Vasculogenesis in kidney organoids upon transplantation

ABSTRACT Human induced pluripotent stem cell-derived kidney organoids have potential for disease modeling and to be developed into clinically transplantable auxiliary tissue. However, they

lack a functional vasculature, and the sparse endogenous endothelial cells (ECs) are lost upon prolonged culture in vitro, limiting maturation and applicability. Here, we use intracoelomic

transplantation in chicken embryos followed by single-cell RNA sequencing and advanced imaging platforms to induce and study vasculogenesis in kidney organoids. We show expansion of human

organoid-derived ECs that reorganize into perfused capillaries and form a chimeric vascular network with host-derived blood vessels. Ligand-receptor analysis infers extensive potential

interactions of human ECs with perivascular cells upon transplantation, enabling vessel wall stabilization. Perfused glomeruli display maturation and morphogenesis to capillary loop stage.

Our findings demonstrate the beneficial effect of vascularization on not only epithelial cell types, but also the mesenchymal compartment, inducing the expansion of ´on target´ perivascular

stromal cells, which in turn are required for further maturation and stabilization of the neo-vasculature. The here described vasculogenic capacity of kidney organoids will have to be

deployed to achieve meaningful glomerular maturation and kidney morphogenesis in vitro. SIMILAR CONTENT BEING VIEWED BY OTHERS PRODUCTION OF KIDNEY ORGANOIDS ARRANGED AROUND SINGLE URETERIC

BUD TREES, AND CONTAINING ENDOGENOUS BLOOD VESSELS, SOLELY FROM EMBRYONIC STEM CELLS Article Open access 22 July 2022 CREATING A KIDNEY ORGANOID-VASCULATURE INTERACTION MODEL USING A NOVEL

ORGAN-ON-CHIP SYSTEM Article Open access 30 November 2022 GENERATION OF THE ORGANOTYPIC KIDNEY STRUCTURE BY INTEGRATING PLURIPOTENT STEM CELL-DERIVED RENAL STROMA Article Open access 01

February 2022 INTRODUCTION Since the publication of protocols for the generation of three-dimensional (3D) kidney organoids from human induced pluripotent stem cells (hiPSCs)1,2,3,4,

expectations regarding their potential for the modeling of kidney development and disease have been very high. Over the past few years, it has become increasingly clear that the level of

maturation in vitro is an important limiting factor in this respect. Although many aspects, including culture environment, restricted time in culture, off-target differentiation, and limited

corticomedullary organization play a role in restricting maturation, the absence of a functional vasculature is of particular importance. In embryology, the development of the kidney

depends on the presence of perfused blood vessels. These supply the tissue with oxygen and nutrients and enable the interaction between podocytes and vascular endothelial cells (ECs) that is

necessary for their maturation and development of the glomerular basement membrane (GBM)5,6,7. In kidney organoids maintained in vitro, this interaction does not occur, limiting their

maturation as well as their resemblance to in vivo kidney tissue. We and others have shown previously that transplantation in mice leads to vascularization and progressive maturation of

hiPSC-derived kidney organoids8,9,10, which even enables functional glomerular filtration11. However, this model is labor intensive and unsuitable to study large numbers of organoids.

Transplantation on the chicken chorioallantoic membrane (CAM) is much more straightforward and was demonstrated to induce vascularization12. Because the CAM takes time to develop, this

method allowed for only 5 days of transplantation, limiting the extent of the vasculature and maturation. Therefore, we aimed to develop an easy, efficient, reproducible and accessible model

to induce and study vascularization and maturation of kidney organoids through transplantation in the coelomic cavity of chicken embryos. Chicken embryos can develop normally after opening

the egg shell, lack a fully functional immune system, and their coelomic cavity is a favorable environment for vascularization and differentiation of transplanted embryonic

tissues13,14,15,16,17. Unlike grafting on the CAM, the coelom permits transplantation as early as day 3–4 of embryonic development, allowing for a longer duration of transplantation, and

quite unlimited expansion of the transplanted tissue in all directions16. Here, we transplanted hiPSC-derived kidney organoids inside the coelom of chicken embryos and extensively

investigated the effect of transplantation on the vasculature, cellular composition and maturation of the organoids using immunofluorescence (IF), single cell RNA sequencing (scRNAseq), and

ultra-large scale transmission electron microscopy (TEM)18. To enable analysis of the 3D organization of organoid glomeruli, we performed serial block face scanning electron microscopy

(SBF-SEM) followed by artificial intelligence-based segmentation, annotation and 3D reconstruction. We demonstrate the development of a chimeric perfused vascular network inside transplanted

organoids, which invades the glomerular structures and induces enhanced nephron maturation as well as improved stromal composition. RESULTS KIDNEY ORGANOIDS ARE VASCULARIZED UPON

INTRACOELOMIC TRANSPLANTATION IN CHICKEN EMBRYOS Chicken embryos are easily accessible, require very little maintenance, and their coelomic cavity is a favorable environment for the

differentiation of embryonic tissues15,16,17,19. Therefore, we aimed to investigate the suitability of intracoelomic transplantation to induce and study vascularization and maturation of

hiPSC-derived kidney organoids. Kidney organoids were generated from hiPSCs using previously published protocols (Fig. 1a)10,20. Although organoids generated in this manner contain some ECs,

these fail to invade glomerular structures and diminish upon prolonged in vitro culture (Supplementary Fig. 1A)21. To ensure the presence of endogenous ECs at the time of transplantation,

we decided to transplant organoids at a relatively early stage, on d7+11 or d7+12 of differentiation. Organoids were bisected and inserted into the coelomic cavity of day 4 (Hamburger

Hamilton stage 23) chicken embryos (Fig. 1a, b)19,22. At this stage of development, the coelom is large enough to accommodate half an organoid, and the body wall is not yet fully closed,

allowing straightforward transplantation that takes around 5 min per embryo. Transplanted organoids and untransplanted controls were harvested 1 day after transplantation for scRNAseq

analysis and 8 days after transplantation for scRNAseq, IF, TEM and SBF-SEM analysis (Fig. 1a). Embryos selected for IF analysis were injected with Rhodamine labeled _lens culinaris

agglutinin_ (LCA)23 to stain the endothelium of perfused blood vessels before sacrificing them. Organoids had most frequently become attached to the liver, but could also be found connected

to the intestines or ribs. Upon visual inspection through a stereo microscope, they appeared to be vascularized with apparent blood perfusion (Fig. 1c). For scRNAseq analysis, we performed

enzymatic dissociation of 94 untransplanted and 75 bisected transplanted organoids (150 chicken embryos) at d7+13 and 93 untransplanted and 46,5 bisected transplanted organoids (93 chicken

embryos) at d7+20 from the same differentiation to single cells. Pooling such a high number of organoids for each condition allows for mitigating potential inter-organoid variability. An

aliquot was taken from the single cell suspension of each condition and processed for sequencing. After species demultiplexing (for transplanted organoids), quality controls and filtering, a

total of 16,290 high quality human organoid cells and 6866 high quality chicken host cells were retained for downstream analyses (Fig. 1d, Supplementary Figs. 2A, 3A, and Supplementary Data

1). Unsupervised clustering of the organoid-derived human cells revealed three main cell populations, which were identified as nephron, mesenchymal and endothelial cells based on expression

of established marker genes (Fig. 1d, e and Supplementary Data 2). The proportion of nephron cell types was the highest at d7+13 (73% of the total cell number in untransplanted and 71% in

transplanted organoids) and had decreased in favor of mesenchymal cell types at d7+20 (60% in untransplanted, 45% in transplanted). As expected, in organoids cultured in vitro the proportion

of ECs decreased over time from 2.6% at d7+13 to 0.4% at d7+20. Upon transplantation, however, we observed an increase to 4% of ECs at d7+20 (Fig. 1f and Supplementary Data 3). Whole mount

IF staining of untransplanted and transplanted organoids at d7+19–20 followed by quantification of organoid derived ECs (human CD31+) confirmed the higher percentage of ECs in transplanted

organoids (Fig. 1g, h). The presence of injected rhodamine-labeled LCA demonstrated perfusion of the vasculature (Fig. 1g and Supplementary Fig. 1B). All investigated transplanted organoids

had become vascularized. In addition, the perfused blood vessels (CD31+, LCA+) were shown to invade the organoid glomerular structures marked by an mTagBFP2 podocyte reporter (MAFB+) or

stained for nephrin (NPHS1+) and align with tubular structures (LTL+/ ECAD+) (Fig. 1I and Supplementary Fig. 1B). Host-derived ECs also appear to contribute to the vasculature, as evidenced

by the presence of CD31−, LCA+ (perfused chicken derived) vessels beside CD31+, LCA+ (perfused organoid derived) and CD31+, LCA− (unperfused organoid derived) vessels (Supplementary Fig.

4A). Unperfused chicken-derived endothelial cells could not be identified, because a specific antibody for these cells is not available. Within the chicken cells collected with transplanted

organoids for scRNAseq, an angiogenic EC population was identified (Supplementary Fig. 3B–D and Supplementary Data 4). Some of these might reflect host-derived ECs that invaded the kidney

organoids upon transplantation in addition to the endogenous human ECs. However, liver-specific sinusoidal ECs, as well as hepatocytes and other non-renal cell types were also identified,

likely emerging from chicken tissue that remained attached to the kidney organoids during collection despite careful dissection, resulting in challenging interpretation of the chicken cell

dataset (Supplementary Fig. 3B–D). Since vascularization was already quite extensive after only 8 days of transplantation, we investigated the development of the vasculature in the organoids

over time. On day 1 after transplantation, perfused blood vessels (LCA+) could already be detected at the edge of the transplanted organoids (Supplementary Fig. 4B). Three days after

transplantation, the first glomerular-like structures (MAFB+) started to be invaded by human ECs (CD31+/LCA+). On day 5, vascularized glomerular structures could be seen throughout the

organoid, and on day 7, the majority of glomerular structures were perfused (Supplementary Fig. 4B). HETEROGENEITY OF NEPHRON CELL TYPES IN KIDNEY ORGANOIDS IS PRESERVED UPON TRANSPLANTATION

Next, we evaluated the effect of transplantation and vascularization on the cellular composition of kidney organoids, starting with the nephron compartment. Unsupervised subclustering of a

total of 10,198 high quality nephron cells obtained from untransplanted (3116 cells from d7+13; 3963 cells from d7+20) and transplanted (1816 cells from d7+13; 1303 cells from d7+20)

organoids resulted in 16 subclusters, which we identified based on the top marker genes for each cluster (Fig. 2a, b, Supplementary Fig. 2b–d, Supplementary Data 5, 6). Organoids from all

conditions contained proliferative cell clusters, nephron precursors (pretubular aggregate/renal vesicle and nephron progenitor cells), tubular cells and podocytes (Fig. 2a). Supporting our

cluster identification, further unsupervised hierarchical clustering of the delineated nephron cell phenotypes revealed higher similarity of the different identified podocyte clusters as

compared with nephron precursors and proliferating nephron cells, as well as a greater distance with tubular cell clusters (Fig. 2c). More specifically, 4 subclusters of tubular cells were

distinguished: 2 types of proximal tubule cells (#1 and 2), loop of Henle-like cells, and distal tubule/collecting duct-like cells (Fig. 2b, c, Supplementary Fig. 2c, and Supplementary Data

5, 7). We did not identify a clear distal convoluted tubule (_SLC12A3_) or pure collecting duct (_AQP2/4_) population. The podocytes displayed more heterogeneity, resulting in 7 subclusters,

mainly reflecting different levels of maturation. The podocyte-committed progenitor cluster highly expressed _OLFM3_, a podocyte-specific lineage marker, but no other markers for podocyte

identity. The immature podocyte, early podocyte #1 and 2 and late podocyte clusters expressed increasing levels of markers for podocyte identity (_MAFB_, _WT1_), and differentiation (_NPHS1_

and _2_, _PODXL_, _CLIC5_) (Fig. 2b, Supplementary Fig. 2b, and Supplementary Data 5). In addition, we found a subcluster of late podocytes expressing genes associated with stress and a

subcluster of hypoxic podocytes (Fig. 2b and Supplementary Fig. 2b, e). Since all organoids from the same condition were pooled, it is not possible to definitively distinguish whether these

podocyte subclusters are the result of different maturation stages within the same or between different organoids. However, based on variation observed through TEM imaging, we believe it

reflects intra- rather than inter-organoid variability. All nephron subclusters were represented in all organoid conditions, and major shifts in populations were not observed (Fig. 2d).

