Modelling kidney disease with CRISPR-mutant kidney organoids derived from human pluripotent epiblast spheroids

Abstract

Human-pluripotent-stem-cell-derived kidney cells (hPSC-KCs) have important potential for disease modelling and regeneration. Whether the hPSC-KCs can reconstitute tissue-specific phenotypes is currently unknown. Here we show that hPSC-KCs self-organize into kidney organoids that functionally recapitulate tissue-specific epithelial physiology, including disease phenotypes after genome editing. In three-dimensional cultures, epiblast-stage hPSCs form spheroids surrounding hollow, amniotic-like cavities. GSK3β inhibition differentiates spheroids into segmented, nephron-like kidney organoids containing cell populations with characteristics of proximal tubules, podocytes and endothelium. Tubules accumulate dextran and methotrexate transport cargoes, and express kidney injury molecule-1 after nephrotoxic chemical injury. CRISPR/Cas9 knockout of podocalyxin causes junctional organization defects in podocyte-like cells. Knockout of the polycystic kidney disease genes PKD1 or PKD2 induces cyst formation from kidney tubules. All of these functional phenotypes are distinct from effects in epiblast spheroids, indicating that they are tissue specific. Our findings establish a reproducible, versatile three-dimensional framework for human epithelial disease modelling and regenerative medicine applications.

Figure 1. hPSCs form cavitated spheroids in 3D culture.

(a) Schematic of spheroid-to-organoid culture protocol. (b) Phase contrast images of ESCs in sandwich (3D) or monolayer (2D) cultures. Consecutive days are shown, with d0 indicating the time point immediately before sandwiching. (c) Confocal optical sections showing PODXL, ZO-1 and βCAT immunofluorescence through a representative spheroid with cavity. Vertical distance from top to bottom row is shown at left. (d) Representative brightfield images of hPSCs in 3D cultures that were dissociated (coloured frames) and passaged (matching coloured arrows). Dashed arrows represent serial passages in the 3D condition. Lower row shows cells plated into 2D cultures from dissociated spheroids from each passage. (e) Hematoxylin and eosin-stained sections of teratomas generated from hPSC serial 3D passages p3, p6 and p9 showing pigmented epithelium (ectoderm), cartilage (mesoderm) and glandular epithelium (endoderm). (f) Cell number (average of duplicate counts for each time point, or AVG of all five time points shown in the last column) in 2D and 3D cultures 72 h after plating. (g) Representative immunofluorescence images showing OCT4 and sex-determining region Y box-2 (SOX2) or tumor rejection antigen 1–60 (TRA-1-60) and NANOG localization in p3 and p7 serially sandwiched hPSCs. Scale bars, 100 μm. Error bars, s.e.m.

Figure 2. Tubular organoids recapitulate kidney development and architecture.

(a) Phase contrast images of spheroid differentiation into tubular organoids. Red arrowheads highlight epithelia. (b) Confocal optical sections showing LTL with nephron progenitor markers Sine oculis homeobox homologue 2 (SIX2), Lin11-Isl1-Mec3 (LIM) homeobox 1 (LHX1), paired box gene 2 (PAX2) and (c) proximal tubule markers low density lipoprotein-related protein 2/megalin and cubilin in tubular organoids. (d) Electron micrographs of a representative tubule, with progressive magnifications of regions in coloured boxes highlighting apical microvilli (arrowheads) and tight junctions (arrows). (e) Wide-field images showing tubule anatomical progression from E-cadherin (ECAD)⁺ to LTL⁺ to PODXL⁺ organoid segments. (f) Low-magnification image of organoid with interlacing tubules and peripheral PODXL⁺ aggregates (left) and high-magnification confocal optical section showing co-localization of synaptopodin (SYNPO) and Wilms tumour protein (WT1) in organoid podocyte-like cells (right). (g) Wide-field immunofluorescence showing co-localization of CD31 with von Willebrand factor (vWF, left), or with nephron markers in tubular organoids derived from hESCs and iPSCs (right). White arrowheads show interactions between tubular, podocyte-like and endothelial compartments. White dashed outline highlights a representative tubular terminus. Images are representative of one hESC line and three iPSC lines from different patients. (h) Number of tubular organoids formed per unit surface area in cultures of hESCs and iPSCs (left) and per cent of these LTL⁺ organoids associated with CD31⁺ and PODXL⁺ cell types within the organoid (right). (i) Confocal images of organoid-derived human tubule (H) with LTL reactivity after 3 weeks of growth inside the developing mouse kidney cortex (m). Scale bars, 100 μm. Error bars, s.e.m (n≥3 experiments).

Figure 3. Tubular organoids create a microphysiological model for kidney nephrotoxicity studies.

(a) Co-localization of KIM-1 antibody AKG7 with LTL or (b) with a second KIM-1 antibody (KIM1ab2). (c) Quantification of organoids expressing KIM-1 after treatment with 50 μM cisplatin (Cispl.) or 5 mM gentamicin (Gent.), compared with vehicle-treated controls (n=3/condition, >50 organoids/experiment). (d) Phase contrast of living cells, or KIM-1 and NANOG immunofluorescence, in epiblast spheroids treated with titrations of cisplatin and gentamicin. No KIM-1 expression is observed at any dose including 5 mM gentamicin, a concentration sufficient to induce KIM-1 upregulation in kidney organoids. Cisplatin was toxic to spheroids at concentrations>1 μM. (e) Phase contrast time course of tubules in culture. After 120 days of culture (d120), tubules were fixed and stained for LTL. (f) Whole well of a 96-well plate showing kidney organoids. Scale bars, 50 μm. Error bars, s.e.m.

