The bHLH transcription factor ASCL1 promotes differentiation of endocrine cells in the stomach and is regulated by Notch signaling

Abstract

Notch signaling regulates gastrointestinal stem cell proliferation and differentiation yet Notch-regulated transcriptional effectors of gastric epithelial cell differentiation are poorly understood. Here we tested the role of the bHLH transcription factor Achaete-Scute homolog 1 (ASCL1) in gastric epithelial cell differentiation, and its regulation by Notch. Newborn Ascl1 null mice showed a loss of expression of markers of neurogenin-3-dependent enteroendocrine cells, with normal expression of enterochromaffin-like cells, mucous cells, chief cells, and parietal cells. In adult mice, Ascl1 gene expression was observed in the stomach, but not the intestine, with higher expression in antral than corpus epithelium. Lineage tracing in Ascl1-CreER^T2; Rosa26-LSL-tdTomato mice revealed single, scattered ASCL1⁺ cells in the gastric epithelium, demonstrating expression in antral gastrin- and serotonin-producing endocrine cells. ASCL1-expressing endocrine cells persisted for several weeks posttamoxifen labeling with a half-life of approximately 2 months. Lineage tracing in Gastrin-CreER^T2 mice demonstrated a similar lifespan for gastrin-producing cells, confirming that gastric endocrine cells are long-lived. Finally, treatment of Ascl1-CreER^T2; Rosa26-LSL-tdTomato mice with the pan-Notch inhibitor dibenzazepine increased the number of lineage-labeled cells in the gastric antrum, suggesting that Notch signaling normally inhibits Ascl1 expression. Notch regulation of Ascl1 was also demonstrated in a genetic mouse model of Notch activation, as well as Notch-manipulated antral organoid cultures, thus suggesting that ASCL1 is a key downstream Notch pathway effector promoting endocrine cell differentiation in the gastric epithelium.NEW & NOTEWORTHY Although Notch signaling is known to regulate cellular differentiation in the stomach, downstream effectors are poorly described. Here we demonstrate that the bHLH transcription factor ASCL1 is expressed in endocrine cells in the stomach and is required for formation of neurogenin-3-dependent enteroendocrine cells but not enterochromaffin-like cells. We also demonstrate that Ascl1 expression is inhibited by Notch signaling, suggesting that ASCL1 is a Notch-regulated transcriptional effector directing enteroendocrine cell fate in the mouse stomach.

Keywords: EC cell; ECL cell; G cell; endocrine cell turnover; gastric epithelial cell differentiation.

Graphical abstract

Figure 1.

Ascl1 is required for endocrine cell differentiation in the postnatal mouse stomach. A:loss of Ascl1 mRNA in Ascl1^CreER/+ (n = 4) and Ascl1^CreER/CreER mice (n = 6, ***P = 0.0001) was confirmed by qRT-PCR analysis of RNA isolated from whole stomachs at P0. B: histological analysis of Ascl1^CreER/+ and Ascl1^CreER/CreER P0 stomach sections stained with hematoxylin and eosin. C: qRT-PCR analysis of the endocrine lineage transcription factor Neurog3 (**P = 0.0012), the bHLH transcription factor Ptf1a (***p = 0.0002), and the pan-endocrine cell marker Chga (***P = 0.0002) in Ascl1^CreER/+ (n = 4) and Ascl1^CreER/CreER (n = 6) P0 stomach RNA. D: qRT-PCR analysis of endocrine cell markers Gast (***P = 0.0006), Sst (****P < 0.0001), Ghrl (****P < 0.0001), and Tph1 (*P = 0.0189) in Ascl1^CreER/+ (n = 4) and Ascl1^CreER/CreER (n = 6) P0 stomach RNA. E: qRT-PCR analysis of the ECL cell markers Hdc (P = 0.3463) and Lhb (P = 0.8179) in Ascl1^CreER/+ (n = 3–4) and Ascl1^CreER/CreER (n = 3–6) P0 stomach RNA. Data are expressed as mRNA fold change (FC), presented as means ± SE and analyzed using Student’s t test. Scale bars: 50 μm. bHLH, basic helix-loop-helix.

