Adaptable haemodynamic endothelial cells for organogenesis and tumorigenesis

Abstract

Endothelial cells adopt tissue-specific characteristics to instruct organ development and regeneration^1,2. This adaptability is lost in cultured adult endothelial cells, which do not vascularize tissues in an organotypic manner. Here, we show that transient reactivation of the embryonic-restricted ETS variant transcription factor 2 (ETV2)³ in mature human endothelial cells cultured in a serum-free three-dimensional matrix composed of a mixture of laminin, entactin and type-IV collagen (LEC matrix) 'resets' these endothelial cells to adaptable, vasculogenic cells, which form perfusable and plastic vascular plexi. Through chromatin remodelling, ETV2 induces tubulogenic pathways, including the activation of RAP1, which promotes the formation of durable lumens^4,5. In three-dimensional matrices-which do not have the constraints of bioprinted scaffolds-the 'reset' vascular endothelial cells (R-VECs) self-assemble into stable, multilayered and branching vascular networks within scalable microfluidic chambers, which are capable of transporting human blood. In vivo, R-VECs implanted subcutaneously in mice self-organize into durable pericyte-coated vessels that functionally anastomose to the host circulation and exhibit long-lasting patterning, with no evidence of malformations or angiomas. R-VECs directly interact with cells within three-dimensional co-cultured organoids, removing the need for the restrictive synthetic semipermeable membranes that are required for organ-on-chip systems, therefore providing a physiological platform for vascularization, which we call 'Organ-On-VascularNet'. R-VECs enable perfusion of glucose-responsive insulin-secreting human pancreatic islets, vascularize decellularized rat intestines and arborize healthy or cancerous human colon organoids. Using single-cell RNA sequencing and epigenetic profiling, we demonstrate that R-VECs establish an adaptive vascular niche that differentially adjusts and conforms to organoids and tumoroids in a tissue-specific manner. Our Organ-On-VascularNet model will permit metabolic, immunological and physiochemical studies and screens to decipher the crosstalk between organotypic endothelial cells and parenchymal cells for identification of determinants of endothelial cell heterogeneity, and could lead to advances in therapeutic organ repair and tumour targeting.

Fig. 1. R-VECs self-assemble into 3D durable vessels in vitro and in vivo.

a, Experimental set-up for vessel formation. A total of 10⁵ control ECs or R-VECs were plated on Matrigel in serum-free StemSpan tube-formation medium (Supplementary Data 2). Lenti-EVT2, lentiviral ETV2 expression construct. b, Z-stack of GFP⁺ R-VEC vessels at week 16. Scale bar, 1,000 μm. c, Quantification of tube formation in control ECs (HUVECs) and R-VECs (HUVEC-ETV2). d, Quantification of R-VEC vessels on Matrigel or LEC matrix. e, Electron microscopy images of stage-3 vessels on Matrigel and LEC matrix. L, lumen. Scale bars, 5 μm. f, Top, schematic of in vivo plug experiment in which control ECs or R-VECs fluorescently labelled with GFP were subcutaneously injected as a single-cell LEC suspension into SCID-beige mice. Bottom, whole-mount confocal images of R-VEC plugs and control EC plugs at five months. A fluorescently labelled antibody against human VEcadherin (hVEcad) was injected retro-orbitally before mice were euthanized. Scale bars, 200 μm. g, Orthogonal projection showing the anastomosis of mouse vessels and human VEcad⁺ vessels. Sections were post-stained for mouse endomucin (mEndomucin). Scale bar, 10 μm. h, Quantification of the density of human vessels in the plugs, defined as the percentage of GFP positive vessels of the scanned area. i, j, Experimental procedure for the decellularized intestine cultures (i, left). R-VECs repopulated the vasculature, lining blood vessels including the distal capillaries. At day 7 the bioreactors were stained for human CD31, imaged (i, right) and quantified (j). Scale bars, 500 μm. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. Source data

Fig. 2. Transcriptome and epigenetic analyses of R-VEC signatures.

