Deriving Human Intestinal Organoids with Functional Tissue-Resident Macrophages All From Pluripotent Stem Cells

Abstract

Background & aims: Organs of the gastrointestinal tract contain tissue-resident immune cells that function during tissue development, homeostasis, and disease. However, most published human organoid model systems lack resident immune cells, thus limiting their potential as disease avatars. For example, human intestinal organoids (HIOs) derived from pluripotent stem cells contain epithelial and various mesenchymal cell types but lack immune cells. In this study, we aimed to develop an HIO model with functional tissue-resident macrophages.

Methods: HIOs and macrophages were generated separately through the directed differentiation of human pluripotent stem cells and combined in vitro. Following 2 weeks of coculture, the organoids were used for transcriptional profiling, functional analysis of macrophages, or transplanted into immunocompromised mice and matured in vivo for an additional 10-12 weeks.

Results: Macrophages were incorporated into developing HIOs and persisted for 2 weeks in vitro HIOs and for at least 12 weeks in HIOs in vivo. These cocultured macrophages had a transcriptional signature that resembled those in the human fetal intestine, indicating that they were acquiring the features of tissue-resident macrophages. HIO macrophages could phagocytose bacteria and produced inflammatory cytokines in response to proinflammatory signals, such as lipopolysaccharide, which could be reversed with interleukin-10.

Conclusions: We generated an HIO system containing functional tissue-resident macrophages for an extended period. This new organoid system can be used to investigate the molecular mechanisms involved in inflammatory bowel disease.

Keywords: Human Intestinal Organoids; Immune Cells; Inflammatory Bowel Disease; Tissue-Resident Macrophages.

Figure 1

HCO-derived erythromyeloid progenitors differentiated into macrophages and persisted in HIOs. (A) Schematic of recombination of HIOs with HCO-derived supernatant containing EMPs. (B) Representative gross images of HIOs and HCOs at Day 21 and 35. Endothelial tubes and EMPs/myeloid cells were observed in Day 21 HCOs. (C) Whole mount stained for macrophages CD163 (red), endothelial and EMPs CD34 (green), and epithelial CDH1 (white) in Day 35 HIOs, HCOs, and HIOs cocultured with HCO’s EMPs (HIOs + HCO’s EMPs) for 2 weeks. Higher magnification image shown of boxed area. Scale bars: 200 μm. (D) Relative expression of monocyte/macrophage markers CD163 and CD14, and patterning marker PDX1 in Day 35 HCOs and HIOs with and without added HCO’s EMPs. Statistical significance was set at ∗P < .05 and ∗∗P < .01, determined using 1-way analysis of variance (n = 4). ActA, Activin A; BMP2, bone morphogenic protein 2; EC, endothelial tubes; EGF, epidermal growth factor; FGF4, fibroblast growth factor 4; MHG Sph, mid/hind gut spheroids.

Figure 2

Direct differentiation of monocytes/macrophages from hPSCs. (A, B) Schematic and representative images of the directed differentiation of hPSCs into HPCs followed by subsequent differentiation into monocytes/macrophages (mono/macro) using M-CSF and GM-CSF. (C) Representative flow cytometric analysis of hPSC-derived myeloid cells stained with viability 7AAD, panhematopoietic CD45, mono/macro CD14 and CD11c, and macrophage CD163. No stain controls show gating strategy. (D, E) Schematic and gross images of Day 21 mono/macro embedded in Matrigel and cultured in HIO media with and without supplemented M-CSF and GM-CSF. (F) Representative flow cytometry of Day 35 mono/macro with and without supplemented M-CSF and GM-CSF. Adv. DMEM, Advanced DMEM; CSF, colony-stimulating factor; EGF, epidermal growth factor; L-Glut, L-glutamine; mono/macro, monocyte and macrophage; P/S, penicillin /streptomycin.

