CCN1 coordinately regulates intestinal stem cell proliferation and differentiation through integrins αvβ3/αvβ5


 Intestinal stem cells (ISCs) at the crypt base contribute to intestinal homeostasis through a balance between self-renewal and differentiation. However, the molecular mechanisms regulating this homeostatic balance remain elusive. Here we show that the matricellular protein CCN1/CYR61 coordinately regulates ISC proliferation and differentiation through distinct pathways emanating from CCN1 interaction with integrins αvβ3/αvβ5. Mice that delete Ccn1 in Lgr5+ ISCs or express mutant CCN1 unable to bind integrins αvβ3/αvβ5 exhibited exuberant ISC expansion and enhanced differentiation into secretory cells at the expense of absorptive enterocytes in the small intestine, leading to nutrient malabsorption. Analysis of crypt organoids revealed that through integrins αvβ3/αvβ5, CCN1 induces NF-κB-dependent Jag1 expression to regulate Notch activation for differentiation and promotes Src-mediated YAP activation and Dkk1 expression to control Wnt signaling for proliferation. Moreover, CCN1 and YAP amplify the activities of each other in a regulatory loop. These findings establish CCN1 as a novel niche factor in the intestinal crypts, providing new insights into how matrix signaling exerts overarching control of ISC homeostasis.


INTRODUCTION
The main function of the intestine is to efficiently absorb nutrients and water while enduring mechanical force, extreme pH variations, and constant exposure to the gut microbiota.
To avoid the accumulation of damaged cells while achieving these functions, the intestine continuously sheds aged epithelial cells and rapidly regenerates the crypt-villus structure every 4-5 days 1 . The villus is a protrusion of epithelium toward the intestinal lumen to maximize the surface area for efficient nutrient absorption and contains a single columnar layer of fully differentiated post-mitotic epithelial cells. Actively dividing stem cells reside at the crypt, an epithelial invagination surrounding the villi, and proliferate for self-renewal or generate progenitor cells that move up to the transit-amplifying (TA) zone to rapidly divide before committing into epithelial cells of specific lineages to replenish the aged and damaged cells in the villi. To maintain homeostasis, the balance between self-renewal and differentiation of the intestinal stem cells (ISCs) needs to be fine-tuned and sustained.
Genetic lineage-tracing identified a stem cell population at the bottom of the crypts that expresses the leucine-rich repeat-containing G protein-coupled receptor 5 (LGR5) and is fully capable of long-term self-renewal and generation of both absorptive cells (enterocytes and M cells) and secretory cells (Paneth cells, goblet cells, enteroendocrine cells, and tuft cells) 1, 2 .
These stem cell activities are largely governed by the surrounding microenvironment 3 , the socalled stem cell niche, which instructs ISCs to proliferate and/or differentiate through cellular signaling driven by several niche factors 4,5,6,7 , including Wnt, Notch ligands, epidermal growth factor (EGF), and bone morphogenic protein (BMP). Especially, Wnt and Notch signaling are crucial determinants in regulating ISC proliferation and differentiation, respectively. The canonical Wnt activation, initiated by the binding of secreted Wnt ligands to the cognate receptors frizzled (FZD) and low-density lipoprotein receptor-related protein 5/6 (LRP5/6), leads to the nuclear accumulation of β-catenin and activation of T cell factor 4 (TCF4) target genes involved in ISC proliferation and maintenance 8 . Mice with systemic 9 or epithelial deletion of 5 Here we provide the first evidence that CCN1 coordinately regulates ISC proliferation and differentiation during homeostasis, and Ccn1 deletion resulted in enhanced ISC proliferation and increased secretory cells at the expense of absorptive cells. Analysis of crypt organoids revealed that CCN1 regulates both Wnt and Notch signaling for proliferation and differentiation of ISCs, respectively, through distinct pathways emanating from the engagement of integrins α v β 3 /α v β 5 . Moreover, a CCN1-yes-associated protein (YAP) regulatory loop amplifies their activities in the crypt. Together, these findings identify CCN1 as a novel niche factor that activates integrins α v β 3 /α v β 5 signaling to coordinately regulate ISC proliferation and differentiation through distinct downstream pathways, thus expanding our understanding of the overarching signaling mechanisms contributing to the maintenance of intestinal homeostasis.

Ccn1 deletion in Lgr5+ ISCs engendered nutrient malabsorption
We examined Ccn1 expression in the normal mouse small intestine to study its functions.
Immunostaining of green fluorescent protein (GFP) in the jejunum of Ccn1:EGFP reporter mice (Ccn1 EGFP/+ ) showed strong CCN1 expression in the lower base of the crypts where stem cells reside (Fig. 1A). Fluorescence in situ hybridization (FISH) also detected Ccn1 mRNA signals from Lgr5-expressing (+) ISCs in Lgr5-EGFP-IRES-CreERT2 mice (Lgr5 EGFP-CreERT ; Fig. 1B). The secreted CCN1 protein was found along the luminal side of epithelial cells at the crypt base by immunohistochemistry, surrounding the Lgr5+ ISCs and the neighboring Paneth cells, although its expression was also seen in a few epithelial cells and mesenchymal cells in the villi and lamina propria (Fig. 1C). A similar CCN1 expression pattern was observed in the crypt of the colon ( Supplementary Fig. S1). Thus, we postulated that CCN1 may play a role in Lgr5+ ISCs in the crypt of the intestine. To test this possibility, we created Ccn1 conditional knock-out mice (Ccn1 flox/flox ; Lgr5 EGFP-CreERT , hereafter Ccn1 ∆Lgr5 ) that delete Ccn1 in Lgr5+ ISCs upon tamoxifen (TM) treatment (Fig. 1C). Indeed, TM administration abolished CCN1 expression in the crypt base without affecting mesenchymal expression (Fig. 1C). Immunoblot analysis further confirmed that Lgr5+ ISCs are responsible for CCN1 expression in the crypt (Fig. 1D).
Ccn1 ∆Lgr5 mice treated with corn oil (vehicle) or TM showed similar body weight within the first 5 days, beyond which TM-treated mice ceased to gain weight, and by day 9 weight loss was evident, implying defects in nutrient uptake (Fig. 1E). Ccn1 flox/flox mice with the same TM regimen did not show body weight loss, ruling out possible side effects of TM. Interestingly, the jejunum of Ccn1 ∆Lgr5 mice exhibited structural changes by histology (H&E) at day 5-post TM treatment (Fig. 1F). The villus length was reduced by ~28% with increased blobs on the epithelium, whereas the crypt became slightly elongated by ~17% with more eosinophilic granules. To assess any effect on nutrient uptake and metabolism, Ccn1 ∆Lgr5 mice at day10-post vehicle or TM treatment were orally gavaged with sucrose and blood glucose was measured thereafter (Fig. 1G). Blood glucose level reached its peak within 15 min and was maintained for up to 60 min before its decline in vehicle-treated mice. By contrast, blood glucose was 50% lower in Ccn1 ∆Lgr5 mice with TM at 15 min and further declined to background level by 30 min, indicating impaired sugar absorption in mice with Ccn1 deletion in Lgr5+ ISCs. Consistently, immunoblot analysis showed a greatly reduced expression of sucrase that hydrolyzes sucrose into glucose and fructose and Na + K + ATPase β1 which is critical for generating Na + gradient for glucose uptake (Fig. 1H). qPCR analysis confirmed that enterocyte-specific genes involved in glucose uptake and transport (Atp1a1, Atpab1, Slc2a1, Slc2a2, and Slc5a1) were downregulated in TM-treated Ccn1 ∆Lgr5 mice (Fig. 1I). In addition, the paracellular transport in the epithelium was also altered, as orally gavaged fluorescein isothiocyanate (FITC)-dextran was detected at significantly increased amounts in the bloodstream of TM-treated Ccn1 ∆Lgr5 mice ( Fig. 1J) and consistent downregulation of genes involved in tight junction formation (Claudin-1, -2, -3, and -8, and Zo-2) was observed in qPCR analysis (Fig. 1K). Thus, Ccn1 is expressed in the intestinal crypts and its deletion in Lgr5+ ISCs engenders architectural and functional alteration of the intestinal epithelium, resulting in nutrition malabsorption.

Ccn1 deletion in Lgr5+ ISCs caused ISC expansion and increased secretory cells
We then monitored the cellular composition in the jejunum of Ccn1 ∆Lgr5 mice. Surprisingly, expansion of Lgr5+ ISCs in the crypts, assessed by GFP expression (Fig. 2A, B), became evident by day 5 of TM treatment and this was confirmed by a marked increase in the expression of olfactomedin 4 (Olfm4), another stem cell marker ( Fig. 2A, B). The expanded Lgr5+ ISCs still maintained their topological position at the crypt base. However, there was no evidence of hyperplasticity even by day 28 of TM treatment ( Supplementary Fig. S2). qPCR analysis on the jejunum of Ccn1 ∆Lgr5 mice with TM also showed increased expression of ISC signature genes, such as Olfm4, Ascl2, