Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Feb 1;2(2):175-188.
doi: 10.1016/j.jcmgh.2015.12.004.

Foxl1-expressing mesenchymal cells constitute the intestinal stem cell niche

Reina Aoki  1 , Michal Shoshkes-Carmel  1 , Nan Gao  1 , Soona Shin  1 , Catherine L May  1 , Maria L Golson  1 , Adam M Zahm  1 , Michael Ray  2 , Caroline L Wiser  2 , Christopher V E Wright  2 , Klaus H Kaestner  1
Affiliations

Foxl1-expressing mesenchymal cells constitute the intestinal stem cell niche

Reina Aoki et al. Cell Mol Gastroenterol Hepatol. .

Abstract

Background & aims: Intestinal epithelial stem cells that express Lgr5 and/or Bmi1 continuously replicate and generate differentiated cells throughout life1. Previously, Paneth cells were suggested to constitute an epithelium-intrinsic niche that regulates the behavior of these stem cells2. However, ablating Paneth cells has no effect on maintenance of functional stem cells3-5. Here, we demonstrate definitively that a small subset of mesenchymal, subepithelial cells expressing the winged-helix transcription factor Foxl1 are a critical component of the intestinal stem cell niche.

Methods: We genetically ablated Foxl1+ mesenchymal cells in adult mice using two separate models by expressing either the human or simian diphtheria toxin receptor (DTR) under Foxl1 promoter control.

Conclusions: Killing Foxl1+ cells by diphtheria toxin administration led to an abrupt cessation of proliferation of both epithelial stem- and transit-amplifying progenitor-cell populations that was associated with a loss of active Wnt signaling to the intestinal epithelium. Therefore, Foxl1-expressing mesenchymal cells constitute the fundamental niche for intestinal stem cells.

Keywords: Intestinal stem cell niche; Wnt; mesenchyme.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Antibodies to the mouse Foxl1 protein are specific, and Foxl1 is expressed in a small subset of subepithelial fibroblasts in the intestine. (A) Immunoblot analysis for the Foxl1 protein in epithelial and mesenchymal fractions of mouse small intestine (jejunum) and colon, as indicated. Two different affinity-purified anti-Foxl1 antibodies were used. Note the absence of the Foxl1 protein band in mesenchymal fractions from Foxl1 null mice. Immunoblot analysis of α-smooth muscle actin (αSMA) confirms separation of epithelium and mesenchyme. mes, mesenchyme. (B) Immunostaining showing expression of Foxl1 (red) in the adult small intestine in a subset of mesenchymal cells located in close apposition to the epithelium (outlined with immunostaining for EpCAM, green). Nuclei are labeled with 4′,6-diamidino-2-phenylindole (DAPI) (blue). Scale bar: 50 μm. (C) Cross-section of crypt region showing the localization of Foxl1+ cells (red) in pericryptal mesenchymal cells. Nuclei are labeled with DAPI (blue). Scale bar: 25 μm.
Figure 2
Figure 2
Foxl1+cells have long processes and express potential niche-supporting factors. (A) Schema of Foxl1–Cre;Rosa–membrane-targeted dimer tomato protein membrane-targeted green fluorescent protein mouse used to drive expression of a membrane-bound green fluorescence protein (GFP) in Foxl1-Cre–positive cells. (B) Foxl1–Cre–driven expression of membrane-bound GFP is restricted to pericryptal mesenchymal cells, which elaborate long extensions into the villus tips and surrounding intestinal crypts. Immunofluorescence staining for GFP (green) and α-smooth muscle actin (αSMA) (red) shows very limited overlap between the 2 markers, indicating that Foxl1+ cells represent a unique cell population. αSMA expression is present in the core of the lamina propria, whereas GFP expression is located at the pericryptal sheath. (C) Immunofluorescence staining for Foxl1–Cre–driven GFP (green) and Myh11 (red) indicates very limited overlap between the 2 markers. (D) A histogram of fluorescence-activated cell sorting of intestinal mesenchymal cells isolated from C57BLl6 control and Foxl1–Cre;Rosa–YFP mice using YFP fluorescence. The x-axis shows YFP fluorescence intensity, and the gate indicated in the Foxl1–Cre;Rosa–YFP plot shows the sorted cells from which RNA was isolated for RNA sequence analysis. Note the absence of high-intensity YFP+ cells in the control histogram. (E) Relative mRNA levels of various markers and key signaling molecules in sorted Lgr5+-epithelial stem cells, differentiated villus epithelial cells, and sorted Foxl1+ cells.
Figure 3
Figure 3
Ablation of Foxl1+mesenchymal cells in diphtheria toxin–treated Foxl1–hDTR mice results in shortening of the gut tube. (A) Schema for the generation of Foxl1–hDTR mice using BAC recombineering. The coding sequence of exon 1 of Foxl1 was targeted by the sequence of the hDTR after destruction of the upstream FoxC2 coding sequence. FRT, flippase recognition target; LA, left homology arm; RA, right homology arm. (B) Expression of the human diphtheria toxin receptor (the HBEGF gene) is confined to the mesenchyme in Foxl1–hDTR mice. Mesenchyme and epithelium of the colon from Foxl1–hDTR mice were separated and cDNA was prepared. Reverse-transcription PCR analysis was performed to detect mesenchymal vimentin, epithelial E-cadherin, and ubiquitous glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcripts. Note that the HBEGF/hDTR mRNA is present exclusively in the mesenchyme. -RT, without reverse transcriptase; Mes, mesenchyme; Epi, epithelium; WT, whole colon from control nontransgenic mice. (C and D) Representative images of small intestine and colon from DT-treated Foxl1–hDTR and control mice. Note the shortening of the gut tube in the DT-treated Foxl1–hDTR mice.
Figure 4
Figure 4
Intestinal epithelial architecture and proliferation are dependent on Foxl1+mesenchymal cells. (A) Representative immunofluorescence images of human diphtheria toxin receptor (green) staining in the jejunum of control and Foxl1–hDTR mice, counterstained with 4′,6-diamidino-2-phenylindole (DAPI) to label nuclei (blue). (B and C) Intestinal morphology of the jejunum of (B) DT-treated control mice or (C) Foxl1–hDTR mice as assessed by H&E staining. Note the reduced number and length of intestinal villi in Foxl1–hDTR mice. (D) Quantification of small intestinal villus length (n = 4; in each animal > 20 villi were measured; ***P < .001). (E and F) Intestinal morphology of the proximal colon from (E) DT-treated control and (F) Foxl1–hDTR mice. Note the decreased depths of the colonic crypts and the irregular localization of epithelial nuclei in Foxl1–hDTR mice. (G) Quantification of colonic crypt depth (n = 4; in each animal > 20 crypts were measured; **P < .005). (H and I) Representative immunofluorescence images of the proliferation marker Ki67 (red) counterstained with DAPI to label nuclei (blue) in the jejunum of (H) control and (I) Foxl1–hDTR mice after toxin injection. (J) Quantification of the number of proliferating cells per crypt unit in jejunum of DT-treated control and Foxl1–hDTR mice (n = 20 crypts each). ***P < .001. (K and L) Ki67 labeling (red) overlaid with DAPI for nuclei (blue) in the colon of (K) DT-treated control and (L) Foxl1–hDTR mice. The number of Ki67-positive cells is reduced dramatically in Foxl1–hDTR mice. Scale bars: 50 μm. (M) Quantification of the number of proliferating cells per crypt unit in the proximal colon of DT-treated control and Foxl1–hDTR mice (n = 20 crypts each). ***P < .001.
Figure 5
Figure 5
Targeted ablation of Foxl1+mesenchymal cells using the iDTR model causes loss of epithelial proliferation. (A) Schema for the generation of Foxl1–iDTR mice. Foxl1–Cre mice were crossed to mice carrying a loxP-stop-loxP simian diphtheria toxin–receptor construct in the Rosa26 locus (RosaiDTR mice). (B and C) Small intestinal (jejunum) morphology was assessed by H&E staining of (B) DT-treated control and (C) Foxl1–iDTR mice. Note the reduced number and length of intestinal villi. (D) Quantification of small intestinal villus length (n = 4; in each animal > 20 villi were evaluated per mouse; ***P < .001). (E and F) Colonic morphology of (E) DT-treated control and (F) Foxl1–iDTR mice. Decreased depths of the colonic crypts is apparent in Foxl1–iDTR mice. (G) Quantification of colonic crypt depth (n = 4; in each animal >20 crypts were evaluated per mouse; *P < .05). (H and I) Representative immunofluorescence images of small intestinal (jejunum) sections stained for the proliferation marker Ki67 (red) and counterstained with 4′,6-diamidino-2-phenylindole (DAPI) to label nuclei (blue) of (H) control and (I) Foxl1–iDTR mice. The number of Ki67-positive epithelial cells is reduced dramatically in Foxl1–iDTR mice. (J) Quantification of the number of proliferating cells per crypt unit in the jejunum of DT-treated control and Foxl1–iDTR mice (n = 20 crypts each). ***P < .001. (K and L) Ki67 labeling (red) overlaid with DAPI to mark nuclei (blue) in the colon of (K) DT-treated control and (L) Foxl1–iDTR mice. The number of Ki67-positive cells is reduced dramatically in Foxl1–iDTR mice. (M) Quantification of the number of proliferating cells per crypt unit in the colon of DT-treated control and Foxl1–iDTR mice (n = 20 crypts each). ***P < .001. Scale bars: 100 μm in the small intestine, 50 μm in the colon.
Figure 6
Figure 6
Foxl1+mesenchymal cells are required for intestinal stem cell marker expression independent of Paneth cells. (A and B) Expression of the intestinal stem cell marker Olfm4 is reduced in (B) DT-treated Foxl1–hDTR mice compared with (A) nontransgenic controls as determined by in situ hybridization. Insets show crypts at high magnification. (C) Quantification of the percentage of small intestinal crypts that are positive for Olfm4 mRNA (>200 crypts were counted for both DT-treated control and Foxl1–hDTR mice). *P < .01. (D and E) Expression of lysozyme (red), a Paneth cell marker, is not affected by ablating Foxl1+ cells. Scale bar: 50 μm. (F) Quantification of lysozyme-positive cells per crypt unit (n = 3 per genotype; >20 crypts were evaluated per animal). DAPI, 4′,6-diamidino-2-phenylindole.
Figure 7
Figure 7
The frequency of mucin-secreting goblet cells is unaffected by ablation of Foxl1+cells. (A and B) Detection of goblet cells by immunostaining for Muc2 (mucin 2; green) and (C and D) Alcian Blue histochemistry (blue) indicates that the frequency of goblet cells in (B and D) DT-treated Foxl1–hDTR mice is not different from that present in (A and C) control mice. DAPI, 4′,6-diamidino-2-phenylindole.
Figure 8
Figure 8
Foxl1+mesenchymal cells provide essential Wnt ligands to the intestinal stem cell compartment. (A and B) Activation of the Wnt pathway in epithelial cells of the jejunum was analyzed by immunohistochemistry for ß-catenin. Reduced nuclear ß-catenin localization is apparent in (B) toxin-treated Foxl–hDTR mice compared with (A) toxin-treated controls. Yellow arrows point to nuclei positive for ß-catenin. Scale bars: 25 μm. (C and D) Immunofluorescence staining of the Wnt target Sox9 (red) shows decreased expression in the jejunum of (D) DT-treated Foxl1–hDTR mice compared with (C) nontransgenic controls. The epithelium is delineated with E-cadherin immunostaining (green). Scale bars: 50 μm. (E and F) Mesenchymal Wnt2b mRNA was detected by in situ hybridization (red) and nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (blue) Wnt2b mRNA is reduced dramatically in the mesenchyme of (F) DT-treated Foxl1–hDTR mice compared with (E) controls. Scale bars: 50 μm. (G and H) Mesenchymal Wnt4 and (I and J) Wnt5a mRNAs were detected by in situ hybridization (red) and nuclei were counterstained with DAPI (blue). Wnt4 and Wnt5a mRNA levels were reduced dramatically in the mesenchyme of (H and J) DT-treated Foxl1–hDTR mice compared with (G and I) controls.
Figure 9
Figure 9
Proposed model of Foxl1-positive subepithelial mesenchymal cells as the intestinal stem cellniche, providing essential Wnt signals, and possibly other signaling molecules, to the epithelium.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Barker N., van Es J.H., Kuipers J. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature. 2007;449:1003–1007. - PubMed
    1. Sangiorgi E., Capecchi M.R. Bmi1 is expressed in vivo in intestinal stem cells. Nat Genet. 2008;40:915–920. - PMC - PubMed
    1. Takeda N., Jain R., LeBoeuf M.R. Interconversion between intestinal stem cell populations in distinct niches. Science. 2011;334:1420–1424. - PMC - PubMed
    1. von Furstenberg R.J., Gulati A.S., Baxi A. Sorting mouse jejunal epithelial cells with CD24 yields a population with characteristics of intestinal stem cells. Am J Physiol Gastrointest Liver Physiol. 2011;300:G409–G417. - PMC - PubMed
    1. Powell A.E., Wang Y., Li Y. The pan-ErbB negative regulator Lrig1 is an intestinal stem cell marker that functions as a tumor suppressor. Cell. 2012;149:146–158. - PMC - PubMed

LinkOut - more resources