat# S8334
Verteporfin	Selleck Chemicals	Cat# S1786
Forskolin	Sigma-Aldrich	Cat# F3917-10MG
GDNF	PeproTech	Cat# 450-10-50ug
PDGF-BB	R&D Systems	Cat# 220-BB-010
R-Spondin 1	R&D Systems	Cat# 4645-RS-100
AICAR	Sigma-Aldrich	Cat# A9978
TRULI	Cayman Chemical	Cat# 36623
MOZ-IN-3	Cayman Chemical	Cat# 27402-1mg
WM-1119	Cayman Chemical	Cat# 30509-5mg
WDR5 degrader	Yali Dou lab, USC	10mM-Stock
MLL1 inhibitor	Yali Dou lab, USC	20mM-Stock
Epigenetics Screening Library (96-Well) (148 Compounds)	Cayman Chemical	Cat# 11076
Tolvaptan	Selleck Chemicals	Cat# S2593
Metformin	TOCRIS	Cat# 2864
CFTRinh172	Cayman Chemical	Cat# 15545-5mg
AZ-505	Cayman Chemical	Cat# 16875-1mg
Tubacin	Cayman Chemical	Cat# 13691-1mg
PTC-209	Cayman Chemical	Cat# 16277-5mg
Staurosporine (STS)	Selleck Chemicals	Cat# S1421
(Z)-4-Hydroxytamoxifen	Sigma-Aldrich	Cat# H7904-5MG
Ethanol (control vehicle for 4-OH-tamoxifen treatment)	VWR	Cat# BDH1156-4LP
Critical commercial assays
TRIzol Reagent	Thermo	Cat# 15596026
Direct-zol RNA MicroPrep Kit	Zymo Research	Cat# R2062
DNA/RNA Shield	Zymo Research	Cat# R1100-50
Quick-RNA Microprep Kit	Zymo Research	Cat# R1051
iScript Reverse Transcription Supermix	Bio-Rad	Cat# 1708841
Ssoadvanced™ Universal SYBR	Bio-Rad	Cat# 1725274
AzuraView GreenFast qPCR Blue Mix LR	Azura Genomics	Cat# AZ-2320
TSA Plus Cyanine 3 Evaluation Kit	PerkinElmer	Cat# NEL744E001KT
KAPA Stranded mRNA-Seq Kit	KAPA Biosystems	Cat# KK8420
CyQUANT® Cell Proliferation Assay Kit	Invitrogen	Cat# C7026
AggreWell™800 6-well Starter Kit	STEMCELL Technologies	Cat# 34860
AggreWell™800 24-well Plate Starter Kit	STEMCELL Technologies	Cat# 34850
Click-iT Plus TUNEL Assay	Invitrogen	Cat# C10618
10x Chromium Next GEM Single Cell Reagent Kit A	10X Genomics	Cat# PN-1000282
10X Chromium Next GEM Chip J	10X Genomics	Cat# PN-2000264
SPRI-select Beads	Beckman Coulter	Cat# B23318
Deposited data
Bulk RNA-seq of mouse and human NPC	NCBI’s GEO	GSE230707
Single Cell Multiome of iNPC-derived nephron organoids	NCBI’s GEO	GSE251862
Bulk RNA-seq of primary mouse NPC87	NCBI’s GEO	GSE78772
Bulk RNA-seq of primary human NPC24	NCBI’s GEO	GSE102230
Bulk RNA-seq of primary human NPC88	NCBI’s GEO	GSE73867
Single Cell Multiome of hPSC-derived kidney organoids58	NCBI’s GEO	GSE213152
Single Cell RNA-seq of kidney organoids55	NCBI’s GEO	GSE124472
Experimental models: Cell lines
SIX2-GFP H1 hESC	This study	N/A
SIX2-GFP/PAX2-mCherry H1 hESC	This study	N/A
SIX2-GFP PKD2−/− H1 hESC	This study	N/A
MAFB-P2A-eGFP H9 hESC	Tran et al., 201955	N/A
Experimental models: Organisms/strains
Swiss Webster mice	Taconic Biosciences	Model # SW-F
