The Centrosomal Protein FGFR1OP Controls Myosin Function in Murine Intestinal Epithelial Cells

Summary

Recent advances in human genetics have shed light on the genetic factors contributing to inflammatory diseases, particularly Crohn's disease (CD), a prominent form of inflammatory bowel disease. Certain risk genes associated with CD directly influence cytokine biology and cell-specific communication networks. Current CD therapies primarily rely on anti-inflammatory drugs which are inconsistently effective, lacking strategies for promoting epithelial restoration and mucosal balance. To understand CD's underlying mechanisms, we investigated the link between CD and the FGFR1OP gene, which encodes a centrosome protein. FGFR1OP deletion in mouse intestinal epithelial cells disrupted crypt architecture, resulting in crypt loss, inflammation, and fatality. FGFR1OP insufficiency hindered epithelial resilience during colitis. FGFR1OP was crucial for preserving non-muscle myosin II activity, ensuring the integrity of the actomyosin cytoskeleton and crypt cell adhesion. This role of FGFR1OP suggests that its deficiency in genetically predisposed individuals may reduce epithelial renewal capacity, heightening susceptibility to inflammation and disease.

enhanced Table of Contents (eTOC) blurb

Trsan et al. show that deletion of the centrosomal protein FGFR1OP in mouse intestinal cells disrupts crypt architecture, impairing regeneration and causing inflammation. FGFR1OP is crucial for non-muscle myosin II activation, cytoskeleton integrity, and cell adhesion, with FGFR1OP defects likely causing Crohn's disease susceptibility by reducing intestinal epithelial renewal capacity.

Graphical Abstract

Introduction

The intestinal epithelium is a single-cell layer lining the lumen of the intestine and comprising a constellation of specialized cell types that functionally interconnect host physiological systems with extrinsic microbial communities, metabolites, and dietary factors¹. Intestinal epithelial cells are compartmentalized into villi and crypts. Villi are finger-like extensions lined by non-dividing, functionally specialized cells, such as enterocytes, Paneth cells, goblet cells, enteroendocrine cells and tuft cells. Crypts are epithelial invaginations at the bases of villi, which contain stem cells that renew the epithelium and Paneth cells that support stem cells². Stem cells differentiate into proliferating progenitors, known as transit-amplifying (TA) cells, that migrate up the crypts and generate mature cells that line the villi^3–5. Epithelial cells are major contributors to the maintenance of intestinal homeostasis by ensuring a physical barrier to gut lumen. Impaired epithelial barrier function or structure lead to intestinal permeability, increased inflammation, and susceptibility to immune disorders.

Crohn’s disease (CD) is caused by aberrant immune responses to environmental triggers that occur in genetically predisposed individuals⁶. CD severely disrupts normal functions and/or proportions of cells in intestinal crypts, including stem cell proliferation and differentiation during epithelial restitution⁷. Genome-wide association studies (GWAS) have identified a dozen CD risk loci, most of which encode molecules that sense microbial stimuli or activate the immune system in response to them, but the majority of CD-associated loci still remain uncharacterized. One such uncharacterized locus is a 100 kb region in high linkage disequilibrium (LD) spanning Fibroblast growth factor receptor 1 oncogene partner (FGFR1OP) and RNASET2. It was reported for CD risk in 2008 (lead variant: rs2301436, FGFR1OP intron)⁸ and later replicated with a stronger significance in the latest meta-analysis (P = 4.5 × 10⁻³¹; rs6939196, intergenic)⁹. Notably, this locus showed pleiotropic associations with many autoimmune and skin diseases, including autoimmune hypothyroidism (P = 3.1 × 10⁻⁴⁵; rs9366078, intergenic), Lichen planus (P = 2.4 × 10⁻²¹; rs35291016, FGFR1OP intron), and basal cell carcinoma (P = 9.1 × 10⁻¹⁹; rs138783046, FGFR1OP intron)¹⁰. These variants are closely correlated with each other and likely tagging the same causal variant(s) in the locus (Figure S1). However, extensive LD in the locus makes it challenging to pinpoint the causal gene(s) solely by genetics; thus the biology underpinning association of FGFR1OP with CD is currently unknown.

FGFR1OP (also known as FOP or centrosomal protein 43, Cep43) is a component of the centrosome^11–13, the microtubule organizing center (MTOC) that coordinates the mitotic spindle, cilia, and cytoskeleton, thereby coordinating cell division, motility, polarity, and adhesion. In vitro studies in immortalized cell lines variably implicated FGFR1OP in microtubule anchoring¹¹, cell cycle progression¹⁴, and ciliogenesis^12,15. An in vivo study of germline Fgfr1op^−/− mice confirmed a prominent function in ciliogenesis during development, leading to defects in skeletal, heart, renal, and lung cells, ultimately resulting in embryonic lethality¹⁶. The observed phenotype in Fgfr1op^−/− mice resembled the short-rib polydactyly syndrome, a prototype of human ciliopathy.

Given the genetic association of the FGFR1OP region with CD, we sought to address the yet unexplored role of FGFR1OP in intestinal homeostasis and functions. We found that induced deletion of Fgfr1op in the mouse intestinal epithelium caused crypt loss, inflammation, and ultimately death. Moreover, we were unable to establish viable cultures of 3D organoids and spheroids from Fgfr1op-deficient crypts in vitro. Fgfr1op haploinsufficiency reduced resistance of intestinal crypts to a colitis model. scRNA-seq analysis of crypts further revealed that Fgfr1op deficiency biased crypt cellular output from early progenitor to progenitor populations with an injury-induced signature. Despite the proposed role of FGFR1OP in cell division and microtubule organization, our structural and biochemical analyses revealed that FGFR1OP has a prominent role in maintaining crypt architecture through modulation of the actin cytoskeleton. FGFR1OP was specifically required for phosphorylation of non-muscle myosin II, which has a major role in actin cross-linking and its contractile properties. The compromised cytoskeleton structure led to significant detachment of crypt cells at the basolateral side, resulting in decreased intestinal barrier function. Overall, these data demonstrate a function of the centrosomal protein FGFR1OP in sustaining a functional actomyosin network in intestinal crypts and maintaining epithelial barrier integrity. Altered function of FGFR1OP in genetically predisposed individuals may increase susceptibility to CD by attenuating the resilience and restitution capacity of intestinal crypts to injuries.

Results

Deficiency of FGFR1OP in the intestinal epithelium disrupts crypts

Analysis of Fgfr1op mRNA in the mouse intestine by in situ hybridization revealed expression in intestinal crypts and villi of the small intestine (Figure 1A). Expression was particularly evident in the cells of the crypts. Analysis of protein expression by immunofluorescence (IF) and transmission electron microscopy (TEM) further showed that FGFR1OP was located in the centrosomes on the apical sides of epithelial cells of the crypts and villi, where it colocalized with gamma tubulin, as well as at the poles of dividing cells (Figures 1B, S2A and S2B). Thus, we hypothesized that FGFR1OP may have an important function in the physiology of the intestinal epithelium. To test this hypothesis, Fgfr1op^fl/fl mice were crossed with mice expressing tamoxifen (TAM)-inducible Cre recombinase under the control of the villin 1 promoter, Vil-CreER^T2 (further in text: Fgfr1op^cKO). Both Fgfr1op^cKO and control Fgfr1op^fl/fl mice were treated with TAM for 5 consecutive days (Figure 1C). FGFR1OP ablation was apparent in the crypts by day 6 post Cre induction and later in the enterocytes lining the intestinal villi, corroborating effective deletion of Fgfr1op (Figure S2C). The delayed abrogation of Fgfr1op expression in the villi may be explained by the time it takes for the cells, after the deletion of Fgfr1op in crypts, to migrate to the villi. Fgfr1op^cKO mice rapidly lost weight starting at day 9 post Cre induction, whereas Fgfr1op^fl/fl mice were unaffected (Figure 1D); all Fgfr1op^cKO mice succumbed between day 12 and day 15 post Cre induction (Figure 1E). Necropsy of Fgfr1op^cKO mice exposed distended bowels and increased content of watery stools (Figure S2D). Histological analysis documented a marked reduction of crypt numbers in Fgfr1op^cKO mice compared to controls by day 9 post Cre induction throughout the small intestine (Figures 1F and 1G) and colon (Figure S2E). The rapid decline in weight and overall health deterioration observed in mice may likely be attributed to increased gut permeability and changes in the function of epithelial cells, which stem from the loss of epithelium.

Figure 1. Induced Fgfr1op deletion disrupts intestinal crypts.

(A) In situ hybridization of small intestinal tissues with a probe for Fgfr1op mRNA. Tissues are counterstained with anti-Epcam to visualize the intestinal epithelium. The image is representative of 6 mice examined. Scale bar 50 μm.

(B) Representative confocal images of FGFR1OP expression in the crypt and villus. Sections were counterstained with anti-α-tubulin. Red arrows, FGFR1OP expression in the apical side of epithelial cells in crypts and villi; white arrows, FGFR1OP staining in the microtubule-organizing center in dividing cells. Images are representative of 6 mice examined. Scale bar 10 μm.

(C-E) Schematic of the experimental time course: Cre was induced by intraperitoneal (i.p.) injection of TAM for 5 consecutive days prior to analysis on days 6, 9, and 12 post Cre induction (C). Weight loss (n=10 mice, mean±SD, Multiple Mann-Whitney test with Holm-Šídák correction; ****P<0.0001, **P<0.01) (D) and survival (E) upon TAM-induced Fgfr1op deletion [Log-rank (Mantel-Cox) test]; data is representative of two independent experiments.

(F-G) H&E images of small intestine on days 6, 9, and 12 post Cre induction (F). Images are representative of 3 independent experiments; Scale bar 100 μm. Crypt numbers in duodenum, jejunum, and ileum (n=4–9 mice per genotype and time point; mean±SD, one-way ANOVA with Tukey’s correction; *P<0.05, **P<0.01, ****P<0.0001) (G). Due to the disappearance of well-oriented crypts as part of the Fgfr1op^cKO phenotype, all crypts in the vision field were counted regardless of their appearance and size in both, Fgfr1op^cKO and control mice.

In addition to Fgfr1op^cKO mice, we also generated Fgfr1op^fl/fl x Vil1-cre 1000Gum/J mice (hereafter: Fgfr1op-Vil-Cre), which have a constitutive deletion of Fgfr1op in the intestinal epithelium. After birth, Fgfr1op-Vil-Cre mice were significantly smaller in size than Fgfr1op^fl/fl mice, but progressively grew and reached comparable weight at 10 weeks of age (Figures S3A–S3C). Fgfr1op-Vil-Cre mice presented altered intestinal architecture with reduced crypt numbers in the small and large intestine throughout their lives (Figure S3D). Further, adults (≥10 weeks of age) were more susceptible to epithelial injury induced by dextran sulfate sodium (DSS) than controls, as assessed by weight loss, survival, blood in stool, colon shortening, and histopathology (Figures S3E–S3I). Thus, Fgfr1op-Vil-Cre mice corroborated the importance of Fgfr1op for the integrity of intestinal crypts. However, intestinal epithelial cells escaping Vil-Cre-induced deletion of Fgfr1op partially replaced Fgfr1op-deficient cells over time (Figure S3J), attenuating the genetic defect and the severity of the pathology in comparison with Fgfr1op^cKO mice in which Fgfr1op deletion was acutely induced. A similar replacement by escapee has been previously reported in Vil-Cre x Polr3b^fl/fl mice¹⁷. Furthermore, compensatory mechanisms for centriole defects have been noted in intestinal development¹⁸, which could aid in the survival of Fgfr1op-Vil-Cre mice during early development and allow adequate time for intestines to repopulate with cells evading recombination.