a, Schematic of RNA-seq and ChIP–seq performed in the induction phase (day 14) on R-VECs and control ECs. b, RNA-seq of R-VECs or control HUVECs in stage 1 (2D monolayers). GO term analysis was performed on differentially expressed genes. GO categories are ordered on the basis of the number of differentially expressed genes. Heat maps for GO categories in red are presented in Fig. 2c and Extended Fig. 6b. ECM, extracellular matrix; Pos. reg., positive regulation. c, Heat map of genes in one top GO category. Values are log₂-normalized counts per million (CPM), centred and scaled by row. ETV2 binding from ChIP–seq at the promoter of each differentially expressed gene is shown in the yellow-and-green heat map (left). d, Heat map of 490 differentially expressed genes across ECs of different tissues (stage 1, induction phase) upon ETV2 expression. Tissue-adjusted log₂-transformed CPM, centred and scaled by row. e, ETV2 ChIP–seq in R-VECs during the induction phase (stage 1; 2D) using an anti-Flag antibody or mouse IgG as control. ChIP for H3K4me3, H3K27ac and H3K27me3 was performed in both control ECs and R-VECs at stage 1. Enriched regions were analysed by ChIP–seq. Horizontal bars underneath peaks represent significantly changed regions. Promoter regions bound by ETV2 are highlighted in cream. Track range ETV2/K27me3/K27ac, 0–0.3; K4me3/input/IgG, 0–1. f, Western blot for active RAP1–GTP compared to total RAP1 input for stage 1 2D control ECs (HUVECs) and R-VECs (HUVEC-ETV2). The quantification of RAP1–GTP compared to total RAP1 is shown below the blot and presented as mean ± s.e.m. g, Quantification of R-VEC vessel formation after treatment with RAP1 inhibitor or dimethyl sulfoxide (DMSO). h, Z-stack confocal images and electron microscopy images of R-VEC vessels treated with RAP1 inhibitor or DMSO at four weeks. Red circles indicate orthogonal cross-sections. Scale bars, 5 μm (top); 2 μm (bottom). Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01. For statistics, see Supplementary Data 1. Source data

Fig. 3. R-VECs haemodynamically and physiologically vascularize human islets.

a, Overview of microfluidic device measuring 5 × 3 × 1 mm and holding 15 μl fibrin gel. b, Representative images of devices with control ECs or R-VECs stained with human VEcad antibody at day 7. Scale bars, 3 mm. c, Orthogonal representation of intact lumen formation in R-VECs. Scale bar, 50 μm. d, Quantification of vessel area in devices with control ECs versus R-VECs. e, Intact heparinized human peripheral blood (100 μl) composed of a full complement of red blood cells, white blood cells, platelets and unperturbed plasma was injected and perfused through the R-VEC vessels. Right, representative live image of blood flow through R-VECs (see also Supplementary Video 1b, c). Scale bar, 25 μm. f, Experimental set-up for co-seeding human islets with control ECs or R-VECs in microfluidic devices. g, Fluorescently labelled human heparinized whole blood (red, PKH26 red fluorescent dye) was perfused through the microfluidic devices (day 4) (see also Supplementary Video 2b–d). Z-stack projections of whole devices of islet explants post-stained with EpCAM and VEcad (day 4). Scale bars, 3 mm. h, Magnified area of direct interaction of R-VECs with co-cultured islets in a microfluidic device. Scale bar, 100 μm. i, Single section and orthogonal projection of human islets vascularized by R-VECs in a microfluidic device. Scale bars, 50 μm. j, Experimental set-up for the glucose-stimulation test in microfluidic devices. k, Insulin levels were measured at 2 mM glucose (t = −10 and 0 min, basal level) and 9 and 24 min after stimulation with 16.7 mM glucose. * represents statistical tests versus islets alone; # represents statistical tests versus islets + control ECs. l, Fold change in insulin levels at the outlet (insulin levels at 16.7 mM/insulin levels at 2 mM), 9 min after high-glucose stimulation. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001, ###P < 0.001. For statistics, see Supplementary Data 1. Source data

Fig. 4. R-VECs arborize and conform to normal and tumour organoids.