Six2tm3(EGFP/cre/ERT2)Amc mice (Six2GCE)	The Jackson Laboratory	JAX # 009600 RRID: IMSR_JAX:009600
Wnt4tm2(EGFP/cre/ERT2)Amc mice (Wnt4GCE)	The Jackson Laboratory	JAX # 032489 RRID: IMSR_JAX:032489
Gt(ROSA)26Sortm1.1(CAG-cas9*,-EGFP)Fezh mice	The Jackson Laboratory	JAX # 026179 RRID: IMSR_JAX:026179
Tg(Hoxb7-Venus*)17Cos mice	The Jackson Laboratory	JAX # 016252 RRID: IMSR_JAX:016252
Gt(ROSA)26Sortm14(CAG-tdTomato)Hze mice	The Jackson Laboratory	JAX # 007908 RRID: IMSR_JAX:007908
Oligonucleotides
Please see Methods S1 for details
Recombinant DNA
lentiCRISPR v2	Sanjana et al.89	Addgene: #52961 RRID: Addgene_52961
lentiGuide-Puro	Sanjana et al.89	Addgene # 52963 RRID: Addgene_52963
pLKO.1-TRC	Moffat et al.90	Addgene # 10878 RRID: Addgene_10878
The Brie genome-wide CRISPR knockout library	Doench et al.91	Addgene # 73632 RRID: Addgene_73632
EF1a_mCherry_P2A_Hygro	Parekh et al.92	Addgene, # 135003 RRID: Addgene_135003
pAAVS1-P-CAG-mCherry	Oceguera-Yanez et al.93	Addgene, # 80492 RRID: Addgene_80492
pXAT2	Oceguera-Yanez et al.93	Addgene, # 80494 RRID: Addgene_80494
pCAS9_GFP	Ding et al.94	Addgene # 44719 RRID: Addgene_44719
Nanog-2A-mCherry	Faddah et al.95	Addgene # 59995 RRID: Addgene_59995
pSpCas9(BB)-2A-Puro (PX459) V2.0	Ran et al.96	Addgene # 62988 RRID: Addgene_62988
pLKO-Puro-gRNA-EGFPx3	This paper	N/A
pLKO-Puro-gRNA-Pkd1×3	This paper	N/A
pLKO-Puro-gRNA-Pkd2×3	This paper	N/A
Software and algorithms
ImageJ (version: 2.1.0)	Schneider et al.97	https://imagej.nih.gov/ij/; RRID: SCR_003070
GraphPad Prism software (version: 9.1.0)	GraphPad Prism	GraphPad Prism (https://graphpad.com); RRID: SCR_015807
FlowJo software (version: 10.0.0)	FlowJo	FlowJo (https://www.flowjo.com/); RRID: SCR_008520
Partek Flow Genomic Analysis software	Partek Genomics Suite	Partek Genomics Suite (http://www.partek.com) RRID: SCR_011860
RStudio software	RStudio	RStudio (https://posit.co/) RRID: SCR_000432
Ingenuity Pathway Analysis software	QIAGEN	Ingenuity Pathway Analysis (http://www.ingenuity.com/products/pathways_analysis.html) RRID: SCR_008653
SnapGene software	SnapGene	SnapGene (http://www.snapgene.com) RRID: SCR_015052
ZEN Digital Imaging for Light Microscopy	ZEISS	ZEN Digital Imaging for Light Microscopy (http://www.zeiss.com/microscopy/en_us/products/microscope-software/zen.html#introduction) RRID: SCR_013672
Olympus Cellsens Standard software	Olympus	Olympus cellSens Software (http://www.olympuslifescience.com/en/software/cellsens) RRID: SCR_014551
Wave software	Agilent Technologies	Seahorse Wave (http://www.agilent.com/en-us/products/cellanalysis-(seahorse)/softwaredownload-for-wavedesktop) RRID: SCR_014526