However, there were some interesting differences between conditions. In transplanted organoids (d7+13 and d7+20), the proportion of hypoxic podocytes appeared increased, which can be

explained by the hypoxia in the intracoelomic cavity prior to vascularization. This did not lead to a larger proportion of stress-induced podocytes, which were most abundant in

untransplanted d7 + 20 organoids. TRANSPLANTATION INDUCES AN INCREASE IN PERIVASCULAR CELL TYPES Mesenchymal cells are essential for kidney development and function. However, they can also

contribute to kidney fibrosis, and subpopulations of stromal cells have been reported in organoids to give rise to off-target cell populations such as cartilage24. To investigate the

identity of the mesenchymal cells in our kidney organoids, we performed unsupervised subclustering of the 5748 high quality mesenchymal cells obtained from untransplanted (1028 cells from

d7+13; 2607 cells from d7+20) and transplanted (653 cells from d7+13; 1460 cells from d7+20) organoids (Fig. 2e). This revealed 12 clusters, which we identified as proliferating cells,

fibroblasts, mural cells, mesenchymal progenitor cells and off-target cell populations based on their top marker gene expression (Fig. 2e, f and Supplementary Data 8). Further unsupervised

hierarchical clustering supported our cluster identification with grouping clusters by cell types: fibroblasts, mural cells, off-target cells derived from neural lineage, and proliferating

cells together with mesenchymal progenitors and derivative off-target chondrocyte-like cells (Fig. 2g). More specifically, off-target cells consisted of chondrocyte-like cells, neural

progenitors, neuron-like cells and melanocyte-like cells. Three types of mural cells, namely pericytes/mesangial cells, smooth muscle cells, and unspecified mural cells could be

distinguished (Fig. 2e, f and Supplementary Fig. 2f). In untransplanted organoids at d7+20, the proportion of off-target cell types and mesenchymal progenitor cells was increased compared to

d7+13, at the expense of mural cells. In transplanted organoids at d7+20, the increase in off-target cell types and mesenchymal progenitors was much less pronounced, but we did observe an

increase in pericytes/mesangial cells and fibroblasts (Fig. 2h and Supplementary Data 3). We confirmed the presence of perivascular stromal cells in transplanted organoids through IF

staining with a PDGFRβ antibody, combined with injected rhodamine-labeled LCA. In untransplanted organoids, PDGFRβ positive stromal cells were present mainly in the periphery of the organoid

and only sporadically in the center scattered between the epithelial structures. In transplanted organoids, perivascular PDGFRβ+ stromal cells supporting the LCA+ perfused blood vessels had

appeared, although some areas with PDGFRβ+ stromal cells were still visible between epithelial structures (Fig. 2I). These data imply that, although transplantation in chicken embryos may

increase the proportion of mesenchymal cells, it favors their differentiation to ‘on-target’ pericytes and mesangial cells, while reducing the amount of off-target cell types compared to in

vitro controls. VESSEL WALL STABILIZATION IN TRANSPLANTED ORGANOIDS Stabilization and maturation of newly formed blood vessels requires the establishment of inter-EC contacts and pericyte-EC

interactions25,26. We evaluated the occurrence of these events in our model through gene set enrichment analysis (GSEA) of ECs in d7+20 transplanted compared with respective untransplanted

organoids and ligand-receptor (LR) network analysis. Differential gene expression analysis (DGEA) followed by GSEA in ECs demonstrated upregulation of, amongst others, gene sets associated

with vascularization, regulation of vasomotor tone, immunity (likely due to the connection to the chicken embryo circulation), and extracellular matrix organization (occurs in but is not

specific for angiogenesis and vessel stabilization). Interestingly, a gene set associated with cell–cell junction organization, an important event in vessel wall stabilization, as well as a

gene set involved in the regulation of vasoconstriction, reflecting maturation of the vasculature were significantly enriched (Fig. 3a and Supplementary Data 9). Indeed, expression of many

individual genes involved in these processes were upregulated in transplanted kidney organoids at d7+20 compared to all other conditions (Fig. 3b). To investigate the interaction of ECs with

pericytes, we performed LR network analysis. Because of the low number of both pericytes/mesangial cells and ECs, especially in untransplanted d7+20 organoids, we decided to pool the cells

from this condition with those from untransplanted and transplanted d7+13 organoids, and compare them to transplanted d7+20 organoids for further LR analysis. Since the observed increase in

proportion and change in localization of pericytes was clearly present in transplanted d7+20 organoids compared to the other conditions, changes in interactions between ECs and pericytes

upon transplantation were expected to remain detectable after pooling of cells from these 3 conditions. Investigation of the differential interaction strength (reflecting the probability of

interactions) for all clusters revealed an increase of inferred interactions between ECs and mesenchymal cell types in transplanted d7+20 organoids compared to pooled controls, especially

with the ECs as receiver (Fig. 3c). We subsequently analyzed which ligand-receptor interactions were newly detected or increased from pericytes to ECs and vice versa (Fig. 3d). We observed

an increase in a large number of LR-pairs, the majority of which have been reported to play a role in angiogenesis (SEMA-PLXN27, PTN-SDC328, EFN-EPH29,30,31,32, IGF1-IGF1R, GDF11-TGFBR1 + 

ACVR2B33), vascular development (CXCL12-CXCR34, FN1-ITGA35), pericyte recruitment to neovasculature (PDGFB-PDGFRB, EDN1-EDNRA), vessel wall stabilization (ANGPT-TEK, JAG/DLL4-NOTCH36,37), or

generally in pericyte-endothelial cell interaction (TGFB2-TGFBR38, VEGF-VEGFR39, ESAM-ESAM40). For the two most relevant identified upregulated pathways, _undefined_ ANGPT-TEK and

PDGF-PDGFR, we generated heatmaps to visualize the expression of the genes encoding the ligands and receptors in all organoid cell clusters (Supplementary Fig. 5a, b). _ANGPT1_ was expressed

by pericytes and unspecified mural cells, but also by podocytes, which is in accordance with the reported expression in kidney development41. _ANGPT2_ was expressed by endothelial cells and

mural cells. As expected, all receptors for angiopoietins were most highly expressed in endothelial cells. _PDGFB_ was strongly expressed by endothelial cells and its receptor _PDGFRB_ in

mural cells and to a lesser degree in other mesenchymal cell types and several subsets of podocytes. Inferred interactions with collagens and laminins were broadly upregulated in

transplanted organoids, which can play a role in but is not specific for vessel maturation. Also, several upregulated LR-pairs have not been described previously in physiological

vascularization, namely _POSTN_-_ITGAV_ + _ITGB5_ (_POSTN_ implicated in kidney fibrosis), _PGF_-_VEGFR1_ (mainly described in pathological angiogenesis), _CD99_-_CD99L2_ and _MIF_-_ACKR3_

(involved in neutrophil extravasation and attraction), _JAM3_-_JAM3_ and _NECTIN3_-_NECTIN2_ (part of endothelial cell tight and adherens junctions) and _LRRC4B_-_PTPRF_ (mainly described in

neurons). These could be examples of unwanted effects of transplantation or previously unrecognized physiological interactions. Distinguishing between these would require further

investigation. Finally, to add spatial information to these sequencing data, we performed TEM to evaluate the presence of perivascular stromal cells in transplanted organoids, and identified

pericyte-like cells supporting extraglomerular capillaries (Fig. 3e). VASCULARIZATION IMPROVES THE MATURATION OF THE KIDNEY ORGANOIDS Cell-cell communication is essential for kidney

development and maintenance. Interestingly, a comparison of the total number of interactions and interaction strength of all cell clusters between conditions showed a marked decrease in

interactions in untransplanted d7+20 organoids compared to all other conditions, suggesting overall loss of cell-cell communication in in vitro culture (Supplementary Fig. 2G). To further

evaluate the effect of transplantation and vascularization on nephron cell types, we compared late stage transplanted organoids with respective untransplanted controls through DGEA followed

by GSEA. This analysis was performed for the late podocyte cluster and tubular epithelial cells (proximal, loop of Henle and distal clusters combined). In transplanted late podocytes, we

observed enrichment of gene sets associated with VEGF production, vasculogenesis, organization of cell–cell junctions, and establishment and maintenance of cell polarity (Supplementary Fig.