Lack of Fgfr1op disrupts the regenerative potential of intestinal crypts and induces inflammation

Renewal of intestinal crypts relies on functional stem cells in the base of the crypt^3,19. Therefore, we sought to first examine the impact of FGFR1OP deficiency on the maintenance of crypt stem cells. IF analysis for the stem cell marker OLFM4²⁰ showed that stem cells were present at day 6 post Cre induction followed by a marked reduction by day 9 in Fgfr1op^cKO mice (Figures 2A and 2B). Next, we examined whether FGFR1OP deficiency impacted cell proliferation. BrdU incorporation, which occurs during the S phase of the cell cycle, and IF for the mitosis marker phospho-histone H3 (PHH3) corroborated a depletion of proliferating cells in Fgfr1op^cKO mice on days 9 and 12 post Cre induction (Figures 2C–2F). Conversely, IF for the Paneth cell marker lysozyme showed Paneth cell persistence in degenerating crypts (Figure S4A). IF analysis of small intestinal tissue specimens also showed the infiltration of cells expressing myeloid markers IBA1 and S100a9 in intestinal crypts of Fgfr1op^cKO at day 9, suggesting inflammation²¹ (Figure 2G). Flow cytometry of intraepithelial and intestinal lamina propria leukocytes confirmed infiltration of monocytes, macrophages, dendritic cells, and granulocytes by day 9 post Cre induction in Fgfr1op^cKO mice but not in controls (Figures 2H and S4B).

Figure 2. Fgfr1op deletion affects proliferating intestinal epithelial cells and associates with myeloid cell infiltration.

(A-F) Mice were treated with TAM and analyzed on days 6, 9, and 12 post Cre induction as in Figure 1C. Representative images (A) and numbers (B) of OLFM4⁺ cells; sections were counterstained with anti-β-Catenin. Representative images (C) and quantification of BrdU incorporation (D). BrdU was injected i.p. 2h before sacrificing; sections were counterstained with anti-EpCAM. Representative images (E) and numbers (F) of PHH3-expressing cells; sections were counterstained with anti-β-Catenin.

(A-F) data are representative of two independent experiments and n=4 mice per genotype and time point; (B,D,F) mean±SD is shown; one-way ANOVA with Tukey’s correction; ns=not significant, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; (A,C,E) Scale bar 50 μm.

(G) Immunofluorescence images of IBA1⁺ myeloid cells and S100a9⁺ neutrophils infiltrating the intestine upon Fgfr1op deletion (representative of two independent experiments, n=4 mice per genotype and time point; sections are counterstained with anti-β-Catenin). Scale bar 50 μm.

(H) Numbers of myeloid cell subsets within the epithelium (IEL) and lamina propria (LP) at day 6 and 9 post Cre induction (n=3–4 mice; mean±SD, one-way ANOVA with Tukey’s correction, *P<0.05, **P<0.01, ****P<0.0001; representative of two independent experiments). See also Figure S4B.

Microarray analysis of mRNAs from whole small intestinal tissue specimens of Fgfr1op^cKO compared to control mice at days 6, 9 and 12 after Cre induction chronicled progressively reduced expression of genes indicative of intestinal epithelial cell types, such as stem cells (e.g. Olfm4) and enterocytes (e.g. Slc5a11, Spink1), while the expression of genes indicating epithelial response to injuries (e.g. Anxa10, Il1rn) and myeloid cell infiltration (e.g. Slc40a1) increased (Figure S4C). Inspection for gene signatures specific for disparate epithelial cell types²² corroborated a depletion of stem cells and enterocytes in Fgfr1op^cKO mice (Figures S4D and S4E), while a gene set indicative of TA cells increased for a brief period following Cre induction in Fgfr1op^cKO mice and then declined to control levels or below. This observation suggests an initial compensatory response of TA cells to stem cell depletion, akin to TA cells actively dividing in response to γ-irradiation to facilitate regeneration^1,23–25. However, this response eventually diminishes over time. RNA sequencing (RNA-seq) of crypts isolated from the small intestine of Fgfr1op^cKO and control mice on days 6 and 9 post Cre induction corroborated reduced expression of stem cell signature genes, paralleled by increased expression of TA cell signature genes²² (Figures S4F and S4G). Together, phenotypic and transcriptional profiling of the intestine suggest that a lack of Fgfr1op dramatically disrupts intestinal crypts, affecting the representation of stem cells, TA cells, and enterocytes as well as inducing inflammation.

FGFR1OP-deficient crypts are unable to generate 3D cultures

We further examined whether Fgfr1op deletion affected the ability of intestinal stem cells to form 3D cultures in vitro. We used Fgfr1op^cKO and Fgfr1op^fl/fl intestinal crypts to generate organoid cultures, which consist of both stem cells and differentiated cells^26,27 (Figure 3A). Organoids were treated with TAM for 3 days, followed by culturing in medium without TAM and passaging. Fgfr1op^cKO crypts yielded significantly less branched organoids after Cre induction in vitro, whereas organoids from Fgfr1op^fl/fl crypts readily formed (Figure 3A). Moreover, Fgfr1op^cKO-derived organoid structures persisting after Cre induction could not be maintained after passaging. Similarly, we were unable to maintain spheroid lines²⁸ derived from Fgfr1op^cKO crypts treated in vitro with TAM or derived from Fgfr1op^cKO mice on day 6 or 9 post Cre induction in vivo (Figures 3B and 3C). To conclusively demonstrate that Fgfr1op deficiency prevents the formation of stable 3D cultures from crypt stem cells, we crossed Fgfr1op^fl/fl with Lgr5-EGFP-IRES-creERT2 and Rosa-LSL-tdTomato mice, such that, after administration of TAM, Lgr5⁺ (EGFP⁺) stem cells that have undergone Fgfr1op deletion also expressed tdTomato, enabling isolation of enriched Fgfr1op-deficient stem cells. Lgr5-EGFP-IRES-creERT2 x Rosa-LSL-tdTomato mice were used as controls. We sorted and plated 10,000 live Lgr5⁺ tdTomato⁺ stem cells from both, Fgfr1op-deficient and control mice. Unlike control Lgr5⁺ tdTomato⁺ stem cells, the Fgfr1op-deficient ones failed to generate viable spheroids in vitro (Figure 3D).

Figure 3. Impact of Fgfr1op deletion on 3D cultures of stem cells.

(A) Organoid cultures were established from crypts isolated from untreated mice, followed by TAM treatment in vitro for 3 days and passaging on day 6. Images were taken on day 5 for both passages. Images are representative of 3 independent experiments. Scale bars 200 μm. The bar graph shows average numbers of crypt-like domains per organoid (n=30–41 organoids per mouse and n=4 mice per genotype; mean±SD, one-way ANOVA with Tukey’s correction, ****P<0.0001).

(B) Spheroid cultures were established from crypts isolated on day 6 and 9 after Cre induction in vivo (as in Figure 1C). P1, P2 – passage 1 and 2. Images were taken on day 3 post passaging. Scale bars 300 μm. Data are representative of two independent experiments.

(C) Spheroid cultures were established from crypts isolated from untreated Fgfr1op^fl/fl and Fgfr1op^cKO mice and then treated with TAM in vitro for 3 days. P0 – day 3 before passaging, P1 – day 3 after passaging. Scale bars 200 μm. Data are representative of three independent experiments.

(D) Fgfr1op deletion was induced with a single injection of TAM in Fgfr1op^fl/fl-Lgr5-EGFP-creERT2-tdTom or Lgr5-EGFP-creERT2-tdTom control mice. LGR5⁺ tdTomato⁺ cells were sorted 7 days later and 10,000 cells were plated in Matrigel and cultured as spheroids for 9 days, followed by passaging. Flow cytometry plots show the gating strategy to isolate EGFP⁺ stem cells in which Cre is expressed and deletes the LoxP-flanked stop cassette allowing tdTomato expression. Spheroid images were taken on day 9 post sorting (P1) and on day 3 post passaging (P2). Data are representative of 3 mice examined. Scale bar 200 μm.

(E) RNA for RNA-seq analysis was isolated from spheroid cultures derived from Fgfr1op^fl/fl and Fgfr1op^cKO mice and treated with TAM for 3 days in vitro. Pathway analysis (Metascape⁵⁵) comparing mRNA profiles of Fgfr1op^cKO and Fgfr1op^fl/fl spheroids (n=3 mice, padj<0.05, log₂FC=−1.2 – 1.2) and representative DEGs associated with actin cytoskeleton, stem cell regeneration, and migration / adhesion are shown. Pink and blue colors denote upregulated and downregulated genes, respectively.

As the Fgfr1op^cKO-derived spheroids seemed intact during the 3-day culture with TAM but showed no growth upon passaging, we collected spheroids from both Fgfr1op^cKO and Fgfr1op^fl/fl for RNA-seq analysis on day 3 of TAM treatment. RNA-seq indicated that deletion of Fgfr1op broadly affects proliferation, morphogenesis, and actin cytoskeleton organization (Figure 3E). In parallel, Fgfr1op^cKO spheroids were enriched in fetal-like signature genes (e.g. Ly6c1, Ly6a, Ly6g6c, Clu, and Anxa1) characteristic of epithelial regeneration following stem cell loss^25,29–32 and consistent with an ongoing repair attempt. We conclude that Fgfr1op deletion impairs the ability of crypt stem cells to generate stable 3D structures in vitro, recapitulating the impaired crypt ability of renewing the intestinal epithelium in vivo.

Haploinsufficiency of Fgfr1op impairs crypts resilience to injury

To test whether reduced expression of Fgfr1op was sufficient to impact intestinal crypts, we examined mice with a deletion of a single copy of the wild-type gene. Crypts of Fgfr1op^fl/+-Vil-Cre-ERT2 (further in text: Fgfr1op^fl/+) mice showed intermediate FGFR1OP expression between those of Fgfr1op^cKO and Fgfr1op^fl/fl crypts on day 6 post Cre induction, corroborating Fgfr1op haploinsufficiency (Figure 4A). Fgfr1op^fl/+ mice did not show obvious defects in intestinal architecture or stem cell marker expression compared to Fgfr1op^fl/fl mice, at least up to day 12 post Cre induction (Figures 4B and 4C). However, crypts with Fgfr1op haploinsufficiency showed reduced capability to generate organoids compared to control crypts (Figure 4D), indicating the presence of a subtle defect in stem cells. This defect might be revealed under conditions that challenge their resilience and renewal capacity, such as aging or tissue damage. To test this hypothesis, we opted to utilize the DSS-induced colitis model, known for inducing epithelial depletion in the colon and commonly employed in studying inflammatory bowel disease in mice. Since DSS primarily affects the colonic epithelium, we initially confirmed a consistent apical FGFR1OP expression pattern in colon crypts (Figure S2A). Moreover, we corroborated comparable histopathology regarding crypt loss between the small intestine and colon of Fgfr1op-deficient mice upon Cre induction in steady state (Figures S2E, S3D, S3J). After DSS-induced injury, Fgfr1op haploinsufficiency reduced epithelial restitution (Figures 4E–4J). While Fgfr1op^cKO mice succumbed to DSS-induced colitis by day 7, Fgfr1op^fl/+ mice survived as did Fgfr1op^fl/fl mice but showed more weight loss and shorter colons (Figures 4E–4G). Moreover, while epithelial damage at day 7 was similar in Fgfr1op^fl/+ and Fgfr1op^fl/fl mice, recovery at day 21 was impaired in Fgfr1op^fl/+ mice, which exhibited persistent open ulcers and lack of wound associated epithelia (Figures 4H–4J). Thus, attenuated expression of FGFR1OP reduces the ability of intestinal crypts to regenerate after injury.

