TCF7L1 Controls the Differentiation of Tuft Cells in Mouse Small Intestine

Continuous and rapid renewal of the intestinal epithelium depends on intestinal stem cells (ISCs). A large repertoire of transcription factors mediates the correct maintenance and differentiation of ISCs along either absorptive or secretory lineages. In the present study, we addressed the role of TCF7L1, a negative regulator of WNT signalling, in embryonic and adult intestinal epithelium using conditional mouse mutants. We found that TCF7L1 prevents precocious differentiation of the embryonic intestinal epithelial progenitors towards enterocytes and ISCs. We show that Tcf7l1 deficiency leads to upregulation of the Notch effector Rbp-J, resulting in a subsequent loss of embryonic secretory progenitors. In the adult small intestine, TCF7L1 is required for the differentiation of secretory epithelial progenitors along the tuft cell lineage. Furthermore, we show that Tcf7l1 promotes the differentiation of enteroendocrine D- and L-cells in the anterior small intestine. We conclude that TCF7L1-mediated repression of both Notch and WNT pathways is essential for the correct differentiation of intestinal secretory progenitors.


Introduction
The gut epithelium plays a central role in immune surveillance, nutrient absorption, and synthesis of hormones. Its constant exposure to pathogens and xenobiotics leads to tissue injury. Abnormalities in the epithelial cell barrier and functions result in both acute disorders, such as colitis and diarrhoea, and chronic diseases, such as cancer. Therefore, gut epithelial cells must be both rapidly replaced and correctly specified. To maintain tissue homeostasis, intestinal stem cells (ISCs) constantly generate transit-amplifying progenitors that, in turn, are specified either along the absorptive enterocyte or the secretory lineages [1]. The secretory progenitors progressively differentiate towards mucous-secreting goblet, hormone-secreting enteroendocrine, immuno-modulating Paneth, and tuft cells.
WNT and Notch signalling molecules are crucial for stem cell self-renewal and balanced generation of absorptive and secretory cell types [1]. WNT signalling governs these processes by regulating a large number of genes encoding for transcription factors and components of signalling pathways [2]. On one side, cell cycle promoting genes, such as c-Myc and cyclin D1 are transcriptionally activated by WNT signals [3]. On the other side, numerous transcription factors essential for the specification of undifferentiated progenitors along the secretory lineage (Atoh1), tuft (Pou2f3), and enteroendocrine cell lineages (Neurog3, Neurod1, Foxa2, and Rfx3) are direct targets of WNT signalling [2]. In the absence of WNT ligands, the downstream components of the WNT signalling T-cell-specific transcription factor 7 like 1 (TCF7L1) and TCF7L2, interact with corepressor proteins, such as Groucho/Transducin-like enhancer of split (GRO/TLE) [4] and C-Terminal-Binding Protein (CtBP) [5], that recruit histone deacetylases to the promoters of WNT target genes, leading Tcf7l1 tm1a(EUCOMM)Wtsi mice were obtained from EMMA. Shh EGFP-Cre mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). CD1 mice were obtained from Charles River Laboratories (Cologne, Germany). ROSA26::FLPe mice were a gift from Thomas Hankeln, JGU, Mainz. Mouse colonies were maintained in a certified SPF animal facility per European guidelines. All mice were housed on a 12-h light/dark cycle with constant access to food and tap water. All the animal experiments were performed according to guidelines of the central animal facility institution (TARC, Mainz University Medical Center) representing those of the German Animal Welfare Act and the European Directive 2010/63/EU for the protection of animals used for scientific purposes. Breeding was approved by the local authorities (Kreisverwaltung Mainz-Bingen, Mainz). Reporting was carried out according to the ARRIVE guidelines for reporting in vivo experiments.