Figure 3

Generation of HIOs containing hPSC-derived macrophages. (A) Schematic of protocol for mixing Day 28 hPSC-derived macrophages (Macro) with Day 21 HIOs during followed by 2 weeks of coculture in Matrigel and standard HIO media (no M-CSF). (B) Representative whole-mount immunofluorescence of Day 35 HIOs and HIO + Macro stained for macrophages CD163 (red), endothelial CD34 (green), and epithelial CDH1 (white). Macro were added to HIOs either by mixing within Matrigel during plating (center) or in media/supernatant overlaying the Matrigel (right). Higher magnification image shown of boxed area. Scale bars: 200 μm for low and 50 μm for high magnifications. (C) Relative expression of mono/macro markers CD163, CD14, ITGAX/CD11c and cytokines IL6, IL10, and TNFα in Day 35 HIO and HIO + Macro. (D) Representative flow cytometric analysis and gating of viable (7AAD-), mono/macro (CD45+ CD14+), macro (CD163+ CD11c+) cells in Day 35 HIOs and HIO + Macro. (E) Quantification of percent of viable CD14+ CD45+ gated cells in Day 35 HIO and HIO + Macro. Statistical significance was set at ∗∗P < .01 and ∗∗∗P < .001, according to t test (n = 4). (F) UMAP visualization of scRNA-sequencing of Day 35 HIOs and HIO + Macro from 2 independent differentiations (N = 2). (G) Seurat clustering UMAP colored by annotated major cell populations. (H) Cell proportions of major cell types present in HIO and HIO + Macro compared with published proportions in 2 human gastrointestinal tract scRNA-seq atlases. Number of cells of each condition listed above the bar. (I) Select feature plots showing expression of markers for epithelial (CDH1), mesenchymal (EMILIN1, CSF1), endothelial (CDH5), and mono/macro (PTPRC/CD45, CSF1R) cell populations. (J) UMAP of reclustered monocyte/macrophage cluster 7 from Day 35 HIO and HIO + Macro datasets. Proportion of cells and relative expression of (K) commonly used monocyte and macrophage markers and cytokines and (L) macrophage subtype-enriched markers. CSF, colony-stimulating factor; EGF, epidermal growth factor; MHG Agg, mid/hind gut aggregates; UMAP, Uniform Manifold Approximation and Projection.

Figure 4

Macrophages had minimal effect on identity and proportion of epithelial and mesenchymal cells in HIOs. UMAP colored by (A) condition/replicate and (B) Seurat cluster of reclustered epithelial cells from Day 35 HIO and HIO + Macro from 2 independent differentiations (N = 2). (C) Cell proportion of Seurat clusters by HIO and HIO + Macro epithelium between the 2 replicates. (D) Proportion of cells and relative expression of epithelial markers enriched in each Seurat cluster. Number of cells of each condition listed above the bar. (E) Select feature plots showing expression of epithelial markers for intestine (CDX2, CDH17, HNF4A, PDX1, SATB2), gastric (SOX2, CLDN18, PDX1), and urothelial/squamous (SOX2, GATA3, KRT13, SATB2). (F) Proportion of cells and relative expression of epithelial markers enriched in each replicate and/or condition. UMAP colored by (G) condition/replicate and (H) Seurat cluster of reclustered mesenchymal and endothelial cells from Day 35 HIO and HIO + Macro. (I) Cell proportion of Seurat clusters by HIO and HIO + Macro mesenchyme between the 2 replicates. (J) Proportion of cells and relative expression of mesenchymal markers enriched in each Seurat cluster. (K) Select feature plots showing expression of mesenchymal markers for subepithelial fibroblasts (PDGFRA, PTCH1, FOXF1), matrix fibroblast/stromal cell (PDGFRA, COL3A1), smooth muscle cell precursor (ACTA2, FOXF1, PTCH1), and intestinal-enriched mesenchyme (NKX2-3). (L) Proportion of cells and relative expression of mesenchymal markers enriched in each replicate and/or condition. ECM, extracellular matrix; SMC, smooth muscle cell; UMAP, Uniform Manifold Approximation and Projection.

Figure 5

Macrophages cocultured with HIOs have a distinct transcriptional signature than HIO-naive macrophages. (A) Schematic of protocol for mixing Day 28 hPSC-derived macrophages (Day 28 Macro) with Day 21 HIOs (HIO + Macro) followed by 2 weeks of coculture in Matrigel standard HIO media (no M-CSF). As a control, Day 28 Macro were cultured for an additional 2 weeks in M-CSF and GM-CSF in the absence of HIO (Day 42 Macro). Macrophages that codevelop with hPSC-derived HCOs were also used as a control. UMAP colored by (B) condition or (C) annotated Seurat clustering in monocyte/macrophages populations bioinformatically reclustered from Day 28 and 42 Macro, Day 35 HIO + Macro, and HCOs (N=1-2). (D) Cell proportions represented in each Seurat cluster by condition. Number of cells of each condition listed above the bar. (E) Select feature plots showing expression of markers for panmyeloid (SPI1/PU.1), monocyte/macrophage (CD68, CSF1R, MRC1/CD206, ITGAX/CD11c, CD14), MHC-II (HLA-DRA, CD14), and developmentally derived tissue resident macrophage (CD163, FOLR2, LYVE1). (F) Proportion of cells and relative expression of myeloid/monocyte/macrophage markers enriched in each Seurat cluster. (G) Proportion of cells and relative expression of myeloid/monocyte/macrophage markers enriched in Day 28 Macro, Day 42 Macro, and HIO + Macro. Genes enriched in HCOs bolded. CSF, colony-stimulating factor; DC, dendritic cell precursor; EGF, epidermal growth factor; UMAP, Uniform Manifold Approximation and Projection.