Figure 4. Fgfr1op^+/− mice have impaired capacity for epithelial regeneration.

(A) Crypts were isolated on day 6 post Cre induction (as in Figure 1C) and analyzed for FGFR1OP expression by immunoblotting. Anti-GAPDH antibody was used as a loading control. Blot is representative of 2 independent experiments and n=5–6 mice per genotype.

(B) H&E images of small intestine on days 6, 9, and 12 post Cre induction. Images are representative of 3 independent experiments and n=6–9 per genotype and time point. Scale bar 100 μm.

(C) OLFM4⁺ staining on day 12 post Cre induction. Scale bar 50 μm. Graph shows number of OLFM4⁺ cells per crypt (mean±SD, Mann-Whitney U test, ns=not significant). Data are representative of two independent experiments and n=4 mice per genotype.

(D) Organoid cultures were established from crypts isolated from untreated mice, followed by TAM treatment in vitro for 3 days. Images were taken on day 5. Scale bar 200 μm. Images are representative of 3 independent experiments. Graph shows average numbers of crypt-like domains per organoid (n=30–40 organoids per mouse and n=4 mice per genotype; mean±SD, one-way ANOVA with Tukey’s correction, ***P<0.001, ****P<0.0001).

(E-G) Female mice were treated with 2.5% DSS in water for 7 days (E). Weight loss (n=15–18 mice per genotype; pool of two independent experiments; Two-way ANOVA; mean ± SD; *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001) (F). Representative images of colons on day 7 (G).

(H-J) Female mice were treated with 2.5% DSS in water for 7 days, followed by 14 days of normal water (H). Weight loss (n=7–8 mice per genotype; Multiple Mann-Whitney test with Holm-Šídák correction) (I). Representative H&E images on day 7 and 21 (J). Scale bar 100 μm.

Lack of FGFR1OP diverges crypt output from normal to injury-induced progenitors

We sought to gain a granular view of the impact of Fgfr1op deletion on intestinal crypt output at early time points, before crypts were entirely disrupted. Thus, we performed scRNA-seq of Fgfr1op^cKO and Fgfr1op^fl/fl crypts on day 6 post Cre induction (Figures 5A–5K and S5). Sequenced cells projected onto a UMAP clustered into 15 populations encompassing stem-TA cells, progenitors and differentiated cells (Figures 5A, 5B, S5A and S5B). The expression of Fgfr1op was the highest in the stem-TA cluster (Figure 5C). The numbers of cells in the stem-TA cluster from Fgfr1op^cKO and Fgfr1op^fl/fl crypts were comparable (Figure 5D). However, Fgfr1op deletion reduced the stem cell signature and increased the TA cell signature²², consistent with microarray and RNA-seq data (Figures S5E and S5F, S4D, S4E and S4G). Deletion of Fgfr1op was followed by upregulation of injury signature genes (Anxa2, Klf6, S100a6, S100a11, Apoa1, Areg, Tnfrsf12a, and Krt19) as well as Prc1, a homeostatic gene that promotes stem cell renewal^{25,29,31,33–35} (Figure 5E and 5F). Gene expression changes caused by Fgfr1op deletion in the stem–TA cluster also indicated anti-inflammatory and stress responses (Rnf186, Nfe2l2, Hbegf, Nlrp6, and Ccl25). Furthermore, differentially expressed genes (DEGs) included genes associated with actin cytoskeleton and adhesion, keratins that form intermediate filaments in gut epithelial cells (Krt8, Krt18, and Krt20), and genes involved in apoptosis. To gain insight into changes associated with cell cycle genes, we overlayed signatures for G1/S and G2/M phases (ref³⁶ and Table S1) with stem-TA cluster genes (Figure S5G). We did not observe an obvious defect in the cell cycle progression at least at day 6 post Cre induction. In addition to inducing changes in the stem-TA cluster, Fgfr1op deletion also markedly reduced the abundance of enterocyte and secretory progenitor clusters (cluster 2 and cluster 11), paralleled by the appearance of injury-induced enterocyte progenitors and mature enterocytes (cluster 0 and cluster 6) (Figures 5A, 5B, 5D, 5E, S5B and S5D).

Figure 5. scRNA-seq analysis reveals reduction of progenitors and appearance of injury-induced cell populations in Fgfr1op-deficient crypts.

(A-F) Crypt cells for scRNA-seq analysis were isolated from Fgfr1op^fl/fl and Fgfr1op^cKO mice on day 6 post Cre induction (n=3 mice per genotype). (A) UMAP representing all the identified clusters. (B) UMAPs representing cluster distributions within each genotype. (C) Dotplot of Fgfr1op expression across clusters. (D) Proportion of all cells for designated clusters, and Fgfr1op^fl/fl and Fgfr1op^cKO genotypes. (E) Heatmap showing DEGs associated with injury and proliferation / apoptosis pathways (Metascape⁵⁵) between the enterocyte progenitors in Fgfr1op^fl/fl mice and the injury-induced enterocyte progenitors in Fgfr1op^cKO mice. (F) Heatmap representing DEGs associated with cytoskeleton, injury, and proliferation / apoptosis pathways (Metascape⁵⁵) in the Stem-TA cluster.

(G-K) Re-clustering of cells from clusters 0, 2, 4, 5 and 12 of Figure 5A. (G) UMAPs representing all clusters identified; TA – transit-amplifying, SP - secretory progenitors, GC - Goblet cells, EP – enterocyte progenitors, IIEP – injury-induced enterocyte progenitors. (H) UMAPs showing how clusters are distributed within each genotype category. (I) Breakdown of genotype distribution within each cluster. (J) Dotplot representing expression of Mki67, Lgr5, Olfm4 and Fgfr1op for each individual cluster. (K) Velocity stream as computed by the dynamic model of scvelo.

(L-M) UMAP of integrated scRNA-seq analysis of human intestinal epithelium in CD, UC and control subjects^7,37 (L) and dotplot of expression of FGFR1OP transcripts in integrated epithelial scRNA-seq atlas categorized by disease status of subjects (M).

To further differentiate how Fgfr1op deficiency affects stem cells and TA cells, we re-clustered stem-TA cells (cluster 4), enterocyte progenitors (cluster 2), injury-induced enterocyte progenitors (cluster 0), secretory progenitors (cluster 5), and goblet cells (cluster 12). This re-clustering process enabled the separation of stem cells from TA cells, which were further categorized into cycling TA-1 and TA-2 cells (Figures 5G and 5H). Our analysis confirmed the loss of expression of stem cell signature markers in Fgfr1op^cKO mice, along with an increase in the TA-2 signature in these mice (Figures 5I and 5J). Moreover, it revealed a clear distinction between enterocyte progenitors in Fgfr1op^fl/fl mice and injury-induced progenitors in Fgfr1op^cKO mice. Notably, Fgfr1op deletion did not adversely affect the proliferation of either stem or TA cells at this time point (Figure 5J and S5H). Furthermore, RNA velocity analysis of the re-clustered data demonstrated a clear directionality of TA-2 cells toward enterocyte progenitors, a trend consistent across both Fgfr1op^fl/fl and Fgfr1op^cKO genotypes (Figure 5K). In summary, the scRNA-seq data support the notion that early Fgfr1op deletion alters the output of stem cells in intestinal crypts and reduces the differentiation of normal progenitors, while simultaneously increasing the generation of cells expressing genes indicative of intestinal injury response.

Human FGFR1OP-expressing intestinal crypt cells harbor pathways associated with increased susceptibility to CD

To examine epithelial cell-specific expression pattern of FGFR1OP in human inflammatory bowel disease, we performed integrative analysis of published scRNA-seq datasets on CD⁷ and Ulcerative Colitis (UC)³⁷ by subsetting the dataset to annotated epithelial cell types. Briefly, genes expressed in > 10 cells were considered in the combined dataset. As shown in Figures 5L and 5M, FGFR1OP expression in epithelial cells was enriched in cycling TA1, TA2 and secretory TA cells. FGFR1OP was not expressed in other mature intestinal epithelial cells. Overall, combined single cell expression patterns, human genetics and detailed functional studies identify a previously underappreciated intestinal crypt compartment that harbors pathways associated with increased susceptibility to CD.

FGFR1OP is required for cell–cell adhesion in the crypts

Given that FGFR1OP deletion affected the cellular composition of intestinal crypts prior to a detectable proliferation defect, we examined whether crypt architecture was also affected at day 6 post Cre induction. Analysis of small intestinal tissue specimens from Fgfr1op^cKO mice on day 6 post Cre induction by TEM showed widening of cell-cell contacts at the basolateral sides of the crypt cells which progressed by day 9 post Cre induction (Figure S6A). This phenotype was accompanied by increased intestinal permeability, as measured by serum levels of FITC-dextran after oral administration (Figure S6B). Further, we noticed fewer intermediate filaments associated with desmosomes of Fgfr1op^cKO mice on day 6 post Cre induction (Figure S6C), particularly at the basolateral sides of the cells (Figure 6A–6C). Conversely, apical side of cells and intermediate filaments associated with apical desmosomes seemed unaffected (Figures S6D–S6F). Typical centrioles with 9 microtubule triplets arranged in a hollow cylinder were evident in both Fgfr1op^cKO and control crypt cells (Figure S6G), indicating that FGFR1OP deficiency did not affect the basic centrioles structure. Thus, these data suggest that FGFR1OP may be required for preserving crypt architecture beyond regulating centriols and stem cell division and imply that FGFR1OP has a role in regulating cell-cell adhesion in intestinal crypts^38,39.

Figure 6. Deletion of FGFR1OP in Caco2 cells recapitulates impaired cytoskeleton phenotype in vivo.

(A) Intestinal crypts were analyzed by TEM on day 6 after Cre induction. Basolateral desmosomes are shown. Arrows indicate intermediate filaments associated with desmosomes. Scale bar 100 μm. Data are representative of n=4 mice per genotype.

(B) Drawing of desmosome complex with intermediate filaments depicting width and length of desmosome, and intermembrane space. Created with BioRender.com.

(C) Bar graphs show width and length measurements (as defined in B) of crypt basolateral desmosomes analyzed by TEM on day 6 after Cre induction (points represent individual desmosomes (n=5–8 desmosomes per mouse) in n=4 mice per genotype; mean±SD; Mann-Whithey U test, ****P<0.0001).

(D) An outline of the experimental design for (E-H) is shown. Caco2 cells stably transduced with FGFR1OP shRNA or control shRNA were treated with 2 μg/ml doxycycline for 3 days prior the analysis. Representative images of ZsGreen expression upon shRNA induction in both cell lines are shown at the bottom of the scheme.

(E-F) TEM analysis of shRNA-expressing Caco2 cell lines. (E) Representative images. Desmosome structures are indicated by red arrows. Images are representative of three independent experiments. Scale bar 500 nm. (F) Bar graphs show desmosome complex measurements (as defined in B); points represent individual desmosomes (n=20–35); mean±SD; one-way ANOVA with Tukey’s correction, ****P<0.0001).

(G-H) Wound healing assay for untreated and doxycycline-treated FGFR1OP or control shRNA-expressing Caco2 cells. (G) Representative images. Scale bar 500 μM. (H) Bar graph representation of individual wound assays with percentage of closed wound area (n=3–4; one-way ANOVA with Tukey’s correction; mean ± SD; ns=not significant, *P<0.05, **P<0.01) (H). Data are representative of three independent experiments.