Low Cell Number RNA-Sequencing
Small intestines were dissected from mouse embryos at day 13.5 (E13.5), cut into 2 mm pieces, and incubated for 10 min with 0.15 mg/mL collagenase (Sigma, Merck KGaA, Darmstadt, Germany) in PBS at 37 • C with shaking at 800 rpm. Single-cell suspensions were collected via centrifugation at 200× g for 5 min and resuspended in 200 mL of PBS supplemented with 2% goat serum. Cells were stained with APC-conjugated anti-EpCAM antibody 1:1000 (eBioscience, San Diego, CA, USA) for 30 min at room temperature. Living cells were gated via DAPI dye exclusion. Fluorescence-activated cell sorting analysis was performed using the BD FACS Aria II SORP cell sorter (85 µM nozzle).
For ultralow cell number RNA-sequencing five hundred EpCAM+ embryonic intestinal cells were isolated by FACS directly in 7 µL of lysis buffer (Takara Bio, Kusatsu City, Japan) supplemented with 5% RNase inhibitor and stored at −80 • C. cDNA was synthesised using SMARTer v4.0 kit (Takara Bio) according to the manufacturer's instructions. Amplification was performed for 15 cycles. After cDNA fragmentation (Covaris, Woburn, MA, USA), libraries were prepared using the Ovation Ultralow v2 Library System (NuGEN, Houston, TX, USA) according to the manufacturer's instructions.

RNA-Sequencing Data Analysis
Libraries were sequenced on an Illumina NextSeq 500 (NextSeq Control software v. 2.1.0.31) with a read length of 84 based and demultiplexed using bclfastq (v.2.19.1). Samples were mapped using STAR v2.5.2b [20] against iGenomes mm9. Reads per gene were counted using feature Counts v. 1.5.1 [21]. Read counts are based on uniquely mapping reads. For mapping as well as read counting the reference gene model downloaded from UCSC on 6 March 2013 was used (as provided with iGenomes). The differential expression analysis was performed using Bioconductor release 3.6 [22] and DESeq2 v1.18.1 [23] using standard parameters for testing and modelling as well as independent filtering. Finally, RPKM values were calculated per gene based on FPM (robust counts per million mapped fragments) values provided by DESeq2 and divided by the gene length and multiplied by 1000.
Sequencing-depth-normalised coverage tracks (bigwig) were generated using Deep-Tools v. 2.4.3 [24]. The raw sequencing data as well as the read counts per gene were deposited on the NCBI Gene Expression Omnibus database. The genes were considered significantly differentially expressed if they fit the following criteria: log 2 FC ≥ 0.5, FDR < 0.01, RPKM ≥ 50. Functional annotation analysis was performed using DAVID Functional Annotation Tool [25].

Quantitative PCR
For qPCR, 10 ng of cDNA generated with SMARTer v4.0 kit (Takara Bio) from EpCAM+ embryonic intestinal cells were used. Expression changes were normalized to Tbp. PCR primers were designed using Primer Blast "http://www.ncbi.nlm.nih.gov/tools/primerblast/ (accessed on 17 May 2023)". PCR was performed using SYBR green containing

RNA In Situ Hybridization
Embryos or adult small intestines were dissected and fixed in 4% PFA overnight. Prior to fixation, adult small intestines were divided into three equal parts-anterior, mid, and posterior. The following day tissues were dehydrated by incubation in a series of 30%, 50%, 70%, and 99% ethanol solutions and washed in xylol twice. Afterwards, tissues were embedded in paraffin and cut into 10 µm sections with the Leica RM2235 Rotary Microtome. Paraffin slides were incubated at 57 • C for 2 h, deparaffinized in xylol and rehydrated in a series of 99%, 70%, 50%, and 30% ethanol solutions, followed by PBS. Rehydrated sections were fixed in 4% PFA for 15 min, bleached in 6% H 2 O 2 for 15 min, and treated with 10 ug/mL proteinase K solution for 10 min. The treated slides were immediately fixed in 4% PFA for another 15 min, washed twice in PBS, acetylated in freshly prepared 0.25% acetic anhydride in 100 mM Tris-Cl (pH 7.5) for 10 min. Then, tissues were equilibrated in 2 × SSC buffer, pH 5 and dehydrated in a series of 30%, 50%, 70%, and 99% ethanol solutions. The dehydrated slides were dried on air and hybridized with digoxygenin-labelled RNA probes for Ghrl, Gip, and Sst overnight at 63 • C. The following day, hybridized sections were washed in 5×, 2×, 1×, and 0.2 × SSC buffer at 60 • C and then proceeded to the blocking stage and were incubated overnight with sheep anti-digoxigenin antibody 1:3000 (Roche, Basel, Switzerland). The following day, slides were washed in TBSX buffer and stained with NBT/BCIP (Roche, Basel, Switzerland) until the signal developed. Afterwards, slides were dehydrated and mounted with ROTI ® Histokitt (Carl Roth GmbH, Karlsruhe, Germany). Images were acquired on the Olympus IX2-UCB microscope.
RNA in situ hybridization for Dclk1 was performed using RNAscope™ 2.5 HD Duplex Assay Kit (ACD Bio-Techne, Newark, CA, USA) following the manufacturer's instructions. Target retrieval was performed for 15 min, and protease pretreatment was performed for 30 min. Images were acquired on the Olympus IX2-UCB microscope.
Slides were incubated with the primary antibody overnight, then washed thrice in PBS, 5 min each wash. Next, biotinylated secondary antibodies (Dianova, Biozol, Hamburg, Germany) were applied on the slides in 1:1000 dilution and incubated for two hours at room temperature. Next, the slides were washed thrice in PBS and incubated with Vectastain Elite ABC Reagent (Vector Laboratories) for 1 h at room temperature. The slides were washed in PBS and incubated with SIGMA FAST 3,3 -Diaminobenzidine Tablets (Sigma-Aldrich, St. Louis, MI, USA) dissolved in PBS. Afterwards, the slides were dehydrated and mounted with ROTI ® Histokitt (Carl Roth GmbH, Karlsruhe, Germany). Images were acquired on the Olympus IX2-UCB microscope.