a, Control ECs or R-VECs were seeded with human normal colon organoids (COs) or colorectal cancer organoids (CRCOs) in Matrigel droplets. b, c, Confocal Z-projections (b) and magnified images (c) of human COs alone or co-cultured with control ECs or R-VECs on day 8. Scale bars, 1 mm (b); 100 μm (c, left image); 50 μm (c, right images). d, Surface area of control ECs or R-VECs interacting with human COs in LEC matrix, quantified over a 72-h time lapse. e, Quantification of colon area (as stained by EpCAM) per field in COs alone or co-cultured with control ECs or R-VECs on day 8. f, Human CRCOs were seeded with control ECs or R-VECs in Matrigel droplets. Confocal images of CRCOs alone or co-cultured with control ECs or R-VECs, post-stained for KRT20 at day 8 after a 4.5-h EdU pulse. Scale bars, 100 μm. g, Surface area of control ECs or R-VECs interacting with human CRCOs in Matrigel, quantified over a 78-h time lapse. h, Quantification of EdU⁺ CRCO cells (in f) in CRCOs alone or CRCOs co-cultured with control ECs or R-VECs on day 8. i, j, R-VECs in single-cell suspension were cultured either alone or mixed with human COs and subjected to scRNA-seq on day 7. i, Uniform manifold approximation and projection (UMAP) plot for the EC fractions of R-VECs alone and R-VECs co-cultured with COs. j, UMAP plot of 9 unique clusters, in ECs from both the R-VECs alone group and the R-VECs +COs group. k, Heat map of differentially expressed genes from cluster 5, enriched among R-VECs but absent in R-VEC-alone cultures. l, m, R-VECs in single-cell suspension were cultured either alone or mixed with human CRCOs and subjected to scRNA sequencing on day 7. l, UMAP plot for the EC fractions of R-VECs alone and R-VECs co-cultured with CRCOs. m, UMAP plot of 8 unique clusters, in ECs from both the R-VECs alone group and the R-VECs + CRCOs group. n, Heat map of differentially expressed genes from cluster 8, enriched among R-VECs in culture with CRCOs. o, Schematic of the adaptation and maladaptation of R-VECs. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. Source data

Extended Data Fig. 1. ETV2 confers mature human ECs with the ability to autonomously self-assemble into lumenized, durable, branching and patterned vessels in vitro without the constraints of bioprinted scaffolds.

a, Overview of experimental set-up for vessel formation in vitro for screen of different media, extracellular matrix components and tissue-specific ECs. b, The proliferation of GFP-transduced R-VECs and control ECs (CTRL-ECs) at each stage of vessel formation. EdU⁺ cells were quantified after a 16-hour EdU pulse. c, Time course of vessel formation on Matrigel for GFP⁺ CTRL-EC and R-VECs over 8 weeks. d, Vessel formation using R-VEC or CTRL-EC in three different enriched pro-angiogenic media (Supplementary Data 2): Serum-free StemSpan with Knockout serum replacement and Cytokines, EGM-2 and complete EC media on Matrigel. R-VEC formed the most robust lumenized vessels in serum-free StemSpan with knockout serum replacement medium and cytokines, as compared to other media with serum. CTRL-EC failed to form durable stable vessels. e, f, Time course (e) and quantification (f) of tube formation for GFP⁺ human Adipose CTRL-EC and human Adipose R-VEC on Matrigel. g, Representative images of tissue-specific GFP⁺ R-VEC and CTRL-EC isolated from adult human heart (cardiac EC), aorta (aortic EC) and skin (dermal EC) demonstrated robust and stable vessels at 4 weeks on Matrigel. h, Representative images of GFP⁺ R-VEC vessels formed on Matrigel or a pre-defined matrix of laminin/entactin and collagen IV (LEC). i, Immunostaining of R-VEC-tubes displayed apicobasal proper polarity with podocalyxin, apical (in red) and laminin, basal (in green). The right image is an orthogonal projection. j, Stiffness measurements by atomic force microscopy (AFM) of adult Adipose and HUVEC ECs with and without ETV2. In both cases, ETV2-transduced ECs are significantly less stiff than their counterparts. The abbreviated box plots indicate the interquartile range and median for each condition. k, HUVECs were transduced with either an empty lentiviral vector or lentiviral vectors with ETV2, myrAKT or ETS1 constructs, and used in a vessel formation assay. Western Blot analysis for expression of ETV2, p-AKT, total AKT and ETS1 in those cells. l, Representative images for ETS1 or myrAKT1 transduced GFP⁺ HUVECs in a vessel formation assay on Matrigel. m, Quantification of vessel area for ETS1, myrAKT1 and ETV2 (R-VEC) cells indicated that ETS1-EC and myrAKT1-EC fail to form robust vessel formation as compared to R-VEC. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2. Source data

Extended Data Fig. 2. Transient ETV2 expression in adult human ECs is sufficient for the generation and maintenance of durable long-lasting R-VEC vessels in vitro.