6A and Supplementary Data 10). This is consistent with the observed vascularization and indicates differentiation of podocytes from immature simple columnar epithelium to a more mature

phenotype displaying apicobasal polarity and forming intercellular junctions at their basal side. Interestingly, a gene set associated with positive regulation of collagen biosynthetic

processes was also enriched in transplanted podocytes. We hypothesized that this was related to the establishment of a GBM, which in embryonic development is formed by the fusion of two

basement membranes produced by both podocytes and ECs. Therefore, we investigated the expression of genes encoding specific collagens, laminins and heparan sulfate proteoglycans (HSPGs)

involved in the formation and maturation of the GBM by podocytes in untransplanted and transplanted organoids (Fig. 4a). We observed upregulation of _COL4A3–5_ and _LAMB2_ in d7+20

transplanted podocytes compared to all other conditions, fitting with the switch from collagen IV α1α2α1 to collagen IV α3α4α5 and from laminin β1 to laminin β2 during GBM maturation. Also,

_SULF1_, a regulator of heparan sulfates, is highly expressed in d7+20 transplanted podocytes. _LAMC1_ and _AGRN_, encoding laminin γ1 and agrin, a dominant laminin and HSPG in the mature

GBM, and _SULF2_ are not upregulated in d7+20 transplanted podocytes, implying the GBM in the organoids is not fully matured. To assess whether glomerular maturation and GBM formation were

also detectable at an ultrastructural level, we performed TEM imaging followed by stitching (Fig. 4b and Supplementary Fig. 1C). In contrast to traditional TEM, in which only small areas

selected based on low-resolution overviews are imaged at higher magnification, this yields large 2D high-resolution virtual slides18. After 8 days of transplantation, the invasion of

organoid glomerular structures by capillaries containing chicken-derived nucleated erythrocytes and leukocytes had resulted in the rearrangement of podocytes around the capillaries and the

deposition of a GBM between the podocytes and ECs. Podocyte cell bodies stretched out along the GBM, cell junctions between adjacent podocytes moved to the basal side of the cells, and

primitive foot processes were formed. Rudimentary slit diaphragms were observed between these primitive foot processes. ECs were variable in appearance but generally contained large nuclei

surrounded by relatively thick cell bodies, lacking the fenestrations typical for mature human glomeruli (Fig. 4b). Mesangial cells were visible in the stalk of some of the glomeruli,

adjacent to the ECs (Fig. 4b). The vascularized glomerular structures were surrounded by a single layer of parietal epithelial-like cells and a clearly distinguishable Bowman’s space had

appeared between the podocytes and parietal epithelial-like cells (Fig. 4b). The GSEA for d7+20 transplanted tubular epithelial cells versus untransplanted controls demonstrated enrichment

of gene sets associated with establishment and maintenance of cell polarity and of potassium ion transport (Supplementary Fig. 6B and Supplementary Data 11). Since the development of

transporters is indispensable for tubular functionality, we evaluated the expression of genes encoding several major tubular transporters in untransplanted and transplanted organoids. We

observed upregulation of genes encoding the sodium-potassium pump Na+/K+-ATP-ase (_ATP1A1, ATP1B1_, _and_ _ATP1B3_), the sodium potassium chloride cotransporter NKCC2 (_SLC12A1_) and the

renal outer medullary potassium channel ROMK (_KCNJ1_) in d7+20 transplanted organoids compared to the other conditions (Fig. 4c). The expression of _SCNN1A_, which encodes the α-subunit of

the epithelial sodium channel (ENaC) in the distal convoluted tubule, connecting tubule and collecting duct, was surprisingly highest in transplanted d7+13 organoids (Fig. 4c). Expression of

_SLC12A3_ and _AQP2_, encoding the sodium chloride cotransporter (NCC) in the distal convoluted tubule and the water channel aquaporin 2 in the collecting duct, was not detected in any

condition, consistent with the lack of clearly distinct distal convoluted tubule and collecting duct cell clusters (Fig. 2a–c). TEM analysis of tubular structures demonstrated that the lumen

widened upon transplantation, the epithelial cells formed a monolayer and, in accordance with the GSEA, showed signs of apicobasal polarity with their nuclei moving towards the basolateral

membrane. Abundant mitochondria were observed in their cytoplasm, and some of the tubular cells displayed cilia and microvilli, further reflecting ongoing maturation (Fig. 4d and

Supplementary Fig. 1D). GLOMERULAR MORPHOGENESIS TO CAPILLARY LOOP STAGE To evaluate the 3D organization of glomerular cells and structures in untransplanted and transplanted organoids, we

performed SBF-SEM. This technique utilizes an ultramicrotome mounted inside the vacuum chamber of a scanning electron microscope to serially image the block face of a tissue sample, of which

ultrathin sections are cut after the generation of each image. The result is a Z-stack of images that allows for the evaluation of the 3D tissue structure as well as ultrastructural

analysis at a subcellular level42. After generating datasets of 4 untransplanted and 3 transplanted glomerular structures (Supplementary Movies 1–4), artificial intelligence-based

segmentation and annotation was performed on 2 datasets (1 untransplanted and 1 transplanted) (Supplementary Movies 1, 2), followed by manual annotation of cell types, segmentation and 3D

visualization (Fig. 5a and Supplementary Movies 5, 6). The untransplanted glomerular structure consisted of a cluster of podocytes, surrounded by a layer of parietal epithelial-like cells

(Fig. 5b, c and Supplementary Movies 1, 5). After 8 days of transplantation, the glomerular structure had become vascularized. A perfused capillary containing erythrocytes as well as

leukocytes was clearly visible running along the side of the glomerular structure, before invading it and forming a single loop reaching deep into the center of the glomerulus before exiting

it (Fig. 5b, c, Supplementary Movies 2, 6). The podocytes reorganized into a layer surrounding the capillary and their cell bodies stretched out along the ECs (Fig. 5c). The parietal

epithelial-like cells assumed a more flattened morphology and formed a nearly complete capsule around the podocytes (Fig. 5c and Supplementary Movie 6). DISCUSSION In this study, we

demonstrate a new model to induce and study vasculogenesis and maturation in hiPSC-derived kidney organoids through transplantation inside the coelomic cavity of chicken embryos. The

procedure is easy and efficient, requiring around 5 min per embryo, and provides a scalable alternative to the labor-intensive transplantation in mice. We limited transplantation duration to

8 days in order to sacrifice the embryos on day 12 of incubation, a day before it is believed they start experiencing pain43,44. An extension to a maximum of 15 days is theoretically

possible, sacrificing the embryos on day 19 of incubation to avoid hatching. Endogenous organoid-derived ECs which diminish in in vitro culture, thrive upon transplantation, invade

glomerular structures and form a chimeric vasculature with invading host-derived endothelium. This implies that organoid vascularization occurs through a combination of vasculogenesis

(human, organoid-derived vessels) and angiogenesis (chicken-derived vessels), and is consistent with the main accepted theory that kidney vascularization in embryology depends on

angiogenic45,46,47,48,49 as well as vasculogenic mechanisms46,50,51,52. The ability of organoid-derived endothelium to contribute to the glomerular vasculature after intracoelomic

transplantation signifies the absence of unidentified cues for vascularization in in vitro culture conditions. It is well known that the production of VEGF-A by podocytes during embryonic

development is essential for recruitment, differentiation and maintenance of glomerular ECs5,53. Although it has been demonstrated that organoids in vitro produce VEGF10, either the levels

produced or the concentration gradient within the organoid could be insufficient. The addition of VEGF to organoid culture media, which increases VEGF level but does not establish a

gradient, supports EC proliferation, but fails to induce glomerular vascularization54. In our transplanted podocytes, we found upregulation of a gene set associated with VEGF production

(Supplementary Fig. 6A), and gene expression of VEGFA was clearly enhanced in podocytes compared to all other organoid cell clusters (Supplementary Fig. 5C) implying that the establishment

of a larger VEGF gradient by the podocytes could be a major facilitator of vascularization upon transplantation. Other factors that could play a role are shear stress and hypoxia. Shear

stress is an important regulator of EC maturation and establishment of the endothelial glycocalyx55,56,57. The static culture of kidney organoids is therefore likely to be suboptimal for

vascular development. Indeed, it has been shown that culturing organoids under flow leads to proliferation of ECs, although invasion of glomerular structures was still rare58. However, shear

stress in this study was much lower than reported as physiological in embryology57,59,60 and modifications were made to the culture conditions, such as addition of fetal bovine serum and

ECM, which could also have contributed to the observed changes. In embryology, nephron progenitor cells are thought to reside in a hypoxic environment until they are properly

vascularized61,62,63. The cells in our organoids, cultured at an air-liquid interface in an incubator with 20% oxygen, are exposed to much higher oxygen levels. The temporary hypoxic

conditions they are exposed to in the coelomic cavity of chicken embryos might therefore mimic physiologic conditions more closely. Unfortunately, the number of ECs in our scRNAseq dataset

was too small to evaluate the influence of the conditions discussed above on transplanted ECs. Analysis of the effect of transplantation on the cellular composition of kidney organoids

revealed an increase in proportion of mesenchymal cells in d7+20 compared to d7+13 organoids, with the highest proportion in transplanted d7+20 organoids. As concerns have been raised about

the proliferation of off-target cell types upon transplantation64, we analyzed the identity of the mesenchymal cells in our organoids. Interestingly, we found a lower percentage of

off-target cells and a higher proportion of pericytes/mesangial cells and fibroblasts in transplanted organoids at d7+20 compared to untransplanted controls. This is consistent with the

reduction of off-target cell types previously reported upon transplantation in mice24. The relatively short duration of transplantation in this paper (14 days) and our model (8 days) may

partly explain the lack of proliferation of off-target cell types, but not the observed difference with timepoint-matched in vitro controls. The increase in the pericyte/mesangial cell

population upon transplantation is an interesting finding. In the human kidney, pericytes play a role in angiogenesis, blood vessel stabilization and blood pressure regulation and have been

implied as a source of mesenchymal stem cells65. Mesangial cells are specialized pericytes that interact with glomerular ECs and podocytes to enable glomerular functionality66. Through LR

network analysis and TEM, we were able to demonstrate that the chimeric neovasculature in transplanted organoids is supported by perivascular stromal cells that interact with the ECs to

stabilize and mature the vascular network. Although the duration of transplantation in our study is relatively short compared to that reported in mouse studies, we observed significant

maturation of transplanted organoids. Comparison of our scRNAseq data to different transplantation methods would be very valuable. However, only one dataset of kidney organoids transplanted

in mice is currently publicly available24. Differences in cell line, differentiation protocol, duration of transplantation and dissociation method between this and our study prevent a fair

comparison. Hence, further studies, including direct comparison of kidney organoids submitted to different transplantation methods, would be required to investigate this. GSEA of the late

podocyte cluster and tubular epithelial cells revealed upregulation of genes involved in maturation of the GBM and in the development of tubular transporters upon transplantation. Evaluation

of the organoids at an ultrastructural level using TEM image mosaics confirmed the deposition of a GBM between the podocytes and the ECs that invaded the glomerular structures. In addition,

it demonstrated maturation of the podocytes with the formation of primitive foot processes and rudimentary slit diaphragms. Mesangial cells had followed the ECs into the glomerular

structures, providing support to the glomerular vasculature and enabling the interaction with podocytes and ECs that in in vivo kidneys is required for functionality66. Tubular epithelial

cells in transplanted organoids formed a monolayer, displayed signs of apicobasal polarization, and the tubular lumen widened. Through SBF-SEM followed by 3D reconstruction, we evaluated the

spatial organization of organoid glomeruli and visualized the trajectory of the vasculature in a transplanted glomerular structure. A single capillary was visible penetrating deep into the

glomerular structure and forming a single loop. It had not yet undergone intussusceptive angiogenesis to form a capillary tuft. The podocytes had reorganized around the capillary. In

conclusion, we have developed a new model to efficiently induce and study the development of a stabilized vascular network in kidney organoids through transplantation in the coelomic cavity

of chicken embryos, and demonstrate the essential role of vascularization for organoid maturation and glomerular morphogenesis. This scalable model can be used to identify essential early

cues for vasculogenesis and maturation of kidney organoids. METHODS IPSC MAINTENANCE AND DIFFERENTIATION hiPSC lines were maintained in Essential 8 medium (E8, Thermo Fisher Scientific)

supplemented with 0.5% Penicillin-Streptomycin (Thermo Fisher Scientific) on recombinant human Vitronectin (Thermo Fisher Scientific). All cell lines were mycoplasma free. hiPSCs were

passaged twice a week as small clumps using 0.5 mM UltraPure EDTA (Thermo Fisher Scientific). hiPSC reporter MAFB:mTagBFP2 (hiPSC-MAFB)67, LUMC0072iCTRL01, and LUMC0020iCTRL6.4 lines