Periodic Acid-SCHIFF (PAS) Staining
Paraffin sections (5 µm) were deparaffinized and hydrated to deionized water. Next, slides were immersed in periodic acid solution (Sigma-Aldrich, St. Louis, MI, USA) for 5 min at room temperature. Then, the slides were rinsed several times in distilled water and immersed in Schiff's Reagent (Sigma-Aldrich, St. Louis, MI, USA) for 15 min. At the end of the incubation, slides were washed in running tap water for 5 min, dehydrated, mounted with ROTI ® Histokitt (Carl Roth GmbH, Karlsruhe, Germany), and imaged on the Olympus IX2-UCB microscope.

Haematoxylin and Eosin (H&E) Co-Staining
Paraffin sections (5 µm) were deparaffinized, hydrated, and stained in Gill's haematoxylin solution 2 (Sigma-Aldrich) for 3 min. After that, the slides were washed in running tap water for 5 min, rinsed in distilled water, and dipped in 95% ethanol. Next, the eosin costaining was performed by immersing slides in the eosin solution (Sigma-Aldrich) for 1 min. Stained slides were washed in 95% ethanol, dehydrated, and mounted with ROTI ® Histokitt (Carl Roth). Images were acquired on the Olympus IX2-UCB microscope.

Statistical Analysis
Information on sample size and statistical tests used for each experiment are indicated in the figure legends. All staining counts were analysed using a two-tailed nested t-test. Data are shown as means with SD, a p-value of ≤0.05 was considered significant. All data were tested for normality using D'Agostino and Pearson, Anderson-Darling, and Shapiro-Wilk tests. All described analyses were performed in GraphPad Prism v9.5.0.