To demonstrate causality between FGFR1OP deficiency and altered cell–cell adhesion, we turned to the human colorectal adenocarcinoma cell line Caco2, since the short life span of organoids upon Cre induction posed technical limitations for structural studies. We generated a Caco2 cell line stably transfected with FGFR1OP shRNA, which enabled doxycycline-inducible silencing of FGFR1OP in vitro (Figures 6D and S7A). TEM analysis showed the formation of desmosomes with reduced intermediate filaments in FGFR1OP-silenced cells, reiterating the phenotype observed in primary crypts of Fgfr1op^cKO mice (Figure 6E and 6F). Conversely, FGFR1OP silencing did not impair cell proliferation (Figure S7B). FGFR1OP silencing also impaired the capacity of Caco2 cells for two-dimensional cell migration, as demonstrated by their reduced ability to fill the gap in a wound healing assay (Figures 6G and 6H). Altogether, these data demonstrate a requirement of FGFR1OP for the maintenance of the cell cytoskeleton and cell–cell adhesion in intestinal crypts and in vitro.

FGFR1OP is required for phosphorylation of non-muscle myosin II

To determine the precise mechanism of FGFR1OP-mediated regulation of cell-cell adhesion, N-terminal and C-terminal FLAG-tagged FGFR1OP proteins were expressed in Caco2 cells, followed by immunoprecipitation with anti-FLAG antibody and mass spectrometry analysis of the immunoprecipitates. This assay identified 13 proteins specifically interacting with N-terminal and/or C-terminal FLAG-tagged FGFR1OP (Figure 7A and Table S2), including the centrosomal protein CEP350 that was previously reported to associate with FGFR1OP¹⁵. FGFR1OP bound to several proteins of the cytoskeleton: the actin-binding protein DBN1⁴⁰, which regulates assembly of cilia; unconventional myosins (MYO1B, MYO1C, and MYO1D) as well as the light (MYL6) and heavy (MYH9) chains of non-muscle myosin II (NMII), which modify actin cytoskeleton through their motor activity⁴¹; and the cytoskeleton-associated protein LIMA1, which inhibits depolymerization of actin filaments and cross-links filaments in bundles⁴². Pathway analysis of FGFR1OP-interacting proteins corroborated a robust signature in the organization of the actin cytoskeleton (Figure 7B). Modulation of the actin cytoskeleton in control Caco2 cells with Latrunculin A provoked a depletion of intermediate filaments associated with desmosomes similar to that induced by FGFR1OP silencing (Figures S7C and 7D), reiterating the link between actomyosin contractility and the integrity of cell–cell junctions. These results suggest that FGFR1OP coordinates the actin cytoskeleton.

Figure 7. FGFR1OP sustains colocalization of myosin with actin filaments through NMII activation.

(A-B) FGFR1OP tagged with C- or N-terminal FLAG was overexpressed in Caco2 cells, and a pull-down assay was performed with anti-FLAG antibody (A). Proteins selectively identified in Caco2 cells expressing C- or N-FLAG-tagged FGFR1OP are shown in orange. Proteins identified in parental Caco2 cells or proteins with low total spectrum counts (cutoff: <7 counts) in all Caco2 lines are shown in grey. Data represent the pool of two independent experiments. (B) Pathway analysis (Metascape⁵⁵) of identified protein interactors.

(C-D) Confocal images of myosin and actin distribution in the crypts on day 6 post Cre induction (C); images are representative of two independent experiments, n=4 mice per genotype. Scale bar 10 μm. (D) Manders’s overlap coefficient calculated for actin and myosin staining. Dots represent coefficients for individual cells from n=3 crypts per genotype (mean±SD, Mann-Whitney U test; ****P<0.0001).

(E) Crypts were isolated on day 6 post Cre induction and analyzed for FGFR1OP, NMII regulatory light chain (RLC), and phosphorylated NMII regulatory light chain (p-RLC) expression by immunoblotting. Blot is representative of 2 independent experiments and n=8 mice per genotype.

(F) Caco2 cells stably transduced with FGFR1OP or control shRNA were treated with 2 μL/ml doxycycline for 3 days and analyzed by immunoblotting. Blot is representative of 2 independent experiments and n=6 replicates per genotype.

(E-F) Anti-GAPDH and anti-β-actin antibodies were used as loading controls.

(G-H) TEM analysis of Caco2 cells stably expressing either FGFR1OP shRNA or control shRNA, and either NMII RLC or constitutively active NMII RLC (see also Figure S7E). (G) Representative images. Desmosome structures are indicated by red arrows. Scale bar 500 nm. Images are representative of two independent experiments and n=4 replicates per genotype and condition. (H) Bar graphs show desmosome complex measurements (as defined in Figure 6B); points represent individual desmosomes (n=24–35 per condition); mean±SD; one-way ANOVA with Tukey’s correction, ***P<0.001, ****P<0.0001, ns=not significant).

Co-precipitation of FGFR1OP with MYL6 and MYH9 suggested that FGFR1OP may specifically impact contractility of the actin cytoskeleton through actin-binding NMII. Both, the actin cytoskeleton and NMII are involved in processes that modify cell architecture or movement, such as cell adhesion and migration. NMII cross-links the actin cytoskeleton and is a common endpoint of mechanochemical signaling pathways that control cellular responses to physical and chemical extracellular cues. Co-staining of NMII and actin in intestinal crypts of Fgfr1op^cKO mice on day 6 post Cre induction revealed a marked reduction of colocalization of NMII and actin filaments (Figures 7C and 7D), while actin itself appeared unaffected, suggesting a dysfunctional NMII–actin interaction. NMII includes two heavy chains, two essential light chains, and two regulatory light chains (RLC)⁴³, and its conformation and function depend on phosphorylation of the RLCs⁴³. To examine the impact of FGFR1OP deletion on RLC phosphorylation, we performed immunoblot analysis of crypts isolated on day 6 post Cre induction. We found a clear reduction in phosphorylation of NMII RLCs in Fgfr1op^cKO crypts (Figure 7E). The expression of total NMII was not affected. Analysis of FGFR1OP-silenced Caco2 cells corroborated the reduction of phosphorylated RLCs compared to control cells (Figure 7F). To confirm the importance of FGFR1OP-dependent NMII activation for the maintenance of cell junctions, FGFR1OP-silenced Caco2 or control cells were modified to stably express a constitutively active form of NMII RLC (Figure S7E). TEM analysis showed that constitutively active RLCs rescued the accumulation of intermediate filaments in desmosomes of FGFR1OP-silenced Caco2 cells (Figure 7G and 7H). Moreover, FGFR1OP-silenced Caco2 cells expressing constitutively active RLCs were as efficient as control Caco2 cells in closing the gap in a wound healing assay (Figures S7F and S7G). Altogether, these data reveal an unprecedented function of FGFR1OP in promoting NMII activation, which enables actomyosin function and maintenance of cell-cell adhesion.

Discussion

As a component of the centrosome and the MTOC, FGFR1OP has been involved in the organization of microtubular structures, such as the mitotic spindle and cilia^11,14,16. Our study demonstrates a role of FGFR1OP in intestinal crypts, which consists of regulating the actomyosin cytoskeleton by promoting NMII phosphorylation. The actomyosin network of crypt epithelial cells is essential to generate tension in the cellular cortex that deforms the cells and shapes the crypt⁴⁴. Our data demonstrate that a defect of FGFR1OP-dependent actomyosin activation in homeostasis dramatically impacts crypt architecture, provoking cell detachment and impairment of barrier function. The role for a centrosomal protein like FGFR1OP in controlling actomyosin is consistent with previous studies showing that the centrosome can function as an actin-organizing center in epithelial cells⁴⁵. Deletion of Fgfr1op was also associated with reduced abundance of intermediate filaments connected to desmosomal cell–cell junctions, suggesting a cross-talk between the actomyosin layer and keratin intermediate filaments, which are anchored to desmosomal adhesions. A link between actomyosin function and desmosomal integrity has been recently supported by a study showing that partial inhibition of NMII by blebbistatin affects desmosome formation and growth⁴⁶. However, apart from a reduction in intermediate filaments, desmosomes themselves remained largely unaffected in Fgfr1op-deficient mice, suggesting that desmosomal malfunction may not be the primary cause of the phenotype, but rather a contributing factor. This is supported by studies on mice lacking desmosomal proteins like desmoglein-2 and desmocollin-2, which do not exhibit the same phenotype seen in Fgfr1op-deficient mice^47–49.

The function of FGFR1OP identified in this study is reminiscent of that of the Drosophila CP190, which is also a centrosomal microtubule-associated protein⁵⁰. Embryos with a CP190 mutation showed unperturbed microtubules but failed in axial expansion, which is an actomyosin-dependent process that distributes the nuclei along the anterior-to-posterior axis. While myosin was disrupted in these embryos, actin was unaffected. The similar capacity of CP190 in early embryos and FGFR1OP in intestinal crypts to sustain the actomyosin cytoskeleton highlights a highly conserved biological mechanism in morphogenesis. In agreement, another centrosomal protein, CEP350, has been shown to contribute to the stabilization of adherens junctions independent of its role in microtubule organization and cell division⁵¹. Deficiency of Lis1, a centrosomal protein with a role in microtubule anchoring, results in decreased attachment of keratin filaments, impaired desmosome function, and loss of epidermal barrier⁵². Thus, centrosome proteins may contribute to the stability of cell junctions in addition to canonical functions in microtubule nucleation. As we did not observe localization of FGFR1OP outside of centrosomes, FGFR1OP may coordinate actomyosin through centrosomal interactions.

Beyond the relevance in intestinal epithelium physiology, the findings reported in this study may be important for human diseases that affect the intestinal epithelium and in particular crypts, such as CD. A reduced resilience to injury may be a relevant predisposing factor among several others. We found that Fgfr1op haploinsufficient mice, despite apparently normal intestinal development, were more susceptible to DSS-induced colitis than their wild-type counterparts. Thus, polymorphisms in the FGFR1OP region may at least in part contribute to disease susceptibility by reducing the ability of intestinal crypts to cope with intestinal stress caused by environmental stimuli. Supporting this conclusion, other genetic defects that reduce intestinal epithelium regeneration and repair lead to prolong mucosal inflammation^53,54. Overall, this study highlights the significance of integrating genetics with scRNA-seq disease atlases. This integration serves as a starting point for understanding gene function and disease pathogenesis. By merging genetics with disease atlases, we discovered that FGFR1OP expression in CD epithelium is primarily limited to cycling and secretory TA cells. Subsequently, we utilized mouse and human model systems to unravel gene function and how cells and tissues are affected in the absence of FGFR1OP. Our study focusing on FGFR1OP has uncovered previously unrecognized cell lineages and dynamic functional states associated with disease pathology. These findings provide a foundation for future mechanistic dissections of CD and may suggest new avenues for therapeutic intervention targeting the actomyosin in epithelial cells.

Limitations of the study

FGFR1OP is one component of the centrosome, which contains hundreds of proteins. While the centrioles appeared intact, it remains to be determined whether the defect in FGFR1OP affects other centrosomal proteins. Although reduced contractility of the actomyosin cytoskeleton was associated with reduced accumulation of intermediate filaments in the desmosomes, the molecular mechanism underlying the cross-talk between the actomyosin cytoskeleton and basolateral desmosomes in the crypts remains unclear. It is possible that the FGFR1OP deficiency interferes with cellular processes responsible for transporting or stabilizing desmosome components. These processes may involve FGFR1OP-mediated interactions with both actomyosin and microtubules. Further investigation is needed to explore these possibilities.

Resource Availability

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by lead contact, Marco Colona (mcolonna@wustl.edu).

Materials Availability

Cell lines generated in this study are available upon request to lead contact.

Data and Code Availability

• Microarray, bulk RNAseq and single-cell RNA-seq data have been deposited at GEO and are publicly available as of the date of publication. Accession numbers are listed in the key resources table.