Figure 2.

Ascl1 is not required for parietal, mucous, or chief cell differentiation in the postnatal mouse stomach. qRT-PCR analysis of parietal cell markers Atp4b and Vegfb (A), mucous cell markers Tff2, Gkn2, Gkn3, and Spdef (B), and chief cell markers Gif and Mist1 (C) in Ascl1^CreER/+ (n = 4) and Ascl1^CreER/CreER (n = 6) P0 stomach RNA. Data are expressed as mRNA FC and presented as means ± SE. Comparisons were not significantly different using Student’s t test. FC, fold change.

Figure 3.

Ascl gene family member expression in the adult mouse GI epithelium. Gene expression analysis of adult mouse corpus, antrum, and duodenum was determined by qRT-PCR analysis of RNA extracted from corpus glands (Corp, n = 4), antral glands (Ant, n = 4), or duodenal crypts (Duod, n = 4). A: tissue purity was validated by measurement of mRNAs for Atp4a (corpus), Gast (antrum), and Cdx2 (duodenum). B: qRT-PCR analysis of Ascl family members showed that Ascl1 mRNA was restricted to the stomach with highest expression in the antrum. mRNA abundance relative to expression of Gapdh is presented as means ± SE. GI, gastrointestinal; N.D., not detected.

Figure 4.

Ascl1 marks a subset of endocrine cells in the mouse stomach. The expression pattern and cellular localization of Ascl1 were determined via lineage tracing in Ascl1^CreER/+; ROSA26^Tom/+ mice 5 days posttamoxifen (TX) treatment. Frozen tissue sections from corpus or antrum, or isolated antral glands (E), were immunostained for E-cadherin (E-cad, epithelial cells) (A, C, and E), CHGA (green, endocrine cells) (B and E), CD45 (green, immune cells) (D), GAST (green, G cells) (F), or TPH1 (green, EC cells) and covisualized with tdTomato fluorescence (red, ASCL1⁺ cells) (G). Arrowheads identify some costained cells, with higher powered views shown in B, E, and F. Arrows mark nonepithelial TPH1⁺ cells. Scale bars: 50 μm. (H): all tdTomato-labeled cells were counted along the entire antrum and then tallied for copositivity with TPH1- or GAST-expressing endocrine cells. Data are presented as means ± SE; n = 3 mice. ASCL1, achaete-scute homolog 1; CHGA, chromogranin A; GAST, gastrin; TPH1, tryptophan hydroxylase 1.

Figure 5.

Ascl1 labels long-lived endocrine cells. Ascl1 lineage labeling persists for several weeks in the gastric antrum. A: immunostaining for GAST (green, G cells) in Ascl1^CreER/+; ROSA26^Tom/+ mice 2 mo post-TX treatment. Arrowhead marks a Tomato⁺ G cell. B: morphometric analysis of lineage-labeled cells in Ascl1^CreER/+; ROSA26^Tom/+ mice at indicated time points post-TX (n = 2–3 mice/group). C: nonlinear regression analysis was used to measure the exponential decay of Ascl1 lineage-labeling in the gastric antrum of Ascl1^CreER/+; ROSA26^Tom/+ mice over time. Data are presented as means ± SE. Scale bar: 50 μm. GAST, gastrin; TX, tamoxifen.

Figure 6.

G cells in the adult mouse gastric antrum are long lived. A–D: lineage tracing in Gast^CreER/+; ROSA26^Tom/+ mice at various time points post-TX treatment. Single, scattered Tomato⁺ cells were observed at the antral gland base, consistent with the location of G cells. E: morphometric analysis of lineage-labeled cells at indicated time points post-TX (n = 3–4 mice/group). Data are presented as means ± SE. Scale bars: 50 μm. TX, tamoxifen.