Figure 6

HIO cocultured monocytes/macrophages are most similar to LYVE1 macrophages of the human fetal intestine. (A) Schematic of protocol for generating and comparing mono/macro from Day 28 Macro, Day 48 Macro, HIO + Macro, and HCOs with myeloid/mono/macro from human fetal, healthy pediatric, pediatric with Crohn’s disease, and healthy adult endodermal organ and intestinal scRNA-seq atlases. (B) UMAP of integrated myeloid cells from hPSC-derived mono/macro (left), Yu et al fetal endodermal atlas (middle), and Elmentaite et al fetal, pediatric, adult intestinal atlas (right). UMAP colored by (C) in vivo age groups or (D) annotated Seurat clustering. (E) Cell proportions represented in each Seurat cluster by condition/region/age. Number of cells of each condition listed above the bar. The “Small Intest.” and “Prox. Small Intest.” in the Yu et al. atlas are from Day 47 and 59 fetal intestine, respectively, whereas other samples are from Day 72–145 fetal tissue. MapQuery projection of hPSC-derived mono/macro onto (F) cell-type annotations or (G) age of Elmentaite et al myeloid reference atlas. (H, I) Ensemble (random forests and support vector machines) similarity score of the expression pattern of macrophages from Day 35 HIO + Macro, HCOs, Day 28 and Day 42 Macro compared with Elmentaite et al fetal, healthy pediatric, pediatric with Crohn’s disease, and healthy adult myeloid atlases as training reference dataset. hPSC-derived mono/macro were compared with (H) age and diagnosis or (I) region and age. (J) MapQuery projection of hPSC-derived mono/macro onto tissue Yu et al fetal endodermal mono/macro reference atlas. (K) Random forest similarity score of the expression pattern of macrophages from Day 35 HIO + Macro, HCOs, Day 28 and Day 42 Macro, and HCOs compared with Yu et al fetal mono/macro atlas training reference datasets split by annotated tissues. CD, Crohn’s disease; cDC, conventional dendritic cell; LN, lymph node; Mono, monocyte; pDC, plasmacytoid dendritic cell; UMAP, Uniform Manifold Approximation and Projection.

Figure 7

Macrophages within HIOs respond to bacterial factors, which could be reversed by IL10. (A) Schematic of functional assay in Day 35 HIO + Macro to phagocytose bacteria and response to inflammatory stimuli LPS alone or LPS and IL10 for 24 hours. (B) Time lapse images of macrophages (generated from GFP-hPSCs) from Day 35 HIO + Macro cultured with/without pHrodo-stained Escherichia coli particles (red). Higher magnification of box region shown. Scale bars: 40 μm and 10 μm. (C) Luminex assay of cytokines and chemokines of supernatant of Day 35 HIO and HIO + Macro (n = 4). (D) Luminex assay of cytokines and chemokines of supernatant of Day 35 HIO + Macro treated with 100 ng/mL LPS or PBS vehicle (Veh) for 24 hours. (E) Expression of cytokines and chemokines from Day 35 HIO + Macro treated with 100 ng/mL LPS or PBS vehicle for 24 hours. (F) Luminex assay of relative secretion of IL6, IL8, CCL3, and CCL4 of Day 35 HIO + Macro with and without 100 ng/mL LPS and with and without LPS and 50 ng/mL IL10 treatment. Statistical significance was set to ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001 according to (E) t test or (C, D, F) 1-way analysis of variance (n = 4). Veh, vehicle/PBS.

Figure 8

Macrophages added to HIOs remained in organoid following transplantation. (A) Schematic of TXP HIOs into an immunocompromised mouse kidney capsule. (B) Representative image of gross HIO and HIO + Macro at Day 35 and following 12 weeks transplanted in vivo. Scale bars: 1 mm. (C) Immunofluorescence staining of human macrophage antibody CD163 (red), mouse-specific macrophage marker F4/80 (green), and nuclei DAPI (blue) from HIO TXP and HIO + Macro TXP. Inset shows higher magnification. (D, E) Representative images and quantification of CD163 (red) macrophages as percentage of total DAPI (blue) cells from HIO + Macro from Day 35 in vitro or following 12-week TXP. Background epithelial CD163 cells were excluded from quantification (n = 3). HIO + Macro TXP with GFP-labeled macrophages (green) stained for (F) CD163 (red, left) or (G) proliferation marker Ki67 (red) and DAPI (blue). Significance: ∗P < .05, determined using t tests. Scale bars: 200 μm and 50 μm. TXP, transplantation.