Figure 4. Differential accumulation of fluorescent cargoes in epiblast spheroids versus kidney organoid tubules.

(a) Experimental schematic and confocal time-lapse images showing hPSC spheroids incubated with lucifer yellow (LY) and rhodamine-conjugated dextran (RD). (b) Experimental schematic and wide-field time course of an hPSC spheroid cavity after microinjection with RD. No autofluorescence is detected (green fluorescent protein channel). (c) An hPSC spheroid treated with fluorescein methotrexate for 4 h, immediately after washout. (d) Time course of an hPSC spheroid incubated with LY or RD molecular dyes for 4 h, after which the media was replaced without dyes (washout). (e) Representative time course of a tubular organoid incubated with RD and LY for 4 h (pulse), followed by incubation in fresh media without dyes (chase), fixation and co-localization with LTL. (f,g) Quantification of tubular organoids that accumulated RD, with or without 2 μM Latrunculin B (n=3). (h) Fluorescein methotrexate and RD distributions in two representative live organoids 1 h and 18 h after washout. Scale bars, 50 μm (a–d) or 100 μm (e–h). Error bars, s.e.m.

Figure 5. Podocalyxin promotes lumenogenesis in epiblast spheroids.

(a) Immunoblot for podocalyxin protein in two representative PODXL^−/− hPSC mutant clones (m1 and m2), compared with CRISPR/Cas9 non-mutant wild-type clones (WT) or cells subjected to scrambled (scr) or podocalyxin (pod) siRNA knockdown. (b) Brightfield images of sandwiched parental ESCs were compared with two mutant or two WT CRISPR/Cas9 clones. (c) Cavitated spheroids as a percentage of total colonies. Data from pools of WT or mutant cell lines were averaged to determine group means (AVG, n≥9) and P values. (d) Podocalyxin and ZO-1 immunofluorescence in naive and primed hLR5 hPSCs or (e) mESCs and EpiSCs. (f) Confocal z-sections of undifferentiated hPSCs showing localization of ZO-1 and βCAT in unmodified (WT) or PODXL^−/− colonies. (g) Filamentous actin (f-actin) and occludin (OCLN) immunofluorescence in undifferentiated WT or PODXL^−/− clones. (h) Averaged TEER measurements in WT or PODXL^−/− monolayers (n≥3). Scale bars, 50 μm or (f,g) 20 μm. Error bars, s.e.m.

Figure 6. Junctional complexes are disrupted in hPSC-derived PODXL^−/− podocyte-like cells.

(a) Confocal optical section of adult human kidney. Podocalyxin is expressed in podocytes and peritubular capillaries, but is absent from tubules (white dotted lines). Auto, autofluorescence. (b) Crumbs3 expression in hPSC-derived kidney organoids (confocal red channel) and human kidney tissue (far right panel, immunohistochemistry). (c) Confocal optical sections showing distributions of ZO-1 with podocalyxin or (d) βCAT in hPSC-derived podocyte-like cell clusters. Arrowheads highlight tracks of junctional complexes between podocyte-like cells. (e) Confocal sections of wild-type or PODXL^−/− podocyte-like cell clusters in tubular organoids. (f) Gap widths between adjacent podocyte-like cell nuclei in these cell lines (n≥100 gaps pooled from two experiments). Scale bars, 50 μm. Error bars, s.e.m.

Figure 7. PKD hPSCs model lineage-specific cyst formation.

(a) Immunoblots showing reduction in full-length polycystin-1 (PC1) in CRISPR-generated PKD1^−/− hPSCs or (b) polycystin-2 (PC2) in PKD2^−/− hPSCs, compared with isogenic controls. PKD2 knockdown is shown for comparison. (c) Epiblast spheroid morphology in representative PKD knockout hPSCs and controls. (d) Representative cyst on day 58 of culture in PKD2 kidney organoids. (e) Quantification of cyst formation rate in PKD knockout organoids and isogenic WT controls as an average of all experiments (n>10) or as scatter plots of individual experiments. (f) Wide-field epifluorescence images and (g) confocal optical sections showing LTL reactivity in cysts. Representative z-sections show hollow center of cyst and associated tubular organoid (arrow). Zoom is shown of red boxed regions. Scale bars, 100 μm. Error bars, s.e.m. *P<0.01.

Figure 8. Models of hPSC-derived epithelia.

(a) Model of hPSC lumenogenesis. ICM-stage hPSCs lack polarized tight junctions, resulting in randomized aggregate formation. Epiblast-stage hPSCs are polarized with continuous tight junctions, and thus organize into a single-cell epithelium. In 3D growth, polarized accumulation of podocalyxin (blue) at the apical membrane results in charge repulsion (negative charges), promoting separation of the cells to form a lumen. Tight junctions (red) permit the entry of small (green) molecules but exclude macromolecules (red), which accumulate in intercellular spaces. Zoom of boxed area is shown, highlighting polarized epithelial cells. (b) Architecture of a proximal tubule within a kidney organoid. An elongated proximal tubule forms a simple columnar epithelium which binds LTL (green) on the apical surface. Surrounding podocyte-like cells express high levels of podocalyxin and form a less organized, aggregate structure at tubular termini. ZO-1 (red) is expressed at a sub-apical position, restricted by podocalyxin (blue). Endothelial cells interact closely with both tubular and podocyte-like compartments. Faded background structures place epithelial structures formed in vitro into their proposed context in vivo.