(Detailed information can be found at https://hpscreg.eu/) were plated as single cells at 15,000–25,000 cells/cm2 one day prior to differentiation using TrypLE Select (Thermo Fisher

Scientific) and the addition of RevitaCell Supplement (ThermoFisher Scientific). hiPSCs were incubated from day 0–4 in 8 μM CHIR99021 (R&D Systems) in STEMdiff APEL2 medium (Stem Cell

Technologies) supplemented with 1% Protein Free Hybridoma Medium II (PFHMII, Thermo Fisher Scientific) and Antibiotic-Antimycotic solution (Thermo Fisher Scientific). From day 4–day 7 cells

were treated with 200 ng mL−1 rhFGF9 (R&D Systems) and 1 μg mL−1 heparin (Sigma Aldrich) in APEL2-medium. After an 1 h 5 μM CHIR pulse on day 7, cells were dissociated using Trypsin-EDTA

(0.25%, Thermo Fisher Scientific), counted and plated on Transwell 0.4 μM pore polyester membranes as pellets of 500,000 cells. These were cultured on an air–liquid interface for another 5

days in APEL2-medium containing 200 ng mL−1 rhFGF9 and 1 μg mL−1 heparin (only bottom compartment). For the remaining days, the organoids were cultured in medium without FGF9 and heparin,

and medium was changed every 2 days. Organoids were maintained on the transwell membranes until transplantation or day 7+20. TRANSPLANTATION IN CHICKEN EMBRYOS In accordance with Dutch law,

approval by the animal welfare committee was not required for these experiments. Fertilized White Leghorn eggs (_Gallus gallus domesticus_, Drost Loosdrecht B.V.) were placed horizontally in

a humidified incubator at 37 °C. After 3 days of incubation, a small hole was made in the tip of each egg by tapping it with the sharp end of a pair of dissecting scissors (Hammacher Karl,

HAMMHSB391-10), and 2–3 mL of albumen was removed using a syringe with a 19 gauge needle to lower the embryo inside the egg. A window was then cut into the egg shell using dissecting

scissors. 2–3 drops of DPBS with calcium and magnesium (Thermo Fisher Scientific) were added to the egg to avoid dehydration before sealing the window and the hole in the tip of the egg with

transparent tape (Tesa 4120). The window was made a day prior to transplantation, because we found that the chorioallantoic membrane, as well as the embryo, have frequently become attached

to the egg shell by day 4 of incubation, leading to bleeding and embryo death when attempting to create a window at this timepoint. On day 4 of incubation (Hamburger Hamilton stage 23),

kidney organoids on day 7+11 or day 7+12 of differentiation were bisected with a surgical knife and half an organoid was transplanted inside the coelomic cavity of each chicken embryo (Fig.

1a)22. Bisection of the organoids was necessary because the coelomic cavity is not large enough to accommodate a whole organoid at this stage of development. Transplantation was performed by

creating a small hole in the chorion and amnion membrane with forceps (Hammacher Karl, HAMMHTC090-11), and inserting the bisected organoid through this hole and the opening in the body wall

that is still present at this stage of development into the coelom using a blunt instrument (Fig. 1b). The eggs were resealed with transparent tape and further incubated for a maximum of 8

days until the end of the experiment. The whole process of transplantation takes about 5 min per egg, allowing for transplantation in up to 100 chicken embryos per person per day. Since

maintenance of the embryos before and after transplantation only requires a humidified incubator, this does not limit the number of embryos used for transplantation. On day 5 and day 12

(respectively 1 and 8 days after transplantation) of incubation, the chicken embryos were sacrificed and the organoids harvested. To enable analysis of the vasculature, some of the embryos

were carefully injected with 20 µL of 2.5 mg mL−1 rhodamine labeled _lens culinaris agglutinin_ (LCA) (RL-1042, Vector Laboratories)23 in the vitelline vein using a glass microcapillary

needle before being sacrificed. The injected lectin was allowed to circulate for 10 min before these embryos were sacrificed. For the timeline experiment (Supplementary Fig. 4B), the volume

of injected LCA was adjusted based on the age of the embryos: 10 µL was injected in day 5–7 embryos, 12 µL in day 8–9 embryos, 15 µL in day 10 embryos and 20 µL in day 11 embryos. To

retrieve the organoids, the abdominal wall of the embryos was opened along the longitudinal axis using forceps under a stereo microscope. Organoids were localized based on tissue morphology

and removed using micro scissors and a surgical knife. IMMUNOFLUORESCENCE ANALYSIS Organoids in vitro were fixed in 2% paraformaldehyde (PFA) at 4 °C for 20 min. Transplanted organoids were

fixed using 4% PFA at 4 °C overnight and thoroughly washed with PBS. They were used for whole mount staining. Non-transplanted organoids or transplanted tissues were permeabilized and

blocked in 0.3% TritonX in PBS containing 10% donkey serum for 2 h. Primary antibodies were diluted in blocking solution and incubated 24–72 h. Upon washing, secondary antibodies were

incubated for 2–4 h at room temperature. Kidney organoids were characterized for NPHS1 (AF4269, dilution 1:100), CD31 (555444, dilution 1:100), ECAD (610181, dilution 1:300), LTL-biotin

(B-1325, dilution 1:300), and PDGFRβ (MAB1263, dilution 1:50). Primary antibodies were detected with donkey-α-sheep Alexa Fluor 568 (A-21099, dilution 1:500) and 647 (A-21448, dilution

1:500), donkey-α-mouse Alexa Fluor 405 (ab175658, dilution 1:500) and 488 (A-212-02, dilution 1:500), streptavidin Alexa Fluor 532 (S11224, dilution 1:200) and 647 (S21374, dilution 1:200)

(Supplementary Table 1). All antibodies and isotype controls were validated in human kidney samples. Nuclei were stained with Hoechst33258 (Thermo Fisher Scientific) and tissues embedded in

ProLong Gold Antifade Mountant (Thermo Fisher Scientific) in 35 mm glass bottom dishes (MatTek corporation) or adhesive microscope slides (StarFrost, Knittel glass). Leica White Light Laser

Confocal Microscope TCS SP8 using LAS-X Image software with 3D module (Leica) was used for analysis of the tissues. QUANTIFICATION OF CD31 POSITIVE ENDOTHELIAL CELLS Quantification of CD31

positive endothelial cells was performed in whole mount half or whole untransplanted and transplanted organoids at d7+19–20. Imaris Software was used to calculate the volume of CD31 positive

cells as a percentage of whole organoid volume. To select the organoid tissue, a surface was generated manually by drawing a contour around the organoids every 5 slides of the Z-stack. This

surface was used to calculate the volume of the organoid with the statistics function in Imaris. Next, a surface was created for the CD31 channel, and the volume of this surface was

obtained in the same manner. In case of transplanted organoids with high background signal in the surrounding chicken tissue, a mask was created for the CD31 channel based on the previously

generated organoid surface. This masked CD31 channel was then used to create a surface and calculate the CD31 volume. Threshold and minimum number of voxels for CD31 surface generation were

set separately for each sample, to be able to correct for differences in background and if necessary, background falsely identified as endothelial cells by the software was removed manually.

This analysis was performed for 6 untransplanted organoids, all from different differentiation batches (3 from iPSC-MAFB, 3 from LUMC0072) and 6 matched transplanted organoids. Graphpad

Prism 9.0.1 was used for statistical analysis. Means were compared between groups using an unpaired two-tailed t-test. The normality of the distribution was tested and confirmed using the

Shapiro–Wilk and Kolmogorov–Smirnov test. Individual data points and mean (SD) are presented. SINGLE CELL RNA SEQUENCING SAMPLE PREPARATION Untransplanted and transplanted (on day 7+12)

organoids from the same differentiation of the hiPSC-MAFB line were dissociated to single cells at day 7+13 and d7+20 using a collagenase I buffer consisting of 600U ml−1 collagenase Type I

(Worthington) and 0.75 U ml−1 DNAse (Sigma Aldrich) in HBSS with calcium and magnesium (Thermo Fisher Scientific) followed by a TrypLE buffer consisting of 5U ml−1 DNAse I (Sigma Aldrich)

and 4 µg ml−1 heparin (Sigma Aldrich) in 80% TrypLE select 10× (Thermo Fisher Scientific) in DPBS (Thermo Fisher Scientific). 5 organoids were placed in Collagenase I buffer and incubated in

a water bath at 37 °C for 40 min with repeated pipetting. The cell suspensions were centrifuged at 300G, and the cell pellet was incubated at 37 °C for another 5 min in TrypLE buffer. The

dissociation was stopped by adding cold HBSS+/+ with 10% FCS, and the suspension further diluted with HBSS+/+. The single cell suspension was centrifuged at 400 G and the cell pellet was

resuspended in PBS+0.1% BSA. All cells from organoids from the same condition were pooled (d7+13 untransplanted: 94 organoids, transplanted: 75 bisected organoids (transplanted in 150

chicken embryos) d7+20 untransplanted: 95 organoids, transplanted: 46,5 bisected organoids (transplanted in 93 chicken embryos)) and 30 mL of PBS + 0.1%BSA was added to enable cell counting.