TCF7L1 Controls the Expression of Secretory Lineage Genes during Gut Development
To study the functions of TCF7L1 during gut development, we have used two mouse strains ( Figure 1A). The first is a conditional Tcf7l1-lacZ line in which the splice acceptor-lacZ gene flanked by FRT sites is inserted in the fifth intron of the Tcf7l1 gene [26]. Splicing of lacZ to the fifth exon of Tcf7l1 generates both a null and a reporter allele. The second strain is a conditional Tcf7l1 in which exon 6 is flanked by loxP sites, generating a frameshift and triggering nonsense-mediated decay of the mutant RNA upon Cre-mediated recombination [26]. We first examined the expression pattern of Tcf7l1 at embryonic day 13.5 (E13.5), the earliest stage when stem cell, secretory and absorptive lineage specific genes are activated in the gut epithelium [10]. LacZ staining revealed that Tcf7l1 is expressed in the epithelium of the anterior small intestine ( Figure 1B). In contrast, the expression of Tcf7l1-lacZ reporter was not observed in the posterior half of the small intestine ( Figure 1B), indicating that Tcf7l1 is differentially expressed along the anterior-posterior axis at this developmental stage.
Mouse embryos lacking Tcf7l1 die around E8.5 [27]. We therefore used the Shh Cre-EGFP allele, which displays Cre activity as early as E9.5 [28,29], to inactivate Tcf7l1 in the developing gut epithelium. The small intestines of Shh Cre-EGFP :Tcf7l1 lox/lox embryos were indistinguishable from wild type controls at E13.5. To explore whether the loss of Tcf7l1 leads to ectopic activation of WNT target genes, we isolated EpCAM-positive intestinal epithelial cells using fluorescence-activated cell sorting (FACS) (Supplementary Figure S1A  Mouse embryos lacking Tcf7l1 die around E8.5 [27]. We therefore used the Shh Cre-EGFP allele, which displays Cre activity as early as E9.5 [28,29], to inactivate Tcf7l1 in the developing gut epithelium. The small intestines of Shh Cre-EGFP :Tcf7l1 lox/lox embryos were indistinguishable from wild type controls at E13.5. To explore whether the loss of Tcf7l1 leads to ectopic activation of WNT target genes, we isolated EpCAM-positive intestinal epithelial cells using fluorescence-activated cell sorting (FACS) (Supplementary Figure S1A) from the mutant and control embryos and performed RNA-sequencing analysis.
We found that around 400 genes were upregulated and that around 800 genes were downregulated (log 2 FC ≥ 0.5, FDR < 0.01, RPKM ≥ 50) upon loss of Tcf7l1 ( Figure 1C-E, Supplementary Figure S1B, and Supplementary Tables S1 and S2). The adult stem cell signature genes Olfm4, Hmgcs2, and Kcne3 were significantly upregulated, whereas Lgr5 expression did not change compared to the wild type ( Figure 1E). Of note, the expression of ISC signature genes varies along the anterior-posterior axis during embryonic development. While Olfm4, Hmgcs2, and Kcne3 are expressed at higher levels in the anterior ISC progenitors, Lgr5 is more highly expressed in the posterior ISCs progenitors [30]. Therefore, the changes in Lgr5 expression in the anterior small intestine, where Tcf7l1 is expressed, Moreover, the expression of enterocyte-specific genes, including Fabp1, Fabp2, Lgals3, Aldob, and Ephx2, were upregulated in Tcf7l1 mutant compared to control cells ( Figure 1C-E). Thus, TCF7L1 prevents precocious differentiation of the embryonic intestinal epithelium both towards ISCs and along the absorptive enterocyte lineage.
Functional annotation analysis of downregulated transcripts revealed enrichment for regulation of transcription (fold enrichment = 1.84, FDR = 5 × 10 −5 ) and kinase activity (fold enrichment = 2.14, FDR = 1.1 × 10 −4 ). Interestingly, we detected a strong decrease in the expression of genes encoding for transcription factors regulating the differentiation of enteroendocrine cells, including Nkx6-2, Nkx6-3, Rfx6, Pdx1, and Foxa2 ( Figure 1C-E and Supplementary Figure S1C-E), indicating that Tcf7l1 is required for the differentiation of the embryonic epithelium along the enteroendocrine lineage. Consistently, we observed a complete loss of ChgB expression ( Figure 1D), which is a marker of the enteroendocrine cells. Furthermore, markers of goblet cells, including Fcgbp, Agr2, and Spdef were also downregulated in Tcf7l1 mutant embryos ( Figure 1D,E). Our results show that Tcf7l1 promotes differentiation of the intestinal epithelium towards both enteroendocrine and goblet cell lineages during embryogenesis.