Imaging data quantification:

For immunostaining quantification, 3–4 different fields of view were randomly selected to count the number of positively stained cell numbers and total cell numbers (as determined by DAPI signals). At least 500 cells per field of view were included. Error bars represent standard derivation between different fields of views.

TUNEL assay

Mini PKD2^−/− organoids were stained using the Click-iT Plus TUNEL Assay kit (Invitrogen, Cat# C10618) following the manufacturer’s instruction. Briefly, Mini PKD2^−/− organoids were incubated with the TdT reaction buffer for 10 min at 37°C in 1.5mL Eppendorf tube. TdT reaction buffer was then removed and replaced by TdT reaction mixture and incubated for 1 hour at 37°C. After the 1-hour incubation, TdT reaction mixture was removed and the organoids were rinsed with PBST (0.3% Triton X-100 in PBS). Next, organoids were incubated with the Click-iT Plus TUNEL reaction cocktail at 37°C for 30 min in dark, followed by PBST washes and immunofluorescent staining as detailed above.

RNA isolation, reverse transcription, qRT-PCR, and immunoblotting

Samples were dissolved in 100 μl TRIzol (Invitrogen, Cat. No. 15596018) or 100 μl DNA/RNA Shield (Zymo Research, Cat. No. R1100–50) and kept in −80°C freezer. RNA isolation was performed using the Direct-zol RNA MicroPrep Kit (Zymo Research, Cat. No. R2062) or Quick-RNA Microprep Kit (Zymo Research, Cat. No. R1051) according to the manufacturer’s instructions. Reverse transcription was performed using the iScript Reverse Transcription Supermix (Bio-Rad, Cat. No. 1708841) following the manufacturer’s instructions. qRT-PCR was performed using SsoAdvanced Universal SYBR^® Green Supermix (Bio-Rad, Cat. No. 1725274) or AzuraView GreenFast qPCR Blue Mix LR (Azura Genomics, Cat. No. AZ-2320) and carried out on an Applied Biosystems Vii 7 RT-PCR system (Thermo Scientific P/N 4453552). Validated gene-specific primers in this study can be found in Methods S1. Fold changes were calculated from ΔCt using Gapdh as a housekeeping gene as previously described.¹⁰⁴ Immunoblotting experiments were performed as described previously.¹⁰⁴

FACS

Cells were dissociated/prepared as described above. FACS sorting was performed on a BD FACSAria^™ III Cell Sorter, operated by experienced core facility staff at USC Stem Cell’s FACS core. Sorted cells were collected in 1.5 ml Eppendorf tubes with 500 μl dissection medium on ice.

QUANTIFICATION AND STATISTICAL ANALYSIS

Data are presented as mean ± SD from at least three biological replicates (n=3). Statistical significance was determined by two-tailed unpaired Student’s t tests; ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001.

Supplementary Material

1

Methods S1: Summary of growth factors and small molecules screened in this study, mNPSR-v2 and hNPSR-v2 medium recipe, mNPC lines established, and oligos, related to STAR Methods.

3

Table S1: Bulk RNA-seq data, related to Figures 1, 4, 5, and 6.

4

Table S2: CRISPR screen data, related to Figure 3.

5

Table S3: Single cell multiome data, related to Figure 5.

6

Video S1: Live imaging showing chick vasculature infiltrating into mouse NPC transplant, related to Figure 1.

Highlights:

• YAP activation allows long term expansion of hPSC-induced nephron progenitor cells (iNPCs)

• iNPC-derived nephron organoids generate mature podocytes with limited off-target cells

• Podocyte-to-NPC reprogramming with hNPSR-v2 medium reveals human podocyte plasticity

• Genome-wide CRISPR screening and PKD modeling from genome-edited NPC lines