  The accession numbers for published datasets analyzed in this study are listed in the key resources table.

  Original western blot images have been deposited at Mendeley and are publicly available as of the date of publication. The DOI is listed in the key resources table.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rat monoclonal anti-mouse CD45, APC/Cyanine7, 30-F11	Biolegend	Cat# 103116; RRID: AB_312981
Rat monoclonal anti-mouse/human CD11b, PerCPCyanine5.5, M1/70	Biolegend	Cat# 101228; RRID:AB_893233
Armenian hamster monoclonal anti-mouse CD11c, Brilliant Violet 421, N418	Biolegend	Cat# 117330; RRID:AB_11219593
Mouse monoclonal anti-mouse CD64, PE, X54-5/7.1	Biolegend	Cat# 139304; RRID:AB_10612740
Mouse monoclonal anti-mouse CD64, Brilliant Violet 605, X54-5/7.1	Biolegend	Cat# 139323; RRID:AB_2629778
Rat monoclonal anti-mouse Ly6C, Brilliant Violet 786, HK1.4	Biolegend	Cat# 128041; RRID:AB_2565852
Rat monoclonal anti-mouse Ly6G, FITC, 1A8	Biolegend	Cat# 127606; RRID:AB_1236488
Rat monoclonal anti-mouse I-A/I-E, Brilliant Violet 510, M5/114.15.2	Biolegend	Cat# 107636; RRID:AB_2561397
Rat monoclonal anti-mouse CD170 (Siglec F), APC, S17007L	Biolegend	Cat# 155508; RRID:AB_2750237
Rat monoclonal anti-CD45R/B220, BUV395, RA3-6B2	BD Biosciences	Cat# 563793; RRID:AB_2738427
Rat monoclonal anti-CD45R/B220, PE, RA3-6B2	BD Biosciences	Cat# 553090; RRID:AB_394620
Mouse monoclonal anti-rat CD90 / mouse CD90.1 (Thy-1.1), Brilliant Violet 510, OX-7	Biolegend,	Cat# 202535; RRID:AB_2562643
Rabbit monoclonal anti-FGFR1OP (EPR9929)	Abcam	Cat# ab156013
Rabbit polyclonal anti-EpCAM	Abcam	Cat# ab71916; RRID:AB_1603782
Rabbit monoclonal anti-Lysozyme (EPR2994(2))	Abcam	Cat# ab108508;b RRID:AB_10861277
Rat monoclonal anti-Tubulin (YL1/2)	Abcam	Cat# ab6160; RRID:AB_305328
Rat monoclonal anti-Brdu (BU1/75 (ICR1))	Abcam	Cat# ab6326; RRID:AB_305426
Rabbit monoclonal anti-Iba-1/AIF-1 (E4O4W)	Cell Signaling Technology	Cat# 17198S; RRID:AB_2820254
Rabbit monoclonal anti-OLFM4 (D6Y5A)	Cell Signaling Technology	Cat# 39141; RRID:AB_2650511
Rabbit polyclonal anti-phospho-Histone H3 (Ser10)	Cell Signaling Technology	Cat# 9701; RRID:AB_331535
Mouse monoclonal anti-β-catenin, cl.14	BD Biosciences	Cat# 610153; RRID:AB_397554
Goat polyclonal anti- S100a9	R&D Systems	Cat# AF2065; RRID:AB_2184263
Rabbit monoclonal anti-non-muscle Myosin IIA (EPR22933-9)	Abcam	Cat# 238131; RRID:AB_2924880
Mouse anti-gamma-tubulin (cl.GTU-88)	MilliporeSigma	Cat# T6557; RRID:AB_477584
Donkey anti-Mouse IgG (H+L), Alexa Fluor 488	Thermo Fisher Scientific	Cat# A-21202; RRID:AB_141607
Donkey anti-Mouse IgG (H+L), Alexa Fluor 594	Thermo Fisher Scientific	Cat# A-21203; RRID:AB_141633
Donkey anti-Rabbit IgG (H+L), Alexa Fluor Plus 594	Thermo Fisher Scientific	Cat# A-32754; RRID:AB_2762827
Goat anti-Rabbit IgG (H+L), Alexa fluor 647	Thermo Fisher Scientific	Cat# A-21245; RRID:AB_2535813
Donkey anti-Rabbit IgG (H+L), Alexa fluor 647	Thermo Fisher Scientific	Cat# A-31573; RRID:AB_2536183
Donkey anti-Rat IgG (H+L), Alexa Fluor 594	Thermo Fisher Scientific	Cat# A-21209; RRID:AB_2535795
Mouse monoclonal anti-Phospho-Myosin Light Chain 2 (Ser19)	Cell Signaling Technology	Cat# 3675; RRID:AB_2250969
Rabbit monoclonal anti-Myosin Light Chain 2 (D18E2)	Cell Signaling Technology	Cat# 8505; RRID:AB_2728760
Rabbit monoclonal anti-GAPDH (14C10; HRP conjugate)	Cell Signaling Technology	Cat# 3683; RRID:AB_1642205
Mouse monoclonal anti-GAPDH (1E6D9)	Proteintech	Cat# 60004-1-Ig; RRID:AB_2107436
anti-mouse IgG, HRP linked	Cell Signaling Technology	Cat# 7076; RRID:AB_330924
anti-rabbit IgG, HRP linked	Cell Signaling Technology	Cat# 7074; RRID:AB_2099233
Mouse monoclonal anti-beta actin (C4) HRP	Santa Cruz Biotechnology	Cat# sc-47778 HRP; RRID:AB_2714189
Anti-FLAG	Millipore Sigma	Cat# F3165; RRID:AB_259529
Chemicals, peptides, and recombinant proteins
Alexa Fluor 647 Phalloidin	Thermo Fisher Scientific	Cat# A22287
Tamoxifen	MilliporeSigma	Cat# T5648
(Z)-4-Hydroxytamoxifen	MilliporeSigma	Cat# H7904
Fluorescein isothiocyanate–dextran	MilliporeSigma	Cat# 46944
Trilogy	MilliporeSigma	Cat# 920P
ProLong Glass Antifade Mountant	Life Technologies	Cat# P36930
Fluoromount	Millipore Sigma	Cat# F4680
Paraformaldehyde (32%)	Electron Microscopy Sciences	Cat# 15714
Tissue-Plus O.C.T. compound	Fisher HealthCare	Cat# 4585
Matrigel	Corning	Cat# 354234
Hoechst 33258	Invitrogen	Cat# H3569
4’,6-Diamidino-2-Phenylindole Dihydrochloride (DAPI)	MilliporeSigma	Cat# D9542
RIPA buffer (10x)	Cell Signaling Technologies	Cat# 9806
Protease/phosphatase inhibitor cocktail (100x)	Cell Signaling Technologies	Cat# 5872
Collagenase/Dispase	Roche	Cat# 10269638001
DNase I	Roche	Cat# 11284932001
Dextran sulfate sodium salt, colitis grade (36,000 – 50,000)	MP Biomedicals	Cat# 0216011080
DSS for colitis	TbD Labs.	N/A
Collagenase from Clostridium histolyticum	MilliporeSigma	Cat# C5138
DL-Dithiothreitol	MilliporeSigma	Cat# 646563
Percoll	Cytiva	Cat # 17089101
Agar	MilliporeSigma	Cat# A7921
Formalin solution, neutral buffered, 10%	MilliporeSigma	Cat# HT501128
Corn oil	MilliporeSigma	Cat# C8267
Xylenes	MilliporeSigma	Cat# 247642
2-propanol	MilliporeSigma	Cat# 190764
Trizma base	MilliporeSigma	Cat# T1503
Y-27632	R&D Systems	Cat# 1254
TrypLE Express	Gibco	12605-010
Lipofectamine 2000 Transfection Reagent	Thermo Fisher Scientific	Cat# 11668027
Puromycin dihydrochloride	MilliporeSigma	Cat# P9620
PenStrep	Gibco	Cat# 15140-122
7AAD Viability Staining	Biolegend	Cat# 420404
1M HEPES Buffer	Corning	Cat# 25-060-Cl
Beta-Mercaptoethanol	Sigma	Cat# M-7522
Bovine Serum Albumin (BSA)	Rockland	Cat# BSA-1000
EDTA, 0.5M	Corning	Cat# 46-034-Cl
Glutamax (100x)	Gibco	Cat# 35050-061
MEM Nonessential amino acids (100x)	Corning	Cat# 25-025-Cl
Nuclease-free water	Invitrogen	Cat# AM9937
PBS (1x)	Leinco	Cat# P364
RPMI 1640	Sigma	Cat# R8758-1L
Sodium Pyruvate 100mM	Corning	Cat# 25-000-Cl
Triton X-100	Sigma	Cat# T8787
Tween 20	Fisher Bioreagents	Cat# BP337-500
Ultra-Pure BSA	Thermo Fisher	Cat# AM2616
10x Tris/Glycine Buffer	Biorad	Cat# 1610734
10x Tris/Glycine/SDS Buffer	Biorad	Cat# 1610732
Tris Buffered Saline (TBS), 10x solution	Fisher bioreagents	Cat# BP2471-1
Protein Assay Reagent A	Biorad	Cat# 500-0113
Protein Assay Reagent B	Biorad	Cat# 500-0114
Protein Assay Reagent S	Biorad	Cat# 500-0115
0.05% Trypsin-EDTA	Gibco	Cat# 25300-054
Collagenase Type I	Gibco	Cat# 17100-017
DMEM/F12	Millipore Sigma	Cat# D6421
Polybrene	Millipore Sigma	Cat# H-9268
Critical commercial assays
RNeasy Plus Mini Kit	QIAGEN	Cat # 74134
IntestiCult Intestinal Organoid Growth Medium	STEMCELL Technologies	Cat # 06005
SuperSignal West Pico PLUS Chemiluminescent Substrate	Thermo Fisher Scientific	Cat# 34577
Restore Western Blot Stripping Buffer	Thermo Fisher Scientific	Cat# 21059
RNAscope Wash Buffer Reagents	ACDBio	Cat# 310091
RNAscope Target Retrieval Reagents	ACDBio	Cat# 322000
RNAscope H2O2 & Protease Plus Reagents	ACDBio	Cat# 322330
RNAscope 2.5 HD Detection Reagent - Red	ACDBio	Cat# 322360
LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit, for 405 nm excitation	Life Technologies	Cat# L34957
QIAPrep Spin Miniprep Kit	Qiagen	Cat# 27104
Monarch DNA Gel Extraction Kit	New England BioLabs	Cat# T1020S
Gibson Assembly Cloning Kit	New England BioLabs	Cat# E5510S
BD Pharmingen APC BrdU kit	BD Biosciences	Cat# 552598
Chromium Next GEM Single Cell 3’ Kit v3.1	10x Genomics	PN-1000268
Chromium Next GEM Chip G Single Cell Kit	10x Genomics	PN-1000120
Dual Index Kit TT Set A	10x Genomics	PN-1000215
Deposited data
Microarray of small intestine	This paper	GEO: GSE194311
RNAseq of small intestine crypts	This paper	GEO: GSE195515
RNAseq of spheroids	This paper	GEO: GSE195515
scRNAseq of small intestine crypts	This paper	GEO: GSE194312
scRNA-seq datasets on Crohn’s disease (CD) and Ulcerative Colitis (UC)	Kong et al.7 Smillie et al.37	Broad Single Cell Portal: SCP18847 and SCP25935 (https://portals.broadinstitute.org/single_cell)
Western blot raw data	This paper	DOI: 10.17632/bry9hw4p4h.1
Experimental models: Cell lines
Caco-2 colorectal adenocarcinoma cells	ATCC	Cat# HTB37
L-WRN	Miyoshi and Stappenbeck28; ATCC	Cat# CRL-3276
Caco2 cells expressing C-terminally or N-terminally expressed FLAG-tagged FGFR1OP	This paper	Colonna lab
Caco2 cells expressing FGFR1OP or control shRNA	This paper	Colonna lab
Caco2 cells co-expressing FGFR1OP or control shRNA and either non activated or activated NMII RLC	This paper	Colonna lab
Fgfr1opfl/fl, Fgfr1opfl/+ and Fgfr1opcKO small intestinal spheroid and organoid cell lines	This paper	Colonna lab
293T	ATCC	CRL-3216
Phoenix-Ampho packaging cell line	Dr. Gary Nolan, Stanford	N/A
2.4G2	ATCC	Cat# HB-197
Experimental models: Organisms/strains
B6NTac; B6N-A<tm1Brd>Fgfr1op<tm1a(EUCOMM)Hmgu>/Cnrm	EMMA	EM:04776
B6.Cg-Tg(Vil1-cre)1000Gum/J	Madison et al.57	JAX# 021504
Vil-CreERT2	el Marjou et al.58	JAX# 020282
B6.129P2-Lgr5tm1(cre/ERT2)Cle/J (Lgr5-EGFP-IRES-creERT2)	Barker et al.3	JAX# 008875
B6.Cg-Gt(ROSA)26Sortm9(CAG-tdTomato)Hze/J (Rosa-LSL-tdTomato)	Madisen et al.59	JAX# 007909
C57BL/6J	Jackson Laboratories	JAX# 000664
Oligonucleotides
RNAscope Probe Mm-Fgfr1op-02	ACDBio	Cat # 585401
Recombinant DNA
pCMV3-mFGFR1OP-Flag	SinoBiological	Cat# MG5A4380-CF
pCMV3-mFGFR1OP-Flag	SinoBiological	Cat# MG5A4380-NF
shERWOOD UltramiR Lentiviral Inducible shRNA target gene set for gene FGFR1OP (pZIP-TRE3G-ZsGreen-Puro vectors ULTRA-3199095 and TLNSU4300)	Transomic technologies	Cat# TLHSU2333
pMXs-IRES-Puro Retroviral Expression Vector	Cell Biolabs, Inc.	Cat# RTV-014
pLVX-EF1a-IRES-mCherry	Clonetech Laboratories, Inc.	No. 631987
psPAX2	Didier Trono	RRID:Addgene_12260
pCMV-VSV-G	Stewart et al.67	RRID:Addgene_8454
Software and algorithms
GraphPad Prism	GraphPad	https://www.graphpad.com/features
R 4.4	The R Foundation	https://www.r-project.org RRID: SCR_001905
Rstudio	The R Foundation	https://posit.co RRID: SCR_000432
CellRanger v6.0.1	10xGenomics	https://www.10xgenomics.com/products/single-cell-gene-expression
Seurat v4.0.1	Hao et al.63	https://satijalab.org/seurat/
Python 3.10	Python Software Foundation	https://www.python.org RRID: SCR_008394
Imaris V8.3	Bitplane	RRID: SCR_007370
CellSense	Olympus	https://www.olympus-lifescience.com/en/software/cellsens/
Zen (Blue edition)	Carl Zeiss Microscopy GmbH	https://www.micro-shop.zeiss.com/en/us/softwarefinder/software-categories/zen-blue/
Zen (Black edition)	Carl Zeiss Microscopy GmbH	https://www.micro-shop.zeiss.com/en/us/softwarefinder/software-categories/zen-black/
Metascape	Zhou et al.55	https://metascape.org
FlowJo (v10)	TreeStar	https://www.flowjo.com/solutions/flowjo/
Photoshop	Adobe	https://www.adobe.com/products/catalog.html
ImageJ/Fiji	Schindelin et al.60	https://imagej.net/software/fiji/ RRID: SCR_002285
PEAKS Studio 10 Plus	Bioinfomatic Solutions	https://www.bioinfor.com
Scaffold Q+S	Proteome Software	https://www.proteomesoftware.com