Figure 7.

Notch signaling regulates gastric antral Ascl1 expression. Analysis of Ascl1 lineage-labeled cells in the gastric antrum after Notch inhibition. A: schematic of the experimental design. Ascl1^CreER/+; ROSA26^Tom/+ mice were treated with either vehicle or the Notch inhibitor DBZ once daily for 5 days, and tissue collected for histological analysis of Tomato-marked cells on day 6. On day 3, all mice were treated with TX to label ASCL1⁺ cells. B: ASCL1 lineage-marked cells (red) in vehicle- or DBZ-treated Ascl1^CreER/+; ROSA26^Tom/+ mice were visualized via tdTomato fluorescence in frozen tissue sections. Arrowheads indicate Tomato⁺ cells. DAPI (blue) used as nuclear counterstain. Scale bars: 50 μm. C: morphometric analysis of Tomato⁺ cells in vehicle (Veh)- or DBZ-treated Ascl1^CreER/+; ROSA26^Tom/+ mice (n = 4/group, *P = 0.04). Analysis of Ascl1 expression and endocrine cell differentiation after Notch activation. D: schematic of experimental design. Lgr5^CreER/+; ROSA26^NICD/+ mice (n = 4–8/group) were treated daily for 5 days with either vehicle or TX, and tissue collected 4 wk later for gene expression analysis. E and F: qRT-PCR analysis of the Notch target gene Hes1, transcription factors Ascl1 and Neurog3, and endocrine cell markers Gast and Tph1. Notch activation resulted in increased Notch target gene Hes1 expression (**P = 0.0014), reduced Ascl1 (***P = 0.0002) and Neurog3 (**P = 0.0042) transcription factor expression, and reduced markers of mature gastrin- (**P = 0.0071) and serotonin-producing (*P = 0.0450) enteroendocrine cells. Data are expressed as mRNA FC, presented as means ± SE and analyzed using Student’s t test. ASCL1, achaete-scute homolog 1; DBZ, dibenzazepine; FC, fold change; TX, tamoxifen.

Figure 8.

Notch regulates Ascl family member expression in gastric antral organoids. Analysis of Ascl family mRNA abundance in wild-type antral organoids after Notch inhibition. A: schematic of the experimental design. Established antral organoids were passaged on day 0 and then treated every other day with vehicle (Veh) or the pan-Notch inhibitor DAPT. RNA was isolated from organoids harvested on day 6. B: qRT-PCR measurement of Ascl1 (*P = 0.0451), Ascl2 (P = 0.4872), and Ascl3 (**P = 0.0047) mRNA abundance in vehicle- (n = 8, 3 wells pooled/sample) or DAPT-treated (n = 7, 3 wells pooled/sample) gastric antral organoids. Analysis of Ascl family mRNA abundance in Notch-activated antral organoids. C: schematic of experimental design. ROSA26^NICD/+ or Lgr5^CreER/+; ROSA26^NICD/+ mice were treated with TX once daily for 5 days and antral organoids initiated from isolated antral glands 1 mo post-TX. After initiation, organoids were passaged at least three times to establish stable organoids, and RNA was isolated for gene expression analysis. D: qRT-PCR measurement of Ascl1 (*P = 0.0237), Ascl2 (**P = 0.0023), and Ascl3 (*P = 0.0354) mRNA abundance in ROSA26^NICD/+ (n = 8, 3 wells pooled/sample) or Lgr5^CreER/+; ROSA26^NICD/+ (n = 8, 3 wells pooled/sample) antral organoids. Quantitative data are presented as means ± SE and analyzed vs. control group using Student’s t test. Prism outlier identification test was used for statistical analysis of Ascl1 in D. DAPT, N-(N-(3,5-difluorophenacetyl-l-alanyl))-(S)-phenylglycine t-butyl ester; TX, tamoxifen.