a, Schematic for ETV2 mRNA and protein levels assessment at each of the three stages of R-VEC vessel formation. b, Quantification of ETV2 mRNA levels at each stage of vessel formation. c, d, Western blot analysis (c) and densitometric quantification (d) of ETV2 protein levels at each stage of vessel formation. GAPDH was used as a loading control. e, A proteasome inhibitor (MG132) restored ETV2 levels by ~sixfold when added to R-VECs during the stabilization stage. f, Densitometric quantification of western blots in e. g, qRT–PCR (g) and western blot (h) assessment of ETV2 levels after doxycycline removal. i, Representative images of GFP⁺ iR-VECs on Matrigel with inducible ETV2 expression at 2 months. ETV2 was turned off at day 0, day 7 and at 4-weeks post start of the remodelling stage 2. j, Quantification of iR-VEC vessels at 2 months. k, Electron microscopy pictures of a lumen present both in vessels in which doxycycline was continuously present and in vessels in which doxycycline was removed after 1 month. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2. Source data

Extended Data Fig. 3. R-VEC vessels are functionally anastomosed to host vessels and not leaky in vivo.

a, Fluorescently labelled R-VEC or CTRL-EC cells in LEC matrix were subcutaneously injected in the flank of SCID beige mice and retrieved at 2 months. Human-specific VEcad antibody (hVEcad) was injected intravitally right before euthanasia. Sections of the plugs were stained for mouse ECs with an anti-mouse endomucin antibody (mEndomucin), identifying properly organized human R-VECs anastomosing with mouse vessels (thickness = 50 μm). Sections were also stained with the nuclear stain DAPI. b, c, Plugs in a were post-stained with hVEcad and a mouse PDGFRβ antibody (b) or mouse SMA antibody (c) (thickness = 50 μm). d, In vivo plug assay, in which mice were subcutaneously injected with either control ECs (HUVECs transduced only with rtTA lentivirus) or stage 1 doxycycline-inducible-ETV2 ECs (iR-VECs: HUVECs transduced with both rtTA and inducible ETV2 lentivirus) in LEC matrix. One group of mice was on doxycycline (ETV2 continuously on) and another group of mice was on doxycycline food diet for 1 week (ETV2 on) and then switched to regular food (ETV2 off). All mice were euthanized 2 months post-implantation. Red indicates the GFP labelled human ECs, white: Anti-VEcad antibody that was retro-orbitally injected before euthanizing the mice. e, Quantification of vessel area for rtTA only plugs, mice on doxycycline for 1 week, and mice continuously on doxycycline diet (ETV2 on). All mice were euthanized 2 months post-implantation. f, 70 kDa fluorescent dextran (in blue) and human VEcad (in white) were injected in mice implanted with fluorescently labelled R-VECs (in red, 5-months post-implantation), iR-VECs (in red, 1 week on doxycycline food and euthanized at 2 months) or K-RAS-HUVECs (K-RAS-EC) (in red, 2-weeks post-implantation) to assess anastomosis and leakiness of vessels. K-RAS-EC vessels showed dextran leakiness, whereas R-VEC and iR-VEC vessels exhibited patency and non-leakiness. Green arrows point at perfused mouse vessels that were also perfused with dextran. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2. Source data

Extended Data Fig. 4. Implanted R-VECs form stable, patterned, branching and durable vessels in vivo without features of vascular malformations, cysts, adenomas, haemangiomas or metastasis.

a, Representative images of non-haemorrhagic R-VEC plugs at 10 months. b, Whole-mount microscopy of R-VEC plugs at 10-months post perfusion with anti-human VEcad antibody (hVEcad). c, d, Representative H&E and Masson staining of R-VEC plugs at 10 months (c). There were no features of cysts or haemangiomas present, in contrast to KRAS-EC plugs (at 4 weeks) that formed an EC tumour (d). e, There was no metastasis of R-VECs to other tissues 10 months after plug implantation and the tissues were assessed to be normal without fibrosis and architectural disruption or tumorigenesis as evaluated by H&E, Masson and picrosirius staining.