TCF7L1 Is Necessary for Tuft Cell Differentiation
Single-cell RNA-sequencing analysis revealed that Tcf7l1 is expressed in tuft cells in the adult small intestine [11]. RNA in situ hybridization analysis showed that Tcf7l1 is expressed in a few spindle-shaped cells located mostly in the intestinal crypts (Figure 2A), suggesting that these cells could be tuft cell progenitors. To determine whether transcriptional changes during embryogenesis affect differentiation of the adult ISCs and to examine the functions of Tcf7l1 in tuft cells, we analysed Shh Cre-EGFP :Tcf7l1 lox/lox small intestines at the age of four months. Adult Shh Cre-EGFP :Tcf7l1 lox/lox mice did not display any gross morphological or behavioural abnormalities. The villi-crypt architecture of mutant mice appeared normal (Supplementary Figure S2A,B).
Immunohistochemical staining for proliferation marker Ki67 was similar between mutant and wild type mice (Supplementary Figure S2C,D). Additionally, the numbers and localization of goblet ( Figure 2B-D) and Paneth cells (Supplementary Figure S2E,F) were similar between Shh Cre-EGFP :Tcf7l1 lox/lox and wild-type mice. In contrast, the number of tuft cells was significantly decreased in Shh Cre-EGFP :Tcf7l1 lox/lox compared to wild-type mice as revealed by RNA in situ hybridization for Doublecortin-like kinase 1 (Dclk1), a marker of tuft cells ( Figure 2E-G). The loss of tuft cells was observed in all parts of the small intestine along the anterior-posterior axis, with a 3-fold reduction in the anterior and posterior and a 3.5-fold in the middle parts ( Figure 2G). Thus, we conclude that TCF7L1 is essential for the differentiation of tuft cells in the adult small intestine. However, transcriptional changes in either ISC signature genes or goblet progenitor markers caused by the loss of Tcf7l1 during embryogenesis do not disturb stem cell proliferation and differentiation along the goblet/Paneth cell lineage in the adult gut, which confirms the results of the previous study [9]. Immunohistochemical staining for proliferation marker Ki67 was similar between mutant and wild type mice (Supplementary Figure S2C,D). Additionally, the numbers and localization of goblet ( Figure 2B-D) and Paneth cells (Supplementary Figure S2E,F) were similar between Shh Cre-EGFP :Tcf7l1 lox/lox and wild-type mice. In contrast, the number of tuft cells was significantly decreased in Shh Cre-EGFP :Tcf7l1 lox/lox compared to wild-type mice as revealed by RNA in situ hybridization for Doublecortin-like kinase 1 (Dclk1), a marker of tuft cells (Figure 2E-G). The loss of tuft cells was observed in all parts of the small intestine along the anterior-posterior axis, with a 3-fold reduction in the anterior and posterior and a 3.5-fold in the middle parts ( Figure 2G). Thus, we conclude that TCF7L1 is essential for the differentiation of tuft cells in the adult small intestine. However, transcriptional changes in either ISC signature genes or goblet progenitor markers caused by the loss of Tcf7l1 during embryogenesis do not disturb stem cell proliferation and differentiation along the goblet/Paneth cell lineage in the adult gut, which confirms the results of the previous study [9].

(G) Quantification of Dclk1-expressing cells in WT (grey) and
Shh Cre-EGFP :Tcf7l1 lox/lox (green) mice. Every dot shows an average number of stained cells per villus in a view field. Scale bar: 50 µm (A,C,E,F) Error bars are ±SD, n = 3 mice, ns stands for not significant, * p < 0.05, ** p < 0.01, and *** p < 0.001 according to a two-tailed nested t-test.

TCF7L1 Is Dispensable for the Differentiation of Enterochromaffin Cells
Our transcriptome analysis of the intestinal epithelium revealed downregulation of ChgB, a marker of the enterochromaffin cells, and Tox3 ( Figure 1D and Supplementary  Table S2) encoding for the HMG-box containing the transcription factor essential for differentiation along the enterochromaffin lineage [31] in Shh Cre-EGFP :Tcf7l1 lox/lox compared to wild-type embryos. To determine whether the enterochromaffin cell population was affected by Tcf7l1 knockout, we performed immunohistochemical analysis for the panmarker of enteroendocrine cells (EECs), Chromogranin A ( Figure 3A,B), and serotonin (5-HT), a marker for enterochromaffin cells (Figure 3D,E). Statistical analysis of the counts for CHGA-positive ( Figure 3C) and 5-HT-positive ( Figure 3F) cells did not show a significant difference between Tcf7l1 KO and wild-type mice, indicating that TCF7L1 is not required for the differentiation of enterochromaffin cells in the adult small intestine.
wild-type embryos. To determine whether the enterochromaffin cell population was affected by Tcf7l1 knockout, we performed immunohistochemical analysis for the panmarker of enteroendocrine cells (EECs), Chromogranin A ( Figure 3A,B), and serotonin (5-HT), a marker for enterochromaffin cells (Figure 3D,E). Statistical analysis of the counts for CHGA-positive ( Figure 3C) and 5-HT-positive ( Figure 3F) cells did not show a significant difference between Tcf7l1 KO and wild-type mice, indicating that TCF7L1 is not required for the differentiation of enterochromaffin cells in the adult small intestine.

TCF7L1 Promotes the Differentiation of L-and D-Cells
The expression of Forkhead box A2 (Foxa2) was reduced eight-fold, and that of regulatory factor X 6 (Rfx6) was reduced three-fold in Shh Cre-EGFP :Tcf7l1 lox/lox embryos compared to wild-type controls (Figures 1C-E and S1C,D and Supplementary Table S2). Winged he-

TCF7L1 Promotes the Differentiation of L-and D-Cells
The expression of Forkhead box A2 (Foxa2) was reduced eight-fold, and that of regulatory factor X 6 (Rfx6) was reduced three-fold in Shh Cre-EGFP :Tcf7l1 lox/lox embryos compared to wild-type controls ( Figure 1C-E and Figure S1C,D and Supplementary Table S2). Winged helix/forkhead box transcription factor FOXA2 controls the differentiation of enteroendocrine progenitors along L-and D-cell lineages [32]. Moreover, Rfx6 encoding for a winged helix transcription factor promotes the differentiation of all peptidergic EECs [33]. While the expression of both transcription factors is restricted to the enteroendocrine cells in the adult gut epithelium, they are broadly expressed in the embryonic epithelium [33,34]. The loss of Foxa2 and Rfx6 expression in Tcf7l1 mutant embryos may change chromatin landscapes in the embryonic ISC progenitors leading to the alteration of adult ISC differentiation potential. Therefore, we examined the effect of TCF7L1 loss on the differentiation of L-and D-cells. Interestingly, we found a significant reduction (1.5-fold) of glucagon-like peptide 1 (GLP-1)-positive L-cells in the anterior small intestine of Tcf7l1 knockout compared to wild-type mice. (Figure 4A-C).
loss of Foxa2 and Rfx6 expression in Tcf7l1 mutant embryos may change chromatin landscapes in the embryonic ISC progenitors leading to the alteration of adult ISC differentiation potential. Therefore, we examined the effect of TCF7L1 loss on the differentiation of L-and D-cells. Interestingly, we found a significant reduction (1.5-fold) of glucagon-like peptide 1 (GLP-1)-positive L-cells in the anterior small intestine of Tcf7l1 knockout compared to wild-type mice. (Figure 4A-C). The numbers of L-cells in the middle and posterior parts were also lower in Tcf7l1KO animals, yet a statistical significance was not reached (p-values for middle and posterior regions are 0.081 and 0.084, respectively). Furthermore, RNA in situ hybridization analysis for the D-cell marker Somatostatin (Sst) showed a significant reduction (1.5-fold) in Dcell number in the anterior part of the small intestine of Tcf7l1 knockout mice compared to controls ( Figure 5D,F). Altogether, these findings suggest that TCF7L1 promotes the differentiation of enteroendocrine cells along D-and L-cell lineages in the anterior small intestine. The numbers of L-cells in the middle and posterior parts were also lower in Tcf7l1KO animals, yet a statistical significance was not reached (p-values for middle and posterior regions are 0.081 and 0.084, respectively). Furthermore, RNA in situ hybridization analysis for the D-cell marker Somatostatin (Sst) showed a significant reduction (1.5-fold) in D-cell number in the anterior part of the small intestine of Tcf7l1 knockout mice compared to controls ( Figure 5D,F). Altogether, these findings suggest that TCF7L1 promotes the differentiation of enteroendocrine cells along D-and L-cell lineages in the anterior small intestine.

