Fig 1.
Transcriptomic analysis of mouse EECs along the entire GI tract.
(A) t-distributed stochastic neighbour embedding (tSNE) plots of single cell RNA-sequencing (scRNA-seq) data of NeuroD1-positive cells from mouse (top to bottom) stomach, upper small intestine (USI), lower SI (LSI), caecum and large intestine (LI). Clusters labelled by standard enteroendocrine cell (EEC) nomenclature defined by expression of known gene markers. Non-EECs are shown in grey. (B) Heatmap of expression for top 10 differentially-expressed genes per EEC cluster, calculated per GI region. Expression is log2(mean counts per million + 1). Differentially-expressed genes defined as padj < 0.05, genes filtered for log2(fold change) > 1.5. (C) tSNE plot for integrated dataset of all EECs from all regions in (A). Samples labelled by EEC defined by clustering in individual datasets. (D) Marker expression per EEC in tSNE coordinates from (C). Expression is log2(mean counts per million + 1).
Fig 2.
Co-expression of hormonal markers.
(A) Dot plots representing co-expression of hormones or cell markers (Tph1, Hdc) per GI region. Dot size represents the percentage of cells expressing the gene on the y-axis, dot colour the expression of the gene on the x-axis in log2(mean counts per million + 1). Open circles labelled “none” represent percentages of cells expressing none of these additional hormonal markers. Bars on the right represent the number of cells expressing the hormone on the y-axis. (B) Venn diagrams demonstrating number of EECs co-expressing: top, Cck, Gcg, Nts and Sct in the upper SI; middle, Sct, Gcg, Pyy and Nts in the lower SI; bottom, Insl5, Gcg, Nts and Pyy in the LI. Darker blue represents larger cell number.
Fig 3.
EECs show sub-clustering based on GPCR cell signalling genes alone.
(A) tSNE plot for integrated dataset generated using genes involved in GPCR cell signalling. Samples labelled by EEC defined by clustering in individual datasets. (B) Left: Relative expression of Cck and Mc4r, which were each differentially-enriched (padj < 0.05) in the I-cell-labelled cluster. Right: CCK secretion, measured as proCCK(21–44) by liquid-chromatography mass spectrometry, from mouse USI primary tissue after 2 hr stimulation with control (Con; 2% DMSO), IBMX (10 μM)/forskolin (10 μM) (IF), or setmelanotide (Set; 10 μM). Concentrations were normalised to mean concentrations in control wells from the same experiment. Horizontal lines represent mean concentration (n = 6 experiments, 3–4 samples per experiment, colour-coded by experiment). Statistical analysis by two-way ANOVA with experimental run as the second factor, and Dunnett’s post-hoc test (*** p < 0.001). (C) Relative expression of selected differentially-expressed (padj < 0.05) genes based on clustering analysis. Panels represent, from top to bottom, K-cells, L- and N-cells, and ECs.
Fig 4.
EECs respond to multiple nutrients.
(A) Mean expression of selected nutrient-sensing GPCRs, Sglt1 (carbohydrates), Ffar1, Ffar4 and Gpr119 (lipids), Gpbar1 (bile acids), and Gpr142 and Casr (aromatic amino acids), per clustered EEC type per GI region. Expression is log2(mean counts per million + 1). (B) Dot plots representing co-expression of nutrient-sensing GPCRs, in cells expressing given hormones, per GI region. Cells represented include Gcg-, Cck-, and Gip-expressing cells in the USI, Gcg-, Nts-, and Gip-expressing cells in the LSI, and Gcg- and Insl5-expressing cells in the LI. Dot size represents the percentage of cells expressing the GPCR on the y-axis, dot colour the expression of the GPCR on the x-axis in log2(mean counts per million + 1). Bars on the right represent the number of cells expressing the GPCR on the y-axis. GPCR and ion channel genes in (A) and (B) are selected as the top 20 most-expressed differentially-expressed (padj < 0.05) genes of that type, per GI region. (C) Example trace of Ca2+ levels in a murine L-cell, from organoids derived from GLU-Cre x Venus mice, treated in series with AM1638 (10 μM; blue), glucose (10 mM; yellow), tryptophan (20 mM; red), and positive-control KCl (70 mM; grey), separated by wash steps. The time over which each treatment was imposed is represented by the coloured vertical bars. Ca2+ levels were measured using fluorescent dye Fura2-AM, and are depicted as the 340/380 ratio. (D) Ca2+ levels (fura2 340/380 ratios) during treatment (R) relative to baseline (R0, measured immediately prior to treatment). Horizontal lines represent mean levels (n = 4 experiments, 3–6 cells per experiment). Statistical analysis by one-tailed Student’s t-test (μ = 1; *** padj < 10–3). (E) Venn diagram of cells from (D) responding to each treatment, demonstrating ability of some cells to respond to multiple treatment types.