Extended Data Fig. 5. Decellularized intestinal scaffolds re-endothelialized with R-VECs engraft in vivo after omental implantation.

a, Schematic of experimental procedure for heterotopic implantation of decellularized intestinal scaffold vascularized using R-VECs. b, Rat intestines were cannulated through lumen, mesenteric artery and mesenteric vein. c, Decellularized intestine preserves native vasculature (green = GFP⁺ R-VECs). d, Seeded GFP labelled R-VECs spread evenly and reach distal capillaries. e, Heterotopic implantation of re-endothelialized intestines in immunodeficient mice omentum shows engraftment after 1 and 4 weeks of GFP⁺ R-VECs and anastomosis to the host vasculature as indicated by intravital intravenous injection of anti-human VEcad antibody (hVEcad). Representative H&E stainings show anatomical normal perfused vessels. f, Quantification of the area covered by R-VEC compared to CTRL-EC in implanted re-endothelialized intestines at 1 week and 4 weeks. g, Quantification of R-VEC and CTRL-EC proliferation and apoptosis in implanted re-endothelialized intestines at 1 and 4 weeks. Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. Source data

Extended Data Fig. 6. ETV2, by directly binding to promoters and enhancers of target genes, regulates differentially expressed genes in R-VECs.

a, Schematic of RNA-sequencing performed on R-VECs and CTRL-ECs derived from different tissue-specific ECs during stage 1 induction phase (2D monolayers). b, R-VECs or CRTL-ECs were analysed by RNA sequencing. Heat maps of selected genes within top enriched GO categories. Values are log₂-normalized CPM, centred and scaled by row. ETV2 binding from ChIP–seq at the promoter of each differentially expressed gene is shown in the yellow-green heat map. c, R-VECs retain essential EC fate genes at stage 1 induction phase across all tissue-specific ECs. The data are presented as log₂(CPM) with no scaling by row or column. d, PCA plot based on the top 1,000 most variable genes across ECs with and without ETV2 from different tissues during stage 1 induction, using log₂-normalized CPM after subtracting tissue-specific effects using limma’s removeBatchEffect function. e, ETV2 ChIP was performed on R-VECs using an anti-flag antibody at the induction stage 1 (2D), along with histone modification ChIP for H3K4me3 and H3K27ac. Enriched regions were analysed by ChIP–seq. f, Genomic distribution of ETV2 peaks in R-VEC (Stage 1). The number in brackets is the number of ETV2 peaks in each region. g, Promoters bound by ETV2 have an increase in both K4me3 and K27ac. h, GO enrichment in genes with ETV2 binding at promoters. For statistics, see Supplementary Data 1.

Extended Data Fig. 7. ETV2 in R-VECs endows ECs with transcriptional adaptability and plasticity.

a, Diagram of EC sample preparation from ETV2 Venus reporter mice by FACS sorting. ETV2⁺ and ETV2⁻ ECs were sorted at day E9.5. ECs were sorted as non-haematopoietic CD31⁺CD45^neg cells. b, Heat map of overlap of differentially expressed genes in ETV2⁺ vs. ETV2⁻ ECs at E9.5 and R-VECs (stage 1) vs. CTRL-EC from different tissues, using tissue-adjusted log₂(CPM), centred and scaled by row. c, Knockdown of RASGRP3 by two different shRNAs in R-VECs, shRNA against Luciferase was used as control. Vessel quantification upon RASGRP3 knockdown. d, Heat map displaying overlapping differentially expressed genes from R-VEC at stabilization stage 3 (4 weeks) vs. R-VEC at induction stage 1, R-VECs in vitro pre-plug (stage 1 induction stage) vs. R-VECs in vivo in plugs (1 month), and freshly isolated vs. cultured HUVECs. Values represent tissue-adjusted log₂(CPM), centred and scaled by row. e, ChIP–seq depicting genes that are differentially expressed in the stabilization stage 3 phase, but that are already directly bound by ETV2 and epigenetically primed for expression at induction stage 1 (2D monolayers). ETV2 ChIP–sequencing was performed on R-VECs using an anti-flag antibody. Mouse IgG was used as a control for ETV2 ChIP. Histone modification ChIP for H3K4me3, H3K27ac and H3K27me3 was performed on both CTRL-EC and R-VEC at the induction stage 1 (2D monolayers). Enriched regions were analysed by ChIP–sequencing. Black bar, ETV2 enriched regions. Green bar, the region with increased K4me3 modification. Blue bar, the region with increased K27ac modification. Promoter regions bound by ETV2 are highlighted in cream. Track range ETV2/K27me3/K27ac/, 0–0.3; K4me3/input/IgG, 0–1. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2. Source data

Extended Data Fig. 8. R-VECs physiologically arborize human pancreatic islet explants and organoids.

