Lrig1 controls intestinal stem-cell homeostasis by negative regulation of ErbB signalling

Abstract

Maintenance of adult tissues is carried out by stem cells and is sustained throughout life in a highly ordered manner. Homeostasis within the stem-cell compartment is governed by positive- and negative-feedback regulation of instructive extrinsic and intrinsic signals. ErbB signalling is a prerequisite for maintenance of the intestinal epithelium following injury and tumour formation. As ErbB-family ligands and receptors are highly expressed within the stem-cell niche, we hypothesize that strong endogenous regulators must control the pathway in the stem-cell compartment. Here we show that Lrig1, a negative-feedback regulator of the ErbB receptor family, is highly expressed by intestinal stem cells and controls the size of the intestinal stem-cell niche by regulating the amplitude of growth-factor signalling. Intestinal stem-cell maintenance has so far been attributed to a combination of Wnt and Notch activation and Bmpr inhibition. Our findings reveal ErbB activation as a strong inductive signal for stem-cell proliferation. This has implications for our understanding of ErbB signalling in tissue development and maintenance and the progression of malignant disease.

Figure 1. Characterisation of Lrig1 expression in the intestine

a) In situ hybridisation for Lrig1 in adult mouse small intestine. Insert shows highest expression in the stem cell niche. b) Indirect immunofluorescence analysis for Lrig1 (magenta) and Lgr5-eGFP (green) in adult intestine of Lgr5^{−eGFP-IRES-CreERT2} mice. Of note Paneth cells do not express Lrig1, the visible membrane staining of ISCs originates from only one cell membrane, whereas in the progenitor compartment the staining is a composite of two neighbouring cell membranes. c) Analysis of Lrig1 expression in Lgr5-GFP^high/mid/low and GFP^neg populations from Lgr5^{−eGFP-IRES-CreERT2} mice. d) Lrig1 is expressed by approximately 30% of all crypt cells as determined by flow cytometry. e) Lgr5-GFP expressing cells express low to medium levels of CD24. f) All Lgr5-GFP expressing cells are Lrig1 positive, and both Lgr5-GFP and Lrig1 expressing cells fall in the CD24^low/mid range. g) Lrig1 negative cells are either CD24^high or negative and Lgr5-GFP negative. h) Relative expression analysis by qPCR for ISC markers, differentiation markers and Lrig1 interaction partners in Lrig1 expressing vs. Lrig1 non-expressing cells. Error bar represents the s.e.m. of four independent samples. nd – not detected. i) Summary of flow cytometric analysis for BrdU in Lrig1 expressing and non-expressing cells following one or several injections of BrdU (BrdU Inj) and analysis following a 1h, 3h, and 1w chase. Error bars represent s.d. from four independent samples. Scale bars: 100μm (a), 25μm (b).

Figure 2. Loss of Lrig1 causes crypt expansion

a-b) Grossly enlarged abdomen in postnatal day 10 Lrig1 KO animals. c-d) Lrig1 KO animals are smaller than control littermates but the length of the gut is not affected. Length of the gut and weight of neonates collected from P4 to P10. (P4:WT/Het n=7, KO n=3; P6:WT/Het n=15; KO:n=7; P8:WT/Het n=9, KO n=5; P10:WT/Het n=16, KO n=7). Error bars represent s.d. and asterisks significant changes (P6: p=0.0028; P8: p=0.011; P10: p=5.5×10⁻⁵). e-f) Loss of Lrig1 causes crypt expansion. H&E staining of proximal jejunum from Lrig1 WT and KO littermates. White lines demarcate crypt structures. g-h) The crypts from Lrig1 KO animals contain more proliferating cells. Phosphorylated histone H3(Ser10) (green) in intestinal samples from Lrig1 WT and KO littermates counterstained with dapi (blue). Scale bars 100μm (e-f), 50μm (f-g).