Deletion of Tcf7l1 Does Not Affect the Differentiation of K-and X-Cells
RFX6 is also required for the differentiation of enteroendocrine progenitors along K-and X-cell lineages [33]. Therefore, we examined whether Tcf7l1 loss in the intestinal epithelium resulted in similar defects. RNA in situ hybridization analysis for the expression of X-cell marker Ghrelin (Ghrl) did not reveal significant differences in the number of Ghrlpositive cells between Shh Cre-EGFP :Tcf7l1 lox/lox and control small intestines ( Figure 5A-C). Additionally, no significant differences were found in the number of Gastric inhibitory peptide (Gip)-positive K-cells between Tcf7l1 knockout and wild-type mice ( Figure 5D-F). These results indicate that TCF7L1 is not necessary for X-and K-cell differentiation.

Deletion of Tcf7l1 Does Not Affect the Differentiation of K-and X-Cells
RFX6 is also required for the differentiation of enteroendocrine progenitors along Kand X-cell lineages [33]. Therefore, we examined whether Tcf7l1 loss in the intestinal epithelium resulted in similar defects. RNA in situ hybridization analysis for the expression of X-cell marker Ghrelin (Ghrl) did not reveal significant differences in the number of Ghrlpositive cells between Shh Cre-EGFP :Tcf7l1 lox/lox and control small intestines ( Figure 5A-C). Additionally, no significant differences were found in the number of Gastric inhibitory peptide (Gip)-positive K-cells between Tcf7l1 knockout and wild-type mice ( Figure 5D-F). These results indicate that TCF7L1 is not necessary for X-and K-cell differentiation.

Discussion
The WNT/β-catenin/TCF signalling pathway plays an important role in many developmental processes as well as during the maintenance and differentiation of adult stem cells. In this study, we examined the functions of TCF7L1 in the developing and adult gut epithelium. We show that in midgestation mouse embryos, TCF7L1 is required for the correct specification of intestinal epithelial cells along all secretory lineages. However, in the adult intestinal epithelium, TCF7L1 is necessary for the differentiation of secretory progenitors along the tuft cell lineage. Furthermore, a loss of Tcf7l1 leads to a reduction in