The single cell suspensions were converted to barcoded scRNA-seq libraries with a targeted cell recovery of 9000 cells/condition using the Chromium Single Cell 3’ v3 Library, Gel Bead &

Multiplex Kit and Chip Kit (10x Genomics). DEEP SEQUENCING AND DATA PRE-PROCESSING The scRNA-seq libraries were converted into DNA nanoballs (DNB) using a standard circularization protocol

optimized for the DNBSEQ-T7 sequencer, and all subsequent steps were carried out following the standard operation procedure68,69. The scRNA-seq libraries were sequenced aiming for at least

50,000 reads/cells, and using 28+100+8 bp paired‐end sequencing to determine (1) the cell barcode and UMI, (2) the transcript and (3) the sample index, respectively. Demultiplexing according

to the sample barcodes, and subsequent read alignment were performed using Cell Ranger (10x Genomics, v3.1.0). For untransplanted kidney organoid samples, reads were aligned to human genome

(GRCh38) only. For transplanted kidney organoid samples, a combined genome including human (GRCh38) and chicken (GRCg6a) genomes, was prepared using the _mkgtf_ and _mkref_ commands from

Cell Ranger (10X genomics), following 10X genomics instructions. Reads associated to these samples were then aligned to this reference. Key sequencing metrics are summarized in Supplementary

Data 1. Raw unfiltered data matrices from the Cell Ranger output were then further processed with R (v4.1.1) and Seurat package (v4.0.4). For multi-species samples, the percentage of genes

belonging to each species was determined using the _PercentageFeatureSet_ function from Seurat package, and cells were labeled either “human” or “chicken” when >85% of total genes were

mapped to the respective species, whereas ambiguous cells (unlabeled) were discarded. Cells from multi-species samples were then segregated per species for downstream processing, resulting

in both a human and a chicken dataset, for each transplanted kidney organoid sample. The following quality control steps were performed for each dataset: (i) genes expressed by less than 10

cells were removed; (ii) cells that expressed fewer than 600 genes (for human datasets) and 200 genes (for chicken datasets) were discarded as low-quality cells (Supplementary Data 1); (iii)

cells with a detected number of genes exceeding a “doublet” threshold as listed in Supplementary Data 1 were excluded (determined by inspecting the cell frequency per total number of genes

expressed, for each sample); (iv) cells with a fraction of mitochondrial genes >15% (for human datasets) and >5% (for chicken datasets) were also removed (cells with compromised cell

membrane/dying or dead cells). After pre-processing, human cells-containing samples were merged as well as chicken cells-containing samples, resulting in 24,988 human cells (27,349 human

cells before removal of dying/dead cells), and in 8018 chicken cells (8279 chicken cells before removal of dying/dead cells). SCRNASEQ DATA ANALYSIS Data were normalized using the

_NormalizeData_ function from Seurat Package, and the top 2000 highly variable genes were identified using _FindvariableFeatures_. Data were scaled with the _Scaledata_ function and reduced

with principal component (PC) analysis using _RunPCA_ function. The top 30 PCs (40 PCs for chicken cells) were used for visualization using uniform manifold approximation and projection

(UMAP) with the _RunUMAP_ function as implemented in Seurat. Next, a shared nearest-neighbor graph (SNN) was determined using _FindNeighbors_ function, and used to calculate clusters with

the _FindClusters_ function. Top marker genes were calculated with the _Findallmarkers_ function using default settings. Remaining low-quality cell and doublet clusters were characterized by

a low number of genes expressed per cell, and an overlap of different cell type markers with high number of genes expressed per cell, respectively, and removed from the dataset.

High-quality cells were split according to condition (transplantation & timepoint) and aligned using _Seurat_ integration tool: _FindintegrationAnchors_ and _IntegrateData_ functions,

with 2000 anchors and 40 PCs as settings. Clustering was performed as described above using 40 PCs and main cell types for human and chicken datasets were identified. For the human dataset,

nephron cell (_NPHS1, NPHS2, MFAB, WT1, EPCAM, KRT19, …_) and mesenchymal cell clusters (_COL1A1, COL1A2, COL3A1, TAGLN, PDGFRA_, …) were subsetted using the _subset_ function for more

in-depth analysis. Briefly, for both the mesenchymal and nephron cell clusters, 2000 highly variable genes were identified, and data were integrated per condition (transplantation &

timepoint) using 2,000 anchors and 40 PCs as settings, as described above. Integrated data were further scaled, reduced within 40 PCs that were further used for UMAP visualization, and

clustering, as described above. Upregulated markers were identified using the _FindAllMarkers_ function, including genes for which Log2FC > 0 only, and using other parameters as default.

Cluster identity was determined according to cell-(sub)type specific top marker genes: podocyte markers as _MAFB, NPHS1, NPHS2, WT1_, epithelial cell markers a_s EPCAM_ and _KRT19_ for

tubular epithelial cells, and fibroblast markers as _COL1A1, COL1A2, COL3A1_, smooth muscle cell markers as _ACTA2_, _TAGLN_, _MYL9_, pericyte/mesangial cell markers as _REN, GATA3, PDGFRB_,

mesenchymal progenitor markers as _PRRX1_ and _PRRX2_, neural-like cell markers as _NTREK2, METRN, GAP43, STMN2_, chondrocyte-like cell markers as _COL9A1, COL9A2, COL9A3, MATN4_, and

melanocyte-like cell markers as _PMEL, MITF, MLANA_. As for the chicken cells, the following markers were used to identify clusters: erythroid cells (_HBA1, HBAD,_ and _HBE1_),

megakaryocytes (_GP9, ITGB3,_ and _GP1BB_), endothelial cells (_PECAM1, CDH5,_ and _PROX1_), macrophages (_CD74_ and _C1QB_), granulocytes (_DEFB4A and AvBD1_), lymphoid cells (_CD3D_ and

_IGLL1_), fibroblasts (_COL3A1_ and _COL14A1_), pericytes (_KCNJ8, RGS5,_ and _ABCC9_), smooth muscle cells (_ACTA2, TAGLN,_ and _ACTG2_), hepatic stellate cells (_COLEC11_ and _HGF_),

epithelioid cells (_KRT24_ and _KRT1_), hepatocytes (_ALB, FGB,_ and _FGG_), Leydig cells (_CYP11A1, STAR,_ and _HSD3B1_), Sertoli cells (_AMH_ and _SOX9_), Schwann cells (_NRN1, SOX10,_ and

_CRYAB_) and neurons (_GAP43, NEFM,_ and _NSG1_). Proliferative cell clusters were identified as either within G1/S phases or G2/M phases according to canonical marker genes, as listed in

Seurat. DOT PLOT AND HEATMAP VISUALIZATION All dotplots were prepared using _DotPlot_ function as implemented in Seurat. All heatmaps were prepared with BIOMEX software (v1.5 -

https://carmelietlab.sites.vib.be/en/biomex)70, except heatmaps displaying gene expression of ligand and receptors, which were prepared with the ComplexHeatmap package (v2.11.1) in R. For

heatmaps prepared in BIOMEX, briefly, count data matrix was exported from the Seurat object using the _GetAssayData_ function (selected assay “RNA”) and uploaded in BIOMEX altogether with

the corresponding metadata information. Data were normalized (“Standard” parameter) and (auto)scaled for heatmap visualization within BIOMEX. Cluster-averaged gene expression was used to

account for cell-to-cell transcriptomic stochastics. Top marker gene heatmaps were prepared, including uniquely upregulated genes resulting from the marker gene identification tool from

BIOMEX. For heatmaps prepared with the ComplexHeatmap package, briefly, the averaged expression profiles per each cluster and for all genes were calculated using the _AverageExpression()_

function from Seurat package. Next, scaled data were obtained from the RNA assay using _GetAssayFunction()_ function and selected gene data were extracted from the resulting matrix. These

values were used as input for the _Heatmap()_ function from the ComplexHeatmap package. CLUSTER SIMILARITY ANALYSIS Hierarchical clustering was performed with BIOMEX using highly variable

genes as identified in BIOMEX, with Euclidean distance and complete linkage. The confidence of each branch of the tree was estimated by the bootstrap resampling approach from the R-package

_pvclust_71, using a confidence score of >0.05. Number of bootstrapping was set at 1000. METADATA QUANTIFICATION Metadata from the Seurat object were exported, as described above and

uploaded in BIOMEX. Metadata were quantified and visualized using the metadata quantification tool, as implemented in BIOMEX. “barplots” was used as plot setting. DIFFERENTIAL GENE

EXPRESSION ANALYSIS (DGEA) For DGEA according to the transplantation condition, the _FindMarkers_ function from Seurat was used on the “RNA” assay. Differential expression was run with the

MAST package. The number of UMIs per cell was used as latent variable to correct for unequal sequencing depth between cells. GENESET ENRICHMENT ANALYSIS (GSEA) DGEA outputs were used for

GSEA using the _ClusterProfiler_ package (v4.0.5). GSEA was performed on gene ontology genesets (biological processes only) with _gseGO_ function, as implemented in _ClusterProfiler_.

Genesets including less than 10 genes were not included. All other parameters were used as default. In order to remove as much as possible redundant terms, results were simplified using the

online Revigo tool (http://revigo.irb.hr) for both significantly up- and downregulated genesets with Normalized Enrichment Scores used as ranking method. The resulting list was aimed to be

small (selected 0.5 parameter). Output from Revigo was exported and plotted using _ggplot_ function, as implemented within the _ggplot2_ package (v3.3.5). LIGAND/RECEPTOR INTERACTION

ANALYSIS (L/R ANALYSIS) For L/R analysis, _CellChat_ package (v1.1.3) was used. First, the melanocyte-like cluster was removed from the analysis since it was absent from the transplanted

d7+20 kidney organoid condition. The normalized data matrix “data” slot, from the “RNA” assay within the Seurat object was extracted for each condition (d7+20 condition, and the other

conditions as pooled, or taken separately for the comparison of the total number of interactions and interaction probabilities between the four conditions) using _GetAssayData()_ function

from the Seurat package. The metadata containing final cluster identities were extracted as a data frame. Data and metadata were used to create CellChat objects using _createCellChat()_

function. The complete human cellchat database was used. Expression data was subsetted to genes present in the database using _subsetData()_ function and overexpressed genes and interactions

were identified for each condition with _identifyOverExpressedGenes()_ and _identifyOverExpressedInteractions()_ functions, respectively. Communication probabilities and inference of the

cellular communication network were calculated using _computeCommunProb()_, _computeCommunProbPathway()_, _aggregateNet()_ and _netAnalysis_computeCentrality()_ functions. Cellchat objects

were merged with the dedicated _mergeCellChat()_ function. Comparison of the total number of interaction and interaction strength was performed using _compareInteraction()_ function setting

the “measure” argument as “weight” when appropriate. The _netVisual_heatmap()_ function, setting the “measure” argument as “weight”, was used to generate a heatmap of differential

interaction strength between cell populations. To identify the upregulated signaling ligand-receptor pairs between ECs and pericytes/mesangial cells, we compared the communication

probabilities between conditions and visualized the results as a dot plot using the _netVisual_bubble()_ function, with “max.dataset” argument set as “2”. TRANSMISSION ELECTRON MICROSCOPY

(TEM) For morphological analysis, small sections of in vitro kidney organoids and transplanted organoids were fixed for 2 h at room temperature in 1.5% glutaraldehyde (Electron Microscopy

Sciences) in 0.1 M sodium cacodylate buffered solution (pH 7.4). The tissues were further rinsed with sodium cacodylate buffer and fixed in a solution of 1% osmium tetroxide (Electron

Microscopy Sciences) in 0.1 M sodium cacodylate buffer for 1 h on ice. Afterwards, samples were washed with sodium cacodylate buffer and dehydrated in a series of 70, 80, 90, and 100%

ethanol. The probes were infiltrated with a mixture of 1:1 Epon LX-112 (Ladd Research) and propylene oxide (Electron Microscopy Sciences) for 1 h, followed by infiltration with pure Epon for