a, Human islet explants were cultured in Matrigel droplets (volume 50 μl) either with GFP labelled CTRL-EC or R-VEC (day 4). b, Insulin secretion fold change after glucose stimulation at 16.7 mM vs. 2mM glucose (2-week time point). c, Vessel area of ECs directly interacting with islets at week 2. d, EpCAM and VEcad staining of islets co-cultured in a Matrigel droplets at 2 weeks. e, Orthogonal projections of R-VECs in co-culture with human islets at two weeks, demonstrating strong interaction of the sprouting R-VEC vessels with islets. f, Human COs were derived from isolated crypts from colon biopsies of healthy human donors. Colon organoids were confirmed to express proper markers by quantitative RT–PCR. g, Quantitative RT–PCR of various colon markers for human COs, co-cultured with CTRL-EC or co-cultured with R-VEC for 8 days. Epithelial cells were sorted out as live CD31^neg non-vascular cells. h, Mouse small intestine organoids were cultured alone, or in the presence of CTRL-EC  or R-VEC (day 8). Confocal representative images of EdU⁺ (proliferating cells), KRT20⁺ (differentiated epithelial cells in blue) and ECs (mCherry - red) of co-culture experiment with mouse intestinal organoids. i, Quantification of vessel area over the course of 7 days in co-cultures of mouse intestine organoids with CTRL-EC or R-VEC. j, Vessel arborization quantified as EC sprouts in direct contact/organoid in CTRL-EC versus R-VEC wells. k, Time-lapse representative images show the progression of interacting ECs with CRCOs. CTRL-EC (in green) did not interact with CRCOs (in red) (top panel), whereas R-VEC (in green) form robust EC tubes to tap and wrap CRCOs (in red) (bottom panel). l, Orthogonal projections of CRCOs co-cultured with R-VECs (day 8). Data are mean ± s.e.m. NS, not significant; *P < 0.05, **P < 0.01, ***P < 0.001. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2. Source data

Extended Data Fig. 9. Endothelial and epithelial cell identification by scRNA-seq from co-cultures of normal COs with R-VECs.

a, Schematic of 10X Chromium scRNA-seq experiments of R-VECs alone, R-VECs co-cultured with human COs, or COs alone. Samples were analysed 7 days post co-culture. The same compatible medium was used across all three conditions. b, UMAP of cells from each condition alone and the three conditions merged. c, Endothelial cells were identified as cells expressing either VEcad, CD31 or VEGFR2 and negative for the epithelial marker EPCAM. Epithelial cells were defined as positive for EPCAM and negative for any of the EC markers VEcad, CD31 or VEGFR2. d, UMAP of the 9 unique clusters identified in the merged samples. e, Endothelial and epithelial cell specific markers were used to confirm the EC clusters (clusters 1 to 7) vs. epithelial cell clusters (clusters 8 and 9). f, The identity of epithelial cells in clusters 8 and 9 was confirmed as colon-specific by expression of marker genes including SATB2, CA4, CA2 and others. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2.

Extended Data Fig. 10. Endothelial and epithelial cell identification by scRNA-seq from co-cultures of CRCOs with R-VECs.

a, Schematic of 10X Chromium scRNA-seq experiments of R-VECs alone, R-VECs co-cultured with human CRCOs or CRCOs alone. Samples were analysed 7 days after co-culture. The same compatible medium was used across all three conditions. b, UMAP of cells from each condition alone and the three conditions merged. c, Endothelial cells were identified as cells expressing either VEcad, CD31 or VEGFR2 and negative for the epithelial marker EPCAM. Epithelial cells were defined as positive for EPCAM and negative for any EC markers VEcad, CD31 or VEGFR2. d, UMAP of the 9 unique clusters identified in the merged samples. e, Endothelial and epithelial cell-specific markers were used to confirm the endothelial cell clusters (clusters 6, 7, 8) vs. epithelial cell clusters (clusters 1, 2, 3, 4, 5, 9). f, UMAP of merged epithelial cell fractions from hCRCO cultured alone or co-cultured with R-VECs. Six unique clusters were identified. g, h, Heat map (g) and dot plot (h) of differentially expressed genes in tumour epithelial cells in cluster 2 and cluster 5 that are enriched in co-culture with R-VECs. Differential expression was performed using the Wilcoxon rank-sum test; FDR-adjusted P < 0.05. For statistics, see Supplementary Data 1. For medium formulations, see Supplementary Data 2.