Discussion
The WNT/β-catenin/TCF signalling pathway plays an important role in many developmental processes as well as during the maintenance and differentiation of adult stem cells. In this study, we examined the functions of TCF7L1 in the developing and adult gut epithelium. We show that in midgestation mouse embryos, TCF7L1 is required for the correct specification of intestinal epithelial cells along all secretory lineages. However, in the adult intestinal epithelium, TCF7L1 is necessary for the differentiation of secretory progenitors along the tuft cell lineage. Furthermore, a loss of Tcf7l1 leads to a reduction in enteroendocrine GLP-1-secreting L-cells and somatostatin-positive D-cells in the anterior small intestine.
In the embryonic intestinal epithelium, Tcf7l1 is expressed at least till E14.5 [13]. The decline of Tcf7l1 transcription coincides with the activation of WNT/β-catenin signalling [10]. The repressor functions of TCF7L1 are essential for the self-renewal and differentiation of pluripotent stem cells and various tissue-specific progenitors. During early gastrulation, TCF7L1 represses pluripotency-associated genes, such as Oct4, Sox2, and Nanog, as well as genes promoting cell specification, including FoxA2, Brachyury, and Lef1 [2,35]. This ensures a timely differentiation of the pluripotent epiblast cells toward the mesendoderm and neuroectoderm [27,35]. During neuronal differentiation, TCF7L1 inhibits the expression of transcription factor NEUROG1 and keeps neural progenitor cells in a self-renewing state to prepare them for further differentiation in the presence of WNT signals [36]. It is possible that TCF7L1 is required to repress genes promoting differentiation of the embryonic intestinal epithelial cells to secure an appropriate growth of the tissue. Consistent with the idea, we have observed that the expression of enterocyte-specific genes, including Aldob, Alpi, Apoa4, Fabp1, and Fabp2, was upregulated in Tcf7l1 mutant embryos. These results indicate that TCF7L1 is required to lock the embryonic gut epithelium in an undifferentiated state. However, the expression of the WNT target genes, such as Axin2, Lgr5, and Slc12a2, was not affected. This confirms that the loss of TCF7L1 alone in the absence of WNT signals does not activate the transcription of WNT target genes.
Moreover, we found that the markers of goblet (Agr2, Fcgbp, and Spdef ) and enteroendocrine cell progenitors (Neurod1, Neurog3, Rfx6, and Foxa2) were downregulated in the absence of Tcf7l1. These results suggest that TCF7L1 is also required either for the generation of secretory progenitors or for the expression of transcription factors promoting differentiation along goblet and enteroendocrine lineages. Previous studies showed that TCF7L1 binds enhancers of Foxa2, Neurog3, Neurod1, and Spdef and represses their transcription [2,37]. Therefore, TCF7L1 is required for differentiation along secretory lineages. A simultaneous increase in enterocyte markers suggests that TCF7L1 may regulate Notch signalling activity. Interestingly, we found that Rbp-j is significantly upregulated in Tcf7l1 mutant epithelium (Supplementary Table S1). Rbp-J was also induced in Tcf7l1 mutant embryonic stem cells [38]. In addition, hydroxymethylglutaryl-CoA synthase (HMGCS2), the expression of which is elevated in Tcf7l1 mutants, promotes Notch signalling in the adult ISCs [39]. Finally, the activation of Olfm4, a target of Notch signalling [40], in Tcf7l1 mutants indicates increased Notch activity. Thus, our data suggest that TCF7L1 might balance Notch signalling in the embryonic intestinal epithelium and that the changes in secretory lineage markers could be secondary to enhanced Notch signalling.
In the adult intestinal epithelium, TCF7L1 promotes differentiation along the tuft cell lineage but does not affect goblet or Paneth cell differentiation. Given that Tcf7l1 is expressed mostly in tuft cells [11], this is an expected finding. Our data on tuft cells can be interpreted in two ways. First, Pou2f3 encoding for the transcription factor essential for the differentiation of the secretory progenitors along the tuft cell lineage [41] is one of the TCF7L1 target genes in ESCs [2]. Therefore, TCF7L1 may be required for the transcriptional activation of Pou3f2 in intestinal epithelial progenitors. This is unlikely, however, because TCF7L1 functions as a repressor. Second, the inhibition of WNT signalling leads to the increase in tuft cell numbers in the lung epithelium [42]. In contrast, upregulation of WNT signalling results in a loss of tuft cells. We propose that TCF7L1 promotes the differentiation of tuft cell progenitors by competing with TCF7L2 and reducing the transcriptional outcomes of WNT signalling.
Likewise, the inhibition of WNT signalling is crucial for the differentiation of enteroendocrine progenitors [43]. We found that the numbers of somatostatin-positive D-cells and GLP-1-positive L-cells were decreased in Tcf7l1 mutant anterior small intestines compared to controls. It is possible that the anterior D-and L-cells are more sensitive to the levels of WNT signalling and thus require TCF7L1-mediated repression. Indeed, a previous study showed that cells located at the +4 position from the crypt bottom do co-express both WNT target genes Lgr5 and Prom1 and enteroendocrine markers Chga, Cck, Gip, and Ghrelin [44], indicating that WNT signalling is active in enteroendocrine progenitors. However, we cannot exclude another possibility-that the action of TCF7L1 on the embryonic intestinal epithelium may change their chromatin landscapes, which in turn would affect later transcriptional programs in adult ISCs. Consistent with this model, we have observed that Tcf7l1 is expressed in the anterior part of the small intestine during embryogenesis. As a result, the transcriptional changes that we detected were specific to the anterior intestinal epithelial cells. Further elucidation of transcriptional programs in Tcf7l1 mutant adult ISCs at a single-cell level could validate this hypothesis.

Conclusions
Taken together, we have shown that TCF7L1 controls the differentiation of the embryonic gut epithelium. The discovery that TCF7L1 negatively regulates the expression of Rbp-J in intestinal epithelial cells provides an additional link between WNT and Notch pathways during embryogenesis, cell lineage specification, and neoplastic transformation. Finally, we propose that TCF7L1 acts to fine-tune levels of WNT-dependent transcription, which is essential for the differentiation of tuft cells in the adult gut.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cells12111452/s1, Figure S1: Transcriptional changes caused by loss of Tcf7l1 in the embryonic gut epithelium; Figure S2: TCF7L1 is dispensable for the proliferation of transit-amplifying cell and the differentiation of goblet and Paneth cells; Table S1: Genes upregulated in Tcf7l1 KO; Table S2: Genes downregulated in Tcf7l1 KO.