2 h. Subsequently, the samples were embedded in pure Epon, mounted in BEEM capsules (Agar Scientific) and polymerized for 48 h at 60 °C. Ultrathin sections (100 nm) were collected onto

copper slot grids (Storck Veco B.V.), covered with formvar film and a 7 nm carbon layer. The sections were contrasted with an aqueous solution of 7% uranyl acetate for 20 min, followed by

Reynolds lead citrate for 10 min. The imaging was performed at an acceleration voltage of 120 kV with a FEI Tecnai G2 Spirit BioTWIN TEM (FEI), equipped with an Eagle 4 K slow-scan

charge-coupled device (CCD) camera (FEI). Data were collected at 18,500× magnification, corresponding to a 1.2 nm pixel size at the specimen level. Large virtual slides showing glomerular

structures were acquired using automated large-scale data collection combined with stitching software18. These extensive digital images provide an overview of entire glomeruli and the

possibility to zoom in to high detail, allowing for qualitative analysis. SERIAL BLOCK FACE SCANNING ELECTRON MICROSCOPY (SBF-SEM) For SBF-SEM analysis, 2 transplanted organoids and 1

untransplanted organoid, all generated using iPSC-MAFB were used. Kidney organoid fixation, staining and embedding was done according to an adapted protocol from Deerink et al.72 Samples

were fixed for 2 h on ice with 2.5% glutaraldehyde and 2% paraformaldehyde (Electron Microscopy Sciences) in 0.15 M sodium cacodylate buffer containing 2 mM calcium chloride. A first

post-fixation step was performed for 1 h on ice, using a solution of 2% osmium tetroxide and 1.5% potassium ferrocyanide in 0.15 M sodium cacodylate buffer containing 2 mM calcium chloride.

Subsequently samples were treated for 20 min at room temperature with 1% aqueous solution of thiocarbohydrazide, followed by 2% aqueous osmium tetroxide for 30 min at room temperature and

incubated overnight at 4 °C in a 1% aqueous solution of uranyl acetate. The probes were then stained for 30 min at 60 °C, using Walton’s lead aspartate, and dehydrated in series of 20% up to

100% ethanol, followed by dehydration in anhydrous acetone. Tissue samples were infiltrated in mixtures of 1:3, 1:1, 3:1 Durcupan ACM resin (Ladd Research) and acetone, in 2 h steps, then

overnight at room temperature in 100% Durcupan. A final infiltration step with fresh Durcupan was performed for 2 h at room temperature. The samples were then placed in BEEM capsules,

embedded in Durcupan and polymerized for 48 h at 60 °C. Sample blocks were trimmed and 300 nm thick sections were cut and stained with toluidine blue for light microscopy to find regions of

interest containing glomerular structures. For SBF-SEM, small blocks of approximately 800 × 800 × 400 µm were trimmed and glued on aluminum pins (Gatan) using Conductive Silver Epoxy Kit

(Agar Scientific), then coated with a thin layer of gold right before imaging. A data set from transplanted and non-transplanted organoid was collected on a GeminiSEM 300 microscope (ZEISS)

equipped with a 3View 2XP system (Gatan). The SEM was operated at a chamber pressure of 5 Pa, in variable pressure mode, using an acceleration voltage of 1.6 kV and the 60 µm aperture.

Back-scattered electron images were acquired using the OnPoint BSE detector (Gatan). 4 datasets were generated of untransplanted glomerular structures and 3 of transplanted glomerular

structures, of which 2 (1 untransplanted and 1 transplanted) were used for 3D visualization. The data set from the non-transplanted organoid was acquired at 10,000× magnification, with a

pixel size of 9 nm, a dwell time of 2.8 µs and a 50 nm slice thickness. The sample size was 61 × 61 × 28 µm, and the image size 6800 × 6800 pixels × 557 slices. The data set from the

transplanted kidney organoid was acquired at 9000× magnification, a pixel size of 10 nm, a dwell time of 1.5 µs and a 60 nm slice thickness. The sample size was 65 × 75 × 27 µm, and the

image size 6500 × 7500 pixels × 451 slices. SBF-SEM image stacks were acquired as dm4 files and aligned using Digital Micrograph software (Gatan). ANNOTATION AND VISUALIZATION OF SBF-SEM

DATA 2 SBF-SEM datasets (1 untransplanted and 1 transplanted) were annotated using artificial intelligence-based image segmentation by Ariadne (Ariadne-Service GmbH). In short, nuclei,

extracellular space (ECS), and plasma membrane labels were predicted using a convolutional neural network (CNN) architecture based on the U-Net73. Stochastic weight averaging was used in the

ER prediction74. Somata were segmented by running the watershed algorithm on the plasma membrane probability map, using the predicted nucleus locations as seeds and excluding the ECS

regions. CNNs were implemented in the ELEKTRONN deep learning library (www.elektronn.org). Image stacks of segmented structures (~12 Gb or ~25 Gb in size) were binned (in X and Y) using

Fiji75 or IMOD76 to decrease the file size, and converted to mrc stacks and the headers were corrected using IMOD. Cell types were annotated by hand according to morphology and location and

surfaces were smoothed, simplified and visualized using AMIRA 3D software for life sciences (Thermo-Fisher Scientific). Surfaces were imported into Cinema4D (Maxon) for final rendering using

Redshift (Maxon). REPORTING SUMMARY Further information on research design is available in the Nature Research Reporting Summary linked to this article. DATA AVAILABILITY The source data

underlying Figs. 1f, h, 2d, h, 3a and Supplementary Fig. 6a, b are provided as a source data file. Both raw and processed sequencing data are available in ArrayExpress under accession number

E-MTAB-11429. Other data are available from the authors upon request. CODE AVAILABILITY Custom code is available upon reasonable request. REFERENCES * Taguchi, A. et al. Redefining the in

vivo origin of metanephric nephron progenitors enables generation of complex kidney structures from pluripotent stem cells. _Cell Stem Cell_ 14, 53–67 (2014). Article  CAS  PubMed  Google

Scholar  * Morizane, R. et al. Nephron organoids derived from human pluripotent stem cells model kidney development and injury. _Nat. Biotechnol._ 33, 1193–1200 (2015). Article  CAS  PubMed

  PubMed Central  Google Scholar  * Takasato, M. et al. Kidney organoids from human iPS cells contain multiple lineages and model human nephrogenesis. _Nature_ 526, 564–568 (2015). Article 

CAS  PubMed  Google Scholar  * Freedman, B. S. et al. Modelling kidney disease with CRISPR-mutant kidney organoids derived from human pluripotent epiblast spheroids. _Nat. Commun._ 6, 8715

(2015). Article  CAS  PubMed  Google Scholar  * Eremina, V. et al. Glomerular-specific alterations of VEGF-A expression lead to distinct congenital and acquired renal diseases. _J. Clin.

Invest._ 111, 707–716 (2003). Article  CAS  PubMed  PubMed Central  Google Scholar  * Kitamoto, Y., Tokunaga, H. & Tomita, K. Vascular endothelial growth factor is an essential molecule

for mouse kidney development: Glomerulogenesis and nephrogenesis. _J. Clin. Invest._ 99, 2351–2357 (1997). Article  CAS  PubMed  PubMed Central  Google Scholar  * Sison, K. et al. Glomerular

structure and function require paracrine, not autocrine, VEGF-VEGFR-2 signaling. _J. Am. Soc. Nephrol._ 21, 1691–1701 (2010). Article  CAS  PubMed  PubMed Central  Google Scholar  *

Sharmin, S. et al. Human induced pluripotent stem cell-derived podocytes mature into vascularized glomeruli upon experimental transplantation. _J. Am. Soc. Nephrol._ 27, 1778–1791 (2016).

Article  CAS  PubMed  Google Scholar  * Bantounas, I. et al. Generation of functioning nephrons by implanting human pluripotent stem cell-derived kidney progenitors. _Stem Cell Rep._ 10,

766–779 (2018). Article  Google Scholar  * van den Berg, C. W. et al. Renal subcapsular transplantation of PSC-derived kidney organoids induces neo-vasculogenesis and significant glomerular

and tubular maturation in vivo. _Stem Cell Rep._ 10, 751–765 (2018). Article  Google Scholar  * van den Berg, C. W., Koudijs, A., Ritsma, L. & Rabelink, T. J. In vivo assessment of

size-selective glomerular sieving in transplanted human induced pluripotent stem cell-derived kidney organoids. _J. Am. Soc. Nephrol._ 31, 921–929 (2020). Article  PubMed  PubMed Central 

Google Scholar  * Garreta, E. et al. Fine tuning the extracellular environment accelerates the derivation of kidney organoids from human pluripotent stem cells. _Nat. Mater._ 18, 397–405

(2019). Article  CAS  PubMed  Google Scholar  * Jankovic, B. D. et al. Immunological capacity of the chicken embryo. I. Relationship between the maturation of lymphoid tissues and the

occurrence of cell-mediated immunity in the developing chicken embryo. _Immunology_ 29, 497–508 (1975). CAS  PubMed  PubMed Central  Google Scholar  * Alkie, T. N. et al. Development of

innate immunity in chicken embryos and newly hatched chicks: a disease control perspective. _Avian Pathol._ 48, 288–310 (2019). Article  PubMed  Google Scholar  * Hamburger, V. Morphogenetic

and axial self-differentiation of transplanted limb primordia of 2-day chick embryos. _J. Exp. Zool._ 77, 379–399 (1938). Article  Google Scholar  * Rawles, M. E. Transplantation of normal

embryonic tissues. _Ann. N. Y Acad. Sci._ 55, 302–312 (1952). Article  CAS  PubMed  Google Scholar  * Rawles, M. E. The development of melanophores from embryonic mouse tissues grown in the

coelom of chick embryos. _Proc. Natl Acad. Sci. USA_ 26, 673–680 (1940). Article  CAS  PubMed  PubMed Central  Google Scholar  * Faas, F. G. et al. Virtual nanoscopy: Generation of

ultra-large high resolution electron microscopy maps. _J. Cell Biol._ 198, 457–469 (2012). Article  CAS  PubMed  PubMed Central  Google Scholar  * Dossel, W. E. New method of intracoelomic

grafting. _Science_ 120, 262–263 (1954). Article  CAS  PubMed  Google Scholar  * Takasato, M., Er, P. X., Chiu, H. S. & Little, M. H. Generation of kidney organoids from human

pluripotent stem cells. _Nat. Protoc._ 11, 1681–1692 (2016). Article  CAS  PubMed  PubMed Central  Google Scholar  * Ryan, A. R. et al. Vascular deficiencies in renal organoids and ex vivo

kidney organogenesis. _Dev. Biol._ 477, 98–116 (2021). Article  CAS  PubMed  PubMed Central  Google Scholar  * Hamburger, V. & Hamilton, H. L. A series of normal stages in the

development of the chick embryo. 1951. _Dev. Dyn._ 195, 231–272 (1992). Article  CAS  PubMed  Google Scholar  * Jilani, S. M. et al. Selective binding of lectins to embryonic chicken

vasculature. _J. Histochem. Cytochem._ 51, 597–604 (2003). Article  CAS  PubMed  Google Scholar  * Subramanian, A. et al. Single cell census of human kidney organoids shows reproducibility

and diminished off-target cells after transplantation. _Nat. Commun._ 10, 5462 (2019). Article  PubMed  PubMed Central  CAS  Google Scholar  * Carmeliet, P. Angiogenesis in health and

disease. _Nat. Med._ 9, 653–660 (2003). Article  CAS  PubMed  Google Scholar  * Dejana, E. Endothelial cell-cell junctions: Happy together. _Nat. Rev. Mol. Cell Biol._ 5, 261–270 (2004).