Experimental model and study participant details

Mouse models

All the protocols and procedures involving animals were approved by Institutional Animal Care and Use Committee (IACUC) at Washington University in St. Louis. Mice used in this study were bred and housed in the specific pathogen-free facilities at Washington University in St. Louis. Up to 5 adult mice were housed in filtered cages with corn cob bedding and provided with standard chow and drinking water ad libitum. Cages were changed at least once a week in the laminar air-flow cabinet. Mice were subjected to 12 hours light: 12 hours dark cycles. All the experiments were conducted with sex-matched 7–12 week old littermates. Fgfr1op^fl/fl (Fgfr1op^tm1c/tm1c) mice were generated from B6N-A^tm1Brd Fgfr1op^{tm1a(EUCOMM)Hmgu}/Cnrm embryos purchased from EMMA (now listed as Cep43^{tm1a(EUCOMM)Hmgu}). Fgfr1op^tm1a/+ mice recovered from thawed embryos were bred to C57BL/6-Tg(CAG-flpe)37Ito/ItoRbrc mice⁵⁶ (RIKEN) to remove the IRES-LacZ/neo^r cassette and generate mice carrying a conditional Fgfr1op allele in which exon 3 is flanked by loxP sites; the CAG-flpe transgene was subsequently bred out. B6.Cg-Tg(Vil1-cre)1000Gum/J⁵⁷, Vil-CreERT2⁵⁸, B6.129P2-Lgr5^{tm1(cre/ERT2)Cle}/J (Lgr5-EGFP-IRES-creERT2)³, and B6.Cg-Gt(ROSA)26Sor^{tm9(CAG-tdTomato)Hze}/J (Rosa-LSL-tdTomato)⁵⁹ mice were purchased from the Jackson Laboratories. Fgfr1op^fl/fl-Vil-Cre-ERT2, Fgfr1op^fl/+-Vil-Cre-ERT2, Fgfr1op^fl/fl-Vil-Cre, and Fgfr1op^fl/fl-Lgr5-Cre-ERT2-tdTom mice were genotyped to exclude any that had undergone germline deletion of Fgfr1op exon 3 due to “leaky” expression of Cre. For inducible gene ablation, tamoxifen (MilliporeSigma) was dissolved in corn oil (MilliporeSigma) at a final concentration of 20 mg/mL and i.p. injected into mice at the dose of 75 mg/kg body weight for five consecutive days.

Method details

Tissue preparation for histological analysis

Dissected full-length small intestines and colons were cleaned of luminal content, opened lengthwise, and pinned out. Tissues for paraffin embedding were fixed in 10% neutral buffered formalin solution (MilliporeSigma) overnight at 4°C. Fixed tissues were washed three times in 70% ethanol and embedded in 2% agar (MilliporeSigma), followed by paraffin embedding, sectioning, and either hematoxylin and eosin staining or immunostaining.

For cryosectioning, intestinal tissues were fixed in 4% PFA overnight at 4°C, followed by incubation in 20% sucrose in PBS overnight at 4°C. Tissues were cryo-embedded in Tissue-Plus O.C.T. compound (Fisher Healthcare) and sectioned at 40 μm.

Immunostaining of fixed tissues

Tissue slides were de-paraffinized in xylenes and 2-propanol, followed by antigen-retrieval in Trilogy (MilliporeSigma) solution (20 min) or Tris-EDTA buffer pH9 (30 min) under boiling water. Blocking was conducted in PBS containing 2% BSA and 0.1% Triton-X for 1 h at room temperature (RT). Incubation with primary antibodies was conducted overnight at 4°C. Slides were washed in PBS before incubation with fluorophore-conjugated secondary antibodies for 1h at RT. Nuclear staining was conducted with Hoechst 33258 dye (Invitrogen). Frozen tissue slides were washed in PBS, followed by immunostaining protocol described above starting from the blocking step. Incubation with fluorophore-conjugated secondary antibodies was performed for 2h at RT. The following primary antibodies were used: anti-FGFR1OP (EPR9929), anti-EpCAM, anti-Lysozyme (EPR2994(2)), anti-Tubulin (YL1/2) and anti-non-muscle Myosin IIA (EPR22933–9) from Abcam; anti-Iba-1/AIF-1 (E4O4W), anti-OLFM4 (D6Y5A) and anti-phospho Histone H3 (Ser10) from Cell Signaling; anti-b-Catenin (cl.14, BD Biosciences); anti-S100A9 (R&D Systems) and anti-γ-tubulin (cl.GTU-88, MilliporeSigma). The following secondary antibodies were used: donkey anti-mouse Alexa Fluor-488 and Alexa fluor-594, donkey anti-rabbit Alexa-fluor-594 and Alexa-fluor-647, goat anti-rabbit Alexa-fluor-647, and donkey anti-rat Alexa Fluor-594 (all Thermo Fisher Scientific). F-actin was visualized by Alexa Fluor 647 Phalloidin staining (ThermoFischer Scientific). For BrdU incorporation analysis, mice were injected i.p. with 2 mg of bromodeoxyuridine (BrdU; MilliporeSigma) solution in PBS two hours before sacrificing. Tissue sections were prepared as described above, and BrdU was visualized with anti-Brdu (BU1/75(ICR1), Abcam) antibody. Bright field histological images were acquired with an Olympus BX51 microscope. Fluorescence images were acquired with Zeiss Axio Imager M2 microscope or Zeiss LSM880 Confocal Laser Scanning Microscope with Airyscan. Image-based quantifications were performed in cellSens software (Olympus) and Fiji⁶⁰. Image processing was performed in Adobe Photoshop CC and Imaris v8.3 (Bitplane) software.

RNAscope in situ hybridization

Small intestinal tissues were fixed in 4% PFA overnight, followed by overnight incubation in 20% sucrose in PBS. Tissues were embedded in Tissue-Plus O.C.T. compound (Fisher Healthcare) before freezing. Seven micron-thick cryosections were used for RNAscope in situ hybridization. Probes were custom ordered from ACDBio and were designed to cover nucleotides 314–813 of the Fgfr1op transcripts NM_001197046.1 and NM_201230.5 tv2; this region is 3’ of the segment encoded by exon 3, which is flanked by loxP sites and hence can be excised by Cre. In situ hybridization was conducted in accordance with the manufacturer’s protocol for the RNAscope 2.5 HD Reagent Kit-RED assay, followed by immunofluorescent staining as described above.

Intestinal leukocytes isolation and flow cytometry

Small intestines were cleaned of luminal content and Payer’s patches, opened lengthwise, and incubated in Hank’s balanced salt solution containing 10% FCS and 5mM EDTA for 20 minutes on the rotator. Samples were vortexed at maximal speed and subjected to another round of rotation and vortexing. For intraepithelial leukocyte isolation, DL- Dithiothreitol (MilliporeSigma) at a final concentration of 5mM was added to the supernatants collected after vortexing. Lamina propria leukocytes were extracted by further tissue digestion in 10% RPMI medium containing collagenase IV (MilliporeSigma) for 40 min with vigorous shaking at 37°C. Both intraepithelial and lamina propria leukocytes were purified by density gradient centrifugation using 40% and 70% Percoll (Cytiva) solutions. Single cell suspensions were incubated in the Fc block (2.4G2) before immunostaining. The following monoclonal anti-mouse antibodies were used: CD45 (30-F11), CD11b (M1/70), CD11c (N418), CD64 (X54–5/7.1), Ly6C (HK1.4), Ly6G, (1A8), I-A/I-E (M5/114.15.2), and CD170 (SiglecF; S17007L) (all Biolegend) and B220 (RA3–6B2, BD Biosciences). Dead cells were excluded with 7-AAD (Biolegend). Cell counting was performed using counting beads (eBioscience). Cells were analyzed on a BD FACSymphony A3 flow cytometer, and data using FlowJo software (BD Biosciences).