Figure 3. Loss of Lrig1 causes crypt and stem cell expansion

a) Dissection of epithelial lineages by flow cytometry based on the relative levels of CD24 and binding of UEA-I. The five identifiable population are enriched in A: Goblet cells (CD24^neg/UEA-I^pos), B: Paneth cells (CD24^pos/UEA-I^pos); C: transit-amplifying cells/enterocytes (CD24^neg/UEA-I^neg); D: stem cells and progenitors (CD24^low/mid/UEA-I^neg); E: enteroendocrine cells (CD24^high/UEA-I^neg). b) Loss of Lrig1 leads to a disproportionate increase in CD24^pos/UEA-I^pos Paneth cells and CD24^low/mid/UEA-I^neg stem cells and progenitors. Error bars represent s.d. from 4 control and 3 Lrig1 KO samples (CD24^neg/UEA-I^neg transit-amplifying cells/enterocytes: p=4.1×10⁻³; CD24^low/mid/UEA-I^neg stem cells and progenitors: p=5.3×10⁻⁴; CD24^pos/UEA-I^pos Paneth cells: p=4.7×10⁻⁴). c) Enrichment of stem cell markers in the CD24^low/mid/UEA-I^neg stem cells and progenitors upon loss of Lrig1. Error bars represent the s.e.m. from 3 Lrig1 KO and 4 control samples (Lgr5: p=2×10⁻⁴; Olfm4: p=0.02; Msi1: p=0.001). d) Loss of Lrig1 causes a progressive expansion of the stem cell niche from postnatal day 6 as detected by in situ hybridisation for Olfm4 and Cryptdin6, respectively. Scale bars 50μm.

Figure 4. Lrig1 controls endogenous signalling via the ErbB pathway

a) Loss of Lrig1 causes increased protein levels and activation of the ErbB pathway. pEgfr, Egfr, pErbB2, ErbB2, pErbB3, ErbB3, cMet were detected by Western blotting in samples enriched for intestinal epithelium. β-actin is used a loading control. b) Relative expression analysis of the receptors by qPCR at P10 shows minor differences. Expression levels are shown relative to control samples (KO/Ctrl). Asterisks indicate significant changes (ErbB2: p=0.004; ErbB3: p=0.04; KO n=4, Ctrl n=3). c-f) Increased activation of MAPK signalling and cMyc signalling upon loss of Lrig1. Immunohistochemical analysis for p-MEK1-2 (c-d) and cMyc (e-f). g-i) Altered Egfr activation dynamics upon loss of Lrig1 KO. i) Average normalised membrane intensity of pEgfr in intestinal samples from KO (n=6) and control animals (n=10) for 6-18 individual crypts per sample. Error bars represent the s.e.m. (positions 1-4: p<0.05). j) Increased Myc activity in progenitors lacking Lrig1. Relative expression of Myc and Myc target genes in CD24^low/mid/UEA-I^neg stem cells and progenitors versus CD24^neg/UEA-I^neg transit-amplifying cells/enterocytes from control (n=4) and Lrig1 KO (n=3) tissues. Error bar represent the s.e.m. (Myc: p=0.001; NSun2: p=0.001; Ncl: p=0.002). k) Lrig1 KO organoids mature in the absence of exogenous ErbB ligands. Maximum intensity projection of confocal images of WT and KO organoids shows incorporation of BrdU (green) and β-catenin as counterstain (magenta). l) EGF signalling is required for organoid maturation. The branching coefficient was determined for varying concentrations of EGF for independently derived samples (n=3 for all except for 50ng/mL and no EGF: n between 6 and 9). Error bars represent the s.e.m. and asterisk significant change (p=0.009). m) Loss of Lrig1 does not affect ErbB ligand expression. The relative levels of ErbB ligands were determined by qPCR on material from organoids grown for 2 and 6 days with or without EGF. Red and green colours reflect increased and decreased deviation from the mean, respectively. The dendogram indicates that ctrl samples grown for 6 days in normal conditions cluster with Lrig1 KO samples grown with and without EGF. Scale bars 50μm (c-h) and 100μm (a-d).

Figure 5. Lrig1 controls ErbB activation in vivo

a-d) Pharmacological inhibition of ErbB activation restores proliferation in the intestinal epithelium and the crypt size to normal in Lrig1 KO animals. Crypt size and proliferation is visualised by the expression of Ki67. e-l) Treatment with Gefitinib rescues the observed effect on the stem cell niche in Lrig1 KO animals. Paneth cells and stem cells are detected by in situ hybridisation for Cryptdin6 and Olfm4, respectively. m-p) pEgfr levels are reduced upon treatment with the ErbB inhibitor Gefitinib. Detection of activated Egfr (pEgfr) in tissues from Lrig1 WT and KO animals at P10 either untreated (m and n) and treated with Gefitinib (o and p). q-t) A loss of function Egfr mutant rescues the Lrig1 KO phenotype. Morphologically proliferation has been restored to normal levels in the rescued animals as detected by expression of Ki67 by immunohistochemistry. All Lrig1 KO mice on an Egfr^wt/wt background have the expected phenotype (r,15 out of 15), however, a large proportion of Lrig1 KO animals heterozygous for the hypomorphic Egfr allele (Egfr^wt/wa-2) have normal intestinal morphology (t, 15 out of 37; p=0.0095) although some still display hyperplasia (s). u) Model of the role of Lrig1 in stem cell homeostasis as a regulator of ErbB signalling. Scale bars: 50μm (a-t).