Article  CAS  PubMed  Google Scholar  * Gu, C. & Giraudo, E. The role of semaphorins and their receptors in vascular development and cancer. _Exp. Cell Res._ 319, 1306–1316 (2013).

Article  CAS  PubMed  PubMed Central  Google Scholar  * Brunskill, E. W. & Potter, S. S. Changes in the gene expression programs of renal mesangial cells during diabetic nephropathy.

_BMC Nephrol._ 13, 70 (2012). Article  PubMed  PubMed Central  Google Scholar  * Kida, Y., Ieronimakis, N., Schrimpf, C., Reyes, M. & Duffield, J. S. EphrinB2 reverse signaling protects

against capillary rarefaction and fibrosis after kidney injury. _J. Am. Soc. Nephrol._ 24, 559–572 (2013). Article  CAS  PubMed  PubMed Central  Google Scholar  * Gerety, S. S. &

Anderson, D. J. Cardiovascular ephrinB2 function is essential for embryonic angiogenesis. _Development_ 129, 1397–1410 (2002). Article  CAS  PubMed  Google Scholar  * Foo, S. S. et al.

Ephrin-B2 controls cell motility and adhesion during blood-vessel-wall assembly. _Cell_ 124, 161–173 (2006). Article  CAS  PubMed  Google Scholar  * Salvucci, O. & Tosato, G. Essential

roles of EphB receptors and EphrinB ligands in endothelial cell function and angiogenesis. _Adv. Cancer Res._ 114, 21–57 (2012). Article  CAS  PubMed  PubMed Central  Google Scholar  * Lu,

L. et al. Growth differentiation factor 11 promotes neurovascular recovery after stroke in mice. _Front. Cell Neurosci._ 12, 205 (2018). Article  PubMed  PubMed Central  CAS  Google Scholar

  * Takabatake, Y. et al. The CXCL12 (SDF-1)/CXCR4 axis is essential for the development of renal vasculature. _J. Am. Soc. Nephrol._ 20, 1714–1723 (2009). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Astrof, S. & Hynes, R. O. Fibronectins in vascular morphogenesis. _Angiogenesis_ 12, 165–175 (2009). Article  CAS  PubMed  PubMed Central  Google Scholar  *

Kofler, N. M., Cuervo, H., Uh, M. K., Murtomaki, A. & Kitajewski, J. Combined deficiency of Notch1 and Notch3 causes pericyte dysfunction, models CADASIL, and results in arteriovenous

malformations. _Sci. Rep._ 5, 16449 (2015). Article  PubMed  PubMed Central  Google Scholar  * Kofler, N. M. et al. Notch signaling in developmental and tumor angiogenesis. _Genes Cancer_ 2,

1106–1116 (2011). Article  PubMed  PubMed Central  CAS  Google Scholar  * Armulik, A., Genove, G. & Betsholtz, C. Pericytes: developmental, physiological, and pathological perspectives,

problems, and promises. _Dev. Cell_ 21, 193–215 (2011). Article  CAS  PubMed  Google Scholar  * Darland, D. C. et al. Pericyte production of cell-associated VEGF is

differentiation-dependent and is associated with endothelial survival. _Dev. Biol._ 264, 275–288 (2003). Article  CAS  PubMed  Google Scholar  * Kosyakova, N. et al. Differential functional

roles of fibroblasts and pericytes in the formation of tissue-engineered microvascular networks in vitro. _NPJ Regen. Med._ 5, 1 (2020). Article  CAS  PubMed  PubMed Central  Google Scholar

  * Woolf, A. S., Gnudi, L. & Long, D. A. Roles of angiopoietins in kidney development and disease. _J. Am. Soc. Nephrol._ 20, 239–244 (2009). Article  CAS  PubMed  Google Scholar  *

Denk, W. & Horstmann, H. Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. _PLoS Biol._ 2, e329 (2004). Article  PubMed  PubMed

Central  CAS  Google Scholar  * ACUC. _ACUC Guideline: The Use and Euthanasia Procedures of Chicken/Avian Embryos_.

https://www.cpp.edu/research/research-compliance/iacuc/docs/iacuc-guidelines-on-euthanasia-of-chicken-and-embryos.pdf (2012). * Aleksandrowicz, E. & Herr, I. Ethical euthanasia and

short-term anesthesia of the chick embryo. _ALTEX_ 32, 143–147 (2015). PubMed  Google Scholar  * Munro, D. A. D., Hohenstein, P. & Davies, J. A. Cycles of vascular plexus formation

within the nephrogenic zone of the developing mouse kidney. _Sci. Rep._ 7, 3273 (2017). Article  PubMed  PubMed Central  CAS  Google Scholar  * Hyink, D. P. et al. Endogenous origin of

glomerular endothelial and mesangial cells in grafts of embryonic kidneys. _Am. J. Physiol._ 270, F886–F899 (1996). CAS  PubMed  Google Scholar  * Rogers, S. A. & Hammerman, M. R.

Transplantation of rat metanephroi into mice. _Am. J. Physiol. Regul. Integr. Comp. Physiol._ 280, R1865–R1869 (2001). Article  CAS  PubMed  Google Scholar  * Takeda, S., Rogers, S. A. &

Hammerman, M. R. Differential origin for endothelial and mesangial cells after transplantation of pig fetal renal primordia into rats. _Transpl. Immunol._ 15, 211–215 (2006). Article  CAS 

PubMed  Google Scholar  * Dekel, B. et al. Human and porcine early kidney precursors as a new source for transplantation. _Nat. Med._ 9, 53–60 (2003). Article  CAS  PubMed  Google Scholar  *

Robert, B., St John, P. L., Hyink, D. P. & Abrahamson, D. R. Evidence that embryonic kidney cells expressing flk-1 are intrinsic, vasculogenic angioblasts. _Am. J. Physiol._ 271,

F744–F753 (1996). CAS  PubMed  Google Scholar  * Sims-Lucas, S. et al. Endothelial progenitors exist within the kidney and lung mesenchyme. _PLoS One_ 8, e65993 (2013). Article  CAS  PubMed

  PubMed Central  Google Scholar  * Loughna, S. et al. A molecular and genetic analysis of renalglomerular capillary development. _Angiogenesis_ 1, 84–101 (1997). Article  CAS  PubMed 

Google Scholar  * Bartlett, C. S., Jeansson, M. & Quaggin, S. E. Vascular growth factors and glomerular disease. _Annu. Rev. Physiol._ 78, 437–461 (2016). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Czerniecki, S. M. et al. High-throughput screening enhances kidney organoid differentiation from human pluripotent stem cells and enables automated

multidimensional phenotyping. _Cell Stem Cell_ 22, 929–940e924 (2018). Article  CAS  PubMed  PubMed Central  Google Scholar  * Arora, S., Lam, A. J. Y., Cheung, C., Yim, E. K. F. & Toh,

Y. C. Determination of critical shear stress for maturation of human pluripotent stem cell-derived endothelial cells towards an arterial subtype. _Biotechnol. Bioeng._ 116, 1164–1175 (2019).

Article  CAS  PubMed  Google Scholar  * Arisaka, T. et al. Effects of shear stress on glycosaminoglycan synthesis in vascular endothelial cells. _Ann. N. Y Acad. Sci._ 748, 543–554 (1995).

Article  CAS  PubMed  Google Scholar  * Adamo, L. et al. Biomechanical forces promote embryonic haematopoiesis. _Nature_ 459, 1131–1135 (2009). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Homan, K. A. et al. Flow-enhanced vascularization and maturation of kidney organoids in vitro. _Nat. Methods_ 16, 255–262 (2019). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Gautam, M., Shen, Y., Thirkill, T. L., Douglas, G. C. & Barakat, A. I. Flow-activated chloride channels in vascular endothelium. Shear stress sensitivity, desensitization

dynamics, and physiological implications. _J. Biol. Chem._ 281, 36492–36500 (2006). Article  CAS  PubMed  Google Scholar  * Buchanan, C. F., Verbridge, S. S., Vlachos, P. P. & Rylander,

M. N. Flow shear stress regulates endothelial barrier function and expression of angiogenic factors in a 3D microfluidic tumor vascular model. _Cell Adh Migr._ 8, 517–524 (2014). Article 

PubMed  PubMed Central  Google Scholar  * Hemker, S. L., Sims-Lucas, S. & Ho, J. Role of hypoxia during nephrogenesis. _Pediatr. Nephrol._ 31, 1571–1577 (2016). Article  PubMed  PubMed

Central  Google Scholar  * Gerosa, C. et al. Low vascularization of the nephrogenic zone of the fetal kidney suggests a major role for hypoxia in human nephrogenesis. _Int. Urol. Nephrol._

49, 1621–1625 (2017). Article  CAS  PubMed  Google Scholar  * Rymer, C. et al. Renal blood flow and oxygenation drive nephron progenitor differentiation. _Am. J. Physiol. Ren. Physiol._ 307,

F337–F345 (2014). Article  CAS  Google Scholar  * Nam, S. A. et al. Graft immaturity and safety concerns in transplanted human kidney organoids. _Exp. Mol. Med._ 51, 145 (2019). Article 

CAS  Google Scholar  * Shaw, I., Rider, S., Mullins, J., Hughes, J. & Péault, B. Pericytes in the renal vasculature: Roles in health and disease. _Nat. Rev. Nephrol._ 14, 521–534 (2018).

Article  CAS  PubMed  Google Scholar  * Vaughan, M. R. & Quaggin, S. E. How do mesangial and endothelial cells form the glomerular tuft. _J. Am. Soc. Nephrol._ 19, 24–33 (2008). Article

  PubMed  Google Scholar  * Vanslambrouck, J. M. et al. A toolbox to characterize human induced pluripotent stem cell-derived kidney cell types and organoids. _J. Am. Soc. Nephrol._

https://doi.org/10.1681/ASN.2019030303 (2019). * Huang, J. et al. A reference human genome dataset of the BGISEQ-500 sequencer. _Gigascience_ 6, 1–9 (2017). Article  PubMed  PubMed Central 

CAS  Google Scholar  * Korostin, D. et al. Comparative analysis of novel MGISEQ-2000 sequencing platform vs Illumina HiSeq 2500 for whole-genome sequencing. _PLoS One_ 15, e0230301 (2020).