DSS-induced colitis

Sex-matched mice were treated for 5 or 7 days (indicated in Figure legends) with 2.5% Dextran sulfate sodium salt, colitis grade (molecular weight 36,000 – 50,000 Da, MP Biomedicals or TbD Labs.) in sterile drinking water, followed by switching to regular drinking water until the end of experiment. Mice were weighed daily and humanly euthanized with weight loss greater than 20% of their initial body weight (in accordance with IACUC guidelines). Stool was examined daily for consistency and presence of blood, and scores were defined as follows: 0, solid pellets without blood; 1, soft pellets with traces of blood; 2, soft bloody pellets; 3, loose bloody stool; 4, bloody diarrhea. Presence of blood was determined with Hemoccult Sensa test kit (Beckman Coulter). In the experiments with Fgfr1op^fl/+-Vil-Cre-ERT2 mice, mice were sacrificed either at the end of DSS treatment (day 7) or water recovery phase (day 21). The end of other experiments was determined as a day at which no survivors were present in Fgfr1op-Vil-Cre group.

Isolation and enrichment of small intestinal crypts

Approximately 5 cm of small intestinal tissue was dissected, longitudinally opened, and extensively washed in PBS. Tissue was placed in 2 mM EDTA in PBS for 30 min on the rotator at 4°C. Tissue fragments were placed in 10 mL of 0.1% BCS and pipetted up and down. Pipetting and 0.1% BCS solution exchange was repeated 6 times. Fractions 4–6 were collected, and crypts were collected by centrifugation.

Analysis of FITC-dextran in the serum

Mice were given 0.25 mg/g of Fluorescein isothiocyanate–dextran (FITC-dextran; average mol wt 4,000; Sigma-Aldrich) solution in PBS by oral gavage. After 4 hours, mice were anesthesized and blood was collected by cardiac puncture. During 4 hours of experiment, mice were deprived of food and water. Presence of FITC-dextran in the serum was analysed by spectrophotometric measurement on Multi-Mode Microplate Reader (Synergy HT) at 485/528 nm (excitation/emission).

Microarray analysis

RNA was extracted from the whole tissue small intestinal samples with an RNeasy Mini Kit (Qiagen). RNA quality control, amplification and hybridization to Affymetrix Mouse Gene 1.0 ST array was conducted by the Genome Technology Access Center (GTAC) in the Department of Genetics at Washington University in St. Louis. To determine differentially expressed genes between genotypes over multiple timepoints, gene expression was fit to linear models using the Limma package with R (v3.5)⁶¹. For gene signature scoring, z-scores were calculated for each gene per sample and gene signature scores were calculated using the averaged z-scores for genes in each signature.

Bulk RNA sequencing analysis

Sequencing of RNA isolated from enriched crypts

Small intestinal crypts were isolated and enriched as described above. Total RNA was extracted using RNeasy Mini Kits (Qiagen) per the manufacturer’s instructions. Samples were prepared according to the library kit manufacturer’s protocol (SMARTer Low Input RNA Kit, Clontech Takara Bio), indexed, pooled, and sequenced on an Illumina NovaSeq 6000. Base calls and demultiplexing were performed with Illumina’s bcl2fastq software and a custom python demultiplexing program with a maximum of one mismatch in the indexing read. RNA-seq reads were then aligned to the Ensembl release 76 primary assembly with STAR version 2.5.1a1. Gene counts were derived from the number of uniquely aligned unambiguous reads by Subread:featureCount version 1.4.6-p52.

Sequencing of RNA isolated from spheroid lines

Total RNA was extracted using RNeasy Mini Kits (Qiagen) per the manufacturer’s instructions. Samples were prepared according to the library kit manufacturer’s protocol (Dynal mRNA Direct Invitrogen), and sequencing was conducted on an Illumina HiSeq. Basecalls and demultiplexing were performed with Illumina’s bcl2fastq software and a custom python demultiplexing program with a maximum of one mismatch in the indexing read. Reads were aligned and processed as described for crypt RNA sequencing.

Downstream Analysis

Gene count tables were processed and analyzed using the DESeq2 package with R⁶². Prior to differential comparison, genes with fewer than 10 counts among all of the samples were excluded. Differentially expressed genes (DEGs) were considered significant when padj<0.05 or as indicated in the Result section and within the Figure legends. Gene ontology analysis was conducted using Metascape⁵⁵ (https://metascape.org).

Single-cell RNA sequencing analysis

Intestinal crypts were isolated and enriched as described above. Crypts were dissociated into single cells by incubation in 0.2 mg/mL DNase1 (MilliporeSigma), 1 mg/mL Collagenase/Dispase (MilliporeSigma), and 10 μM Y-27632 (R&D Systems) at 37°C for a total of 10 min, with vigorous shaking every 2 min. Cell sorting was performed on a BD FACS Aria II. FACS-purified live cells were resuspended in 0.04% BSA in PBS at a final concentration of 1,000 cells/μl. cDNA was prepared after the GEM generation and barcoding, followed by the GEM-RT reaction and bead cleanup steps. Purified cDNA was amplified for 11–13 cycles before being cleaned up using SPRIselect beads. Samples were then run on a Bioanalyzer to determine the cDNA concentration. GEX libraries were prepared as recommended by the 10x Genomics Chromium Single Cell 3’ Reagent Kits User Guide (v3.1 Chemistry Dual Index) with appropriate modifications to the PCR cycles based on the calculated cDNA concentration. For sample preparation on the 10x Genomics platform, the Chromium Next GEM Single Cell 3’ Kit v3.1, 16 rxns (PN-1000268), Chromium Next GEM Chip G Single Cell Kit, 48 rxns (PN-1000120), and Dual Index Kit TT Set A, 96 rxns (PN-1000215) were used. The concentration of each library was accurately determined through qPCR utilizing the KAPA library Quantification Kit according to the manufacturer’s protocol (KAPA Biosystems/Roche) to produce cluster counts appropriate for the Illumina NovaSeq6000 instrument. Normalized libraries were sequenced on a NovaSeq6000 S4 Flow Cell using the XP workflow and a 50×10×16×150 sequencing recipe according to the manufacturer’s protocol. A median sequencing depth of 50,000 reads/cell was targeted for each Gene Expression Library.

FASTQ files were aligned to the mm10 genome using CellRanger v6.0.1 (10x Genomics) to produce a count matrix of cells and UMI counts per gene. Count matrices for all samples (3 Fgfr1op^cKO and 3 Fgfr1op^fl/fl) were combined for downstream analysis. To filter cells with uncharacteristically low library size or genes with low counts across all cells, we removed cells with log₁₀ library size > 3.2 and genes with log₁₀ mean expression across all cells > 3.5, corresponding to the tails of each distribution. Cells with greater than 20% mitochondrial reads were removed; ribosomal genes and noncoding RNAs were also removed. We then conducted library size and log-normalization in addition to scaling by a factor of 10,000 in Seurat v4.0.1⁶³. We further removed doublets, as computed by Scrublet⁶⁴, and all lymphocyte clusters. The top 10 principal components of the gene expression data were used for Louvain clustering at resolution 0.4 and for a two-dimensional embedding of the data using the RunUMAP() function with default parameters. To define markers for each cluster, we used the FindMarkers() function and specified “Wilcox” as the statistical test. Genes were considered significantly differentially expressed at FDR P < 0.05. Cell cycle signatures were obtained from Tirosh et al.³⁶, and stem/TA signatures were obtained from Haber et al²². Signature scores were computed for single cells using the AddModuleScore() function in Seurat with default parameters. Comparisons of signature scores between cells were computed using a two-sided Wilcoxon test. RNA velocity analysis was performed using Python 3.10 and scvelo 0.3.1. Integrative analysis of published scRNA-seq datasets on Crohn’s disease (CD)⁷ and Ulcerative Colitis (UC)³⁷ was performed by subsetting the dataset to annotated epithelial cell types. Briefly, genes expressed in > 10 cells were considered in the combined dataset. Dimensionality reduction and batch correction was performed on each individual sample using Harmony⁶⁵ package. Top 60 principal components were considered for clustering using the leiden clustering implement in scanpy package, which was visualized using the UMAP embedding. Wilcoxon test was performed to define the markers specific to individual clusters. Post-hoc analysis was performed on identified clusters to remove poor quality cells. Specifically, clusters expressing both non-epithelial lineage markers along with epithelial lineage markers were removed as doublet cells.

Stem cell spheroid lines

Small intestinal stem cell spheroid lines were obtained as previously described²⁸. Approximately 1 cm segment of intestinal tissue was minced with scissors and digested in 2 mg/ml of Collagenase I (Thermo Fisher Scientific) at 37°C for 30–40 min, with vigorous pipetting every 5 min. The solution was filtered through a 70 μm cell strainer, and crypts were collected by centrifugation and plated on 24-well plates in 20 μL of Matrigel (Cornig). Following polymerization of the Matrigel for 10 min at 37°C, 50% L-WRN conditioned medium was added. Spheroids were maintained in the culture in the cycles of three days before dissociation with TrypLE express (Thermo Fisher Scientific) and re-plating. For TAM-induced gene ablation, spheroids were cultured for 3 days in 50% L-WRN medium containing 250 nM Z-4-Hydroxytamoxifen (MilliporeSigma). After re-plating, spheroids were cultured in 50% L-WRN without Z-4-Hydroxytamoxifen. Images of spheroid cultures were taken with a Zeiss AxioObserver inverted microscope.

Small intestinal organoid lines

Small intestinal organoids were obtained as previously described^26,66. For the establishment of organoid lines, crypts were plated on 24-well plates in 20 μL of Matrigel and cultured in IntestiCult Intestinal Organoid Growth Medium (STEMCELL Technologies) for 6 days, with medium exchange on day 3. For TAM-induced gene ablation, organoids were treated with 250 nM Z-4-Hydroxytamoxifen for 3 days, followed by medium exchange to Intesticult medium without Z-4-Hydroxytamoxifen. Images of organoid cultures were taken on day 5 with a Zeiss AxioObserver inverted microscope.

Single cell spheroid formation assay

LGR5+tdTom+ cells were labeled by a single injection of TAM (75 mg/kg body weight). On day 7 post TAM injection, ~5 cm of small intestinal tissue was dissected, cleaned of luminal content, and longitudinally opened. Tissue was treated with 30mM EDTA, 1.5mM DTT, and 10 μM Y-27632 (R&D Systems) solution in PBS for 20 min on ice, followed by 10 min incubation in 30 mM EDTA and 10 μM Y-27632 solution in PBS at 37°C with vigorous shaking. Cell supernatants were centrifuged, and pellets were resuspended in an enzymatic cocktail containing 0.2 mg/mL DNase1 (MilliporeSigma), 1 mg/mL Collagenase/Dispase (MilliporeSigma), and 10 μM Y-27632. Cells were incubated at 37°C for 10 min with vigorous shaking every 2 min to obtain single cell suspensions. Lgr5+tdTom+ cells were sorted on a BD FACS Aria II instrument, and 10,000 cells per 24-well were plated in Matrigel and 50% L-WRN medium. To assess spheroid formation capacity, cells were cultured for 9 days, followed by dissociation and re-plating.

Generation of stable Caco2 cell lines expression FLAG-tagged FGFR1OP

Plasmids expressing Fgfr1op with either C- or N-terminally expressed FLAG were purchased from Sinobiological. Sequences were cloned into pMXs-Puro-IRES retroviral expression vectors (Cell Biolabs, Inc.) and confirmed by sequencing. Transfection into Phoenix retrovirus producer cells was conducted with Lipofectamine 2000 (Thermo Fisher Scientific) per the manufacturer’s protocol. Supernatants were collected after 48h and filtered through 0.45 μm filters. Transduction was conducted with 1 mL of supernatant containing polybrene at a final concentration of 8 mg/mL in the 12-well plate, with centrifugation at 900g for 1.5h. Transfected cells were selected with puromycin (10 μg/ml) added in culture 3 days post transduction. Stably transduced cells were subsequently sorted based on FLAG expression using an anti-FLAG M2 antibody (MilliporeSigma).

Immunoprecipitation and mass spectrometry

Caco2 cells stably expressing FGFR1OP with C- or N-terminal FLAG and control Caco2 cells were lysed in RIPA buffer with Lysonase (MilliporeSigma) and EDTA-free protease inhibitors. Soluble fractions were isolated by centrifugation at 17,000g for 10 minutes at 4°C. Immunoprecipitation was conducted with an anti-FLAG antibody (MilliporeSigma) bound to Dynabeads (Thermo Fisher scientific). Immunoprecipitated samples were resuspended in a Tris/urea buffer, reduced, alkylated, and digested with trypsin at 37°C overnight. This solution was subjected to solid phase extraction to concentrate the peptides and remove unwanted reagents, followed by injection onto a Waters NanoAcquity HPLC equipped with a self-packed Aeris 3 m C18 analytical column 0.075 mm by 20 cm (Phenomenex). Peptides were eluted using standard reverse-phase gradients. The effluent from the column was analyzed using a Thermo Orbitrap Elite mass spectrometer (nanospray configuration) operated in a data-dependent manner for 90 minutes. The resulting fragmentation spectra were correlated against Refseq database entries concatenated to user-provided target sequences and an in-house contaminants database using PEAKS Studio 10 Plus (Bioinfomatic Solutions) for posttranslational modifications and sequence coverage of target proteins. Scaffold Q+S (Proteome Software) was used to provide consensus reports for all identified proteins.

Generation of a Caco2 cell line expressing shRNA targeting FGFR1OP

shERWOOD UltramiR Lentiviral Inducible shRNA target gene set for gene FGFR1OP was purchased from Transomic technologies. Lentiviral packaging was performed in the 293T packaging cell line (ATCC, CRL-3216) with Lipofectamine 2000 (Thermo Fisher Scientific) per the manufacturer’s protocol. Cells were transfected with psPAX2 packaging plasmid (Addgene No. 12260), pCMV-VSV-G envelope plasmid⁶⁷ (Addgene No. 8454), and pZIP-TRE3G-ZsGreen-Puro vector expressing either shRNA targeting FGFR1OP (ULTRA-3199095) or control pZIP-TRE3G-ZsGreen-Puro vector expressing non targeting shRNA (TLNSU4300) per the manufacturer’s instructions. Supernatants were collected after 48h and filtered through 0.45 μm filters. Transduction was conducted with 1 mL of supernatant containing polybrene at a final concentration of 8 mg/mL in the 12-well plate, with centrifugation at 900g for 1.5 h. The medium was changed next day. Transfected cells were selected with puromycin (10 μg/ml) added in culture 3 days post transduction.

Silencing of FGFR1OP in stably transduced Caco2 cells was induced by adding doxycycline (2 μg/ml; MilliporeSigma) in the cell culture medium. Cells were kept under induction for 3 days, and doxycycline-containing medium was changed on day 2. Efficiency of FGFR1OP shRNA induction and FGFR1OP silencing was confirmed by visualizing the expression of the ZsGreen reporter with an Olympus BX51 microscope and by Western blot, respectively.

Treatment of shRNA-expressing Caco2 cells with 10 μM Latrunculin A was conducted on day 3 of doxycycline treatment for 1h prior processing cells for TEM analysis.

Generation of Caco2 cell lines expressing activated myosin

Myosin regulatory chain (RLC, Myl12a) was expressed in Caco2 cells stably expressing FGFR1OP or control shRNA (generated as described above). RLC or the active form of RLC containing phosphomimetic glutamic acids in place of phosphorylatable threonine and serine residues (Thr19Ser20) were cloned into pLVX-EF1a-IRES-mCherry vectors (Clonetech Laboratories, Inc., No. 631987) by Gibson Assembly Cloning Kit (NEB) and confirmed by sequencing. Packaging of pLVX-EF1a-IRES-mCherry expressing either form of RLC and transduction of Caco2 cells stably expressing FGFR1OP or control shRNA were performed as described above. Cells were cultured for the next five days, and stably transduced cells were sorted based on mCherry expression on a BD FACS Aria II instrument.

Transmission electron microscopy

Small intestinal samples were cleaned of luminal content, opened lengthwise, and pinned out in 2% PFA / 2.5% glutaraldehyde (Ted Pella, Inc., Redding, CA) in 100 mM sodium cacodylate buffer, pH 7.2 for 2 hours at RT. Samples were washed in sodium cacodylate buffer and postfixed in 1% osmium tetroxide (Ted Pella, Inc.) for 1 hour. Samples were then rinsed extensively in dH₂0 prior to en bloc staining with 1% aqueous uranyl acetate (Ted Pella Inc.) for 1 hour. After several rinses in dH₂0, samples were dehydrated in a graded series of ethanol and embedded in Eponate 12 resin (Ted Pella Inc.). Sections of 95nm thickness were cut with a Leica Ultracut UCT ultramicrotome (Leica Microsystems Inc., Bannockburn, IL) and collected on Cu grids, followed by staining with uranyl acetate and lead citrate. Imaging was performed on a JEOL 1200 EX transmission electron microscope (JEOL USA Inc., Peabody, MA) equipped with an AMT 8 megapixel digital camera and AMT Image Capture Engine V602 software (Advanced Microscopy Techniques, Woburn, MA).

For ultrastructural analyses of centrioles, intestinal samples were fixed in a freshly prepared mixture of 1% glutaraldehyde and 1% osmium tetroxide in 50 mM phosphate buffer at 4°C for 2 hours. Samples were then rinsed extensively in cold dH₂0 prior to en bloc staining and were further processed as described above.

For immunolocalization at the ultrastructural level, small intestinal samples were fixed in 4% paraformaldehyde/0.05% glutaraldehyde in 100mM PIPES/0.5mM MgCl2, pH 7.2 for 1 hr at 4°C, embedded in 10% gelatin, and infiltrated overnight with 2.3M sucrose/20% polyvinyl pyrrolidone in PIPES/MgCl2 at 4°C. Samples were trimmed, frozen in liquid nitrogen, and sectioned with an ultramicrotome. Sections of 65nm thickness were blocked with 5% FBS/5% NGS for 30 min and subsequently incubated with an anti-FGFR1OP antibody for 1 hr, followed by secondary goat anti-rabbit IgG+IgM conjugated to 18 nm colloidal gold (Jackson ImmunoResearch Laboratories, Inc., West Grove PA) for 1 hr. Sections were stained with 0.3% uranyl acetate/2% methyl cellulose and analyzed by TEM as described above. All labeling experiments were conducted in parallel with controls omitting the primary antibody. These controls were consistently negative at the concentration of colloidal gold conjugated secondary antibodies used in these studies. TEM was performed on at least 20 cells in Caco2 cells and crypts per sample. Crypt images are representative of at least 10 crypts per sample. Measurements of desmosomes were performed in Fiji⁶⁰.

Western blot

Small intestinal crypts were isolated and enriched on day 6 post Cre induction (as described above). Caco2 cells transduced with either control or FGFR1OP shRNA (as described above) were collected on day 3 of doxycycline treatment. Crypts and transduced Caco2 cells were washed in PBS and lysed in RIPA buffer (Cell Signaling Technology). Lysates were centrifuged at 12,000g at 4°C for 15 min, and supernatants were collected. Protein concentration was determined using the DC Protein Assay (Bio-Rad) per the manufacturer’s instructions. Samples were denatured and reduced in 4X Laemmli sample buffer (Bio-Rad) containing 10% β-mercaptoethanol with heating at 95°C for 5 min. Proteins were separated by gel electrophoresis on Mini PROTEAN TGX gels 4–20% (Bio-Rad) and blotted to PVDF membrane (Millipore). Membranes were blocked in either 5% BSA (Rockland) or 5% nonfat dried milk (Bio-Rad) in TBST buffer (Tris buffered saline (TBS, Fisher bioreagents) containing 0.05% Tween 20 (Fisher scientific)) for 1h at RT. Membranes were incubated with the following primary antibodies at 4°C on the rotator overnight: anti-FGFR1OP (EPR9929, Abcam), anti-myosin light chain 2 (D18E2), and anti-phospho myosin light chain 2 (Ser19) (both Cell Signaling Technologies), and anti-GAPDH (1E6D9, Proteintech). Incubation with HRP-linked secondary antibodies, anti-mouse IgG and anti-rabbit IgG (Cell Signaling Technologies) as well as HRP-labeled anti-GAPDH (14C10, Cell Signaling Technologies) was performed at RT for 2 hours. Incubation with HRP-labeled anti-β-actin (C4, Santa Cruz) was done for 3 hours at RT. Bands were visualized with SuperSignal West Pico Plus chemiluminescent substrate (Thermo Fisher Scientific). Membrane stripping was performed with Restore Western Blot Stripping Buffer (Thermo Fisher Scientific) for 20–30 min.

Wound-healing assay

FGFR1OP or control shRNA-expressing Caco2 cells, or FGFR1OP or control shRNA-expressing Caco2 cells co-expressing either RLC or active form of RLC were treated with 2 μg/ml doxycycline for 4 days in total. On day two of doxycycline treatment, 4×10⁴ cells were seeded into each chamber of a Culture-Insert 2 Well in μ-Dish 35 mm (Ibidi) and incubated at 37°C overnight to reach confluence. On day three of treatment the silicone insert was removed, followed by image acquisition (defined as day 0 of the assay). Next day images were taken for day 1 timepoint. Untreated FGFR1OP or control shRNA-expressing Caco2 cells were seeded at a density of 4×10⁴ cells per chamber as a control, followed by removal of insert next day (day 0) and imaging on day 1. Images were acquired on a Zeiss AxioObserver D1 Inverted Microscope. Percentage of wound closure was calculated based on uncovered wound areas at day 0 and day 1 post insert removal. Area was calculated in Fiji⁶⁰.

Quantification and statistical analysis

Statistical analysis was conducted using GraphPad Prism 9 software (GraphPad Software, USA). Statistical parameters, methods, and significance are reported in the Results section, within Figure legends and in the Method details. The data are presented as mean with standard deviation (SD). The data were considered statistically significant when p<0.05. Image analysis was conducted in Fiji⁶⁰. All experiments were repeated independently at least twice with reproducible results, except microarray and RNA seq analysis experiments, which were performed once.

Supplementary Material

1

Document S1. Figures S1–S7 and supplemental references

2

Table S1. The list of stem cells, TA cells and enterocytes signature genes²² and cell cycle signature genes³⁶ used in RNA sequencing analysis for defining cell identity and cell cycle phase, respectively, related to Figures 5, S4 and S5

3

Table S2. The list of proteins identified in Caco2 cells expressing C- or N-FLAG-tagged FGFR1OP by mass spectrometry from two independent experiments, related to Figure 7

Highlights.

• GWAS have linked CD with the FGFR1OP, which encodes a centrosome protein

• FGFR1OP deletion in mouse intestinal epithelial cells disrupted crypts architecture

• FGFR1OP promoted activation of non-muscle myosin II, ensuring crypt cell adhesion

• FGFR1OP defects may heighten CD susceptibility by reducing gut epithelial renewal