Article  CAS  PubMed  PubMed Central  Google Scholar  * Taverna, F. et al. BIOMEX: An interactive workflow for (single cell) omics data interpretation and visualization. _Nucleic Acids Res._

48, W385–W394 (2020). Article  CAS  PubMed  PubMed Central  Google Scholar  * Suzuki, R. & Shimodaira, H. Pvclust: An R package for assessing the uncertainty in hierarchical clustering.

_Bioinformatics_ 22, 1540–1542 (2006). Article  CAS  PubMed  Google Scholar  * Deerinck, T. J., Bushong, E. A., Thor, A. & Ellisman, M. H. NCMIR Methods for 3D EM: A new protocol for

preparation of biological specimens for serial block face scanning electron microscopy. _NCMIR_ https://ncmir.ucsd.edu/sbem-protocol (2010). * Ronneberger, O., Fischer, P. & Brox, T.

_Medical Image Computing and Computer-Assisted Intervention (MICCAI)_ 9351 234–241 (Springer, 2015). * Izmailov, P., Podoprikhin, D., Garipov, T., Vetrov, D. & Wilson, A. G. Averaging

weights leads to wider optima and better generalization. _arXiv preprin_t https://doi.org/10.48550/arXiv.1803.05407 (2018). * Schindelin, J. et al. Fiji: An open-source platform for

biological-image analysis. _Nat. Methods_ 9, 676–682 (2012). Article  CAS  PubMed  Google Scholar  * Kremer, J. R., Mastronarde, D. N. & McIntosh, J. R. Computer visualization of

three-dimensional image data using IMOD. _J. Struct. Biol._ 116, 71–76 (1996). Article  CAS  PubMed  Google Scholar  Download references ACKNOWLEDGEMENTS We are grateful to Christian Freund

(hiPSC core facility, LUMC, Leiden, the Netherlands) for providing hiPSC lines (LUMC0072 and LUMC0020), and Melissa Little (Murdoch Children’s Research Institute, Melbourne, Australia) for

iPSC-MAFB. We acknowledge the support of Saskia van der Wal-Maas (Department of Anatomy & Embryology, LUMC, Leiden, the Netherlands), Conny van Munsteren (Department of Anatomy &

Embryology, LUMC, Leiden, the Netherlands), Manon Zuurmond (LUMC, Leiden, the Netherlands), George Galaris (LUMC, Leiden, the Netherlands), and Annemarie de Graaf (LUMC, Leiden, the

Netherlands). This work is supported by the partners of Regenerative Medicine Crossing Borders (RegMedXB) and Health Holland, Top Sector Life Sciences & Health. M.Koning is supported by

‘Nephrosearch Stichting tot steun van het wetenschappelijk onderzoek van de afdeling Nierziekten van het LUMC’. S.J. Dumas is supported by a Marie Skłodowska-Curie fellowship (grant

agreement No 846615) from the European Union’s Horizon 2020 research and innovation program. M. Borri is supported by the ‘Fonds voor Wetenschappelijk Onderzoek’ (FWO). L. Lin is supported

by the DFF Sapere Aude Starting grant (8048-00072A). Y. Luo is supported by the Danish Research Council (9041-00317B) and European Union’s Horizon 2020 research and innovation program under

grant agreement No 899417. H.S. Spijker is supported by a Kolff grant from the Dutch Kidney Foundation. P. Carmeliet is supported by Grants from Methusalem funding (Flemish government), the

Fund for Scientific Research-Flanders (FWO-Vlaanderen), the European Research Council ERC Advanced Research Grant EU- ERC74307, and NNF Laureate Research Grant from Novo Nordisk Foundation

(Denmark). C.W. van den Berg is supported by the Wiyadharma fellowship (Bontius stichting-LUMC). C.W. van den Berg and T.J. Rabelink are supported by The Novo Nordisk Foundation Center for

Stem Cell Medicine (reNEW), The Novo Nordisk Foundation Center for Stem Cell Medicine is supported by Novo Nordisk Foundation grants (NNF21CC0073729). AUTHOR INFORMATION Author notes * These

authors contributed equally: Marije Koning, Sébastien J. Dumas. * These authors jointly supervised this work: Cathelijne W. van den Berg, Ton J. Rabelink. AUTHORS AND AFFILIATIONS *

Department of Internal Medicine - Nephrology, Leiden University Medical Center, Leiden, the Netherlands Marije Koning, Ellen Lievers, Loes E. Wiersma, H. Siebe Spijker, Bernard M. van den

Berg, Cathelijne W. van den Berg & Ton J. Rabelink * Einthoven Laboratory of Vascular and Regenerative Medicine, Leiden University Medical Center, Leiden, the Netherlands Marije Koning, 

Ellen Lievers, Loes E. Wiersma, H. Siebe Spijker, Bernard M. van den Berg, Cathelijne W. van den Berg & Ton J. Rabelink * Laboratory of Angiogenesis and Vascular Metabolism, Department

of Oncology and Leuven Cancer Institute (LKI), KU Leuven, VIB Center for Cancer Biology, VIB, Leuven, 3000, Belgium Sébastien J. Dumas, Elda Meta, Mila Borri & Peter Carmeliet *

Department of Cell and Chemical Biology, Leiden University Medical Center, Leiden, the Netherlands M. Cristina Avramut & Roman I. Koning * Lars Bolund Institute of Regenerative Medicine,

Qingdao-Europe Advanced Institute for Life Sciences, BGI-Qingdao, BGI-Shenzhen, Qingdao, China Xue Liang & Yonglun Luo * MGI, BGI-Shenzhen, Shenzhen, China Lin Xie, Ping Liu & Fang

Chen * Department of Biomedicine, Aarhus University, Aarhus, Denmark Lin Lin & Yonglun Luo * Steno Diabetes Center Aarhus, Aarhus University Hospital, Aarhus, Denmark Lin Lin & 

Yonglun Luo * Department of Pediatrics, Leiden University Medical Center, Leiden, the Netherlands Jaap Mulder * Division of Nephrology, Department of Pediatrics, Erasmus Medical Center,

Rotterdam, the Netherlands Jaap Mulder * Sorbonne Université, IBPS, CNRS UMR7622, Inserm U 1156, Developmental Biology Laboratory, Paris, France Thierry Jaffredo * Laboratory of Angiogenesis

and Vascular Heterogeneity, Department of Biomedicine, Aarhus University, Aarhus, 8000, Denmark Peter Carmeliet * State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun

Yat-Sen University, Guangzhou, Guangdong, People’s Republic of China Peter Carmeliet * The Novo Nordisk Foundation Center for Stem Cell Medicine (reNEW), Leiden University Medical Center,

Leiden, the Netherlands Cathelijne W. van den Berg & Ton J. Rabelink Authors * Marije Koning View author publications You can also search for this author inPubMed Google Scholar *

Sébastien J. Dumas View author publications You can also search for this author inPubMed Google Scholar * M. Cristina Avramut View author publications You can also search for this author

inPubMed Google Scholar * Roman I. Koning View author publications You can also search for this author inPubMed Google Scholar * Elda Meta View author publications You can also search for

this author inPubMed Google Scholar * Ellen Lievers View author publications You can also search for this author inPubMed Google Scholar * Loes E. Wiersma View author publications You can

also search for this author inPubMed Google Scholar * Mila Borri View author publications You can also search for this author inPubMed Google Scholar * Xue Liang View author publications You

can also search for this author inPubMed Google Scholar * Lin Xie View author publications You can also search for this author inPubMed Google Scholar * Ping Liu View author publications

You can also search for this author inPubMed Google Scholar * Fang Chen View author publications You can also search for this author inPubMed Google Scholar * Lin Lin View author

publications You can also search for this author inPubMed Google Scholar * Yonglun Luo View author publications You can also search for this author inPubMed Google Scholar * Jaap Mulder View

author publications You can also search for this author inPubMed Google Scholar * H. Siebe Spijker View author publications You can also search for this author inPubMed Google Scholar *

Thierry Jaffredo View author publications You can also search for this author inPubMed Google Scholar * Bernard M. van den Berg View author publications You can also search for this author

inPubMed Google Scholar * Peter Carmeliet View author publications You can also search for this author inPubMed Google Scholar * Cathelijne W. van den Berg View author publications You can

also search for this author inPubMed Google Scholar * Ton J. Rabelink View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS M.K. designed and

conducted the experiments, analyzed the data and wrote the manuscript. S.J.D. contributed to sample preparation, designed the scRNAseq experiments, analyzed the scRNAseq data and wrote the

manuscript. M.K. and S.J.D. contributed equally to the manuscript. M.C.A. and R.I.K. processed, imaged and analyzed TEM and SBF-SEM samples and datasets. E.M. contributed to sample

preparation, designed the scRNAseq experiments and contributed to scRNAseq data analysis. E.L. and L.E.W. provided technical support. M.B. contributed to sample preparation. X.L. and L.L

pre-processed scRNAseq data. L.X., P.L., and F.C. performed sequencing of single cell libraries. Y.L. provided resource, supervised sequencing and pre-processing steps. J.M. and H.S.S.

contributed to the manuscript revision. T.J. contributed to establishing intracoelomic transplantation technique. B.M.B. contributed to data interpretation and manuscript revision. P.C.

provided resource, supervised S.J.D., E.M., and M.B., designed the scRNAseq experiments and contributed to scRNAseq data interpretation. C.W.B. and T.J.R. conceived the project, provided

resource, interpreted data, supervised M.K., revised the manuscript. CORRESPONDING AUTHOR Correspondence to Ton J. Rabelink. ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no

competing interests. ADDITIONAL INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

SUPPLEMENTARY INFORMATION SUPPLEMENTARY INFORMATION SUPPLEMENTARY MOVIE 1 SUPPLEMENTARY MOVIE 2 SUPPLEMENTARY MOVIE 3 SUPPLEMENTARY MOVIE 4 SUPPLEMENTARY MOVIE 5 SUPPLEMENTARY MOVIE 6

SUPPLEMENTARY DATA 1 SUPPLEMENTARY DATA 2 SUPPLEMENTARY DATA 3 SUPPLEMENTARY DATA 4 SUPPLEMENTARY DATA 5 SUPPLEMENTARY DATA 6 SUPPLEMENTARY DATA 7 SUPPLEMENTARY DATA 8 SUPPLEMENTARY DATA 9

SUPPLEMENTARY DATA 10 SUPPLEMENTARY DATA 11 RELATED MANUSCRIPT FILE REPORTING SUMMARY RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0

International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the

source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative

Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by

statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Koning, M., Dumas, S.J., Avramut, M.C. _et al._ Vasculogenesis in kidney organoids

upon transplantation. _npj Regen Med_ 7, 40 (2022). https://doi.org/10.1038/s41536-022-00237-4 Download citation * Received: 24 February 2022 * Accepted: 20 July 2022 * Published: 19 August

2022 * DOI: https://doi.org/10.1038/s41536-022-00237-4 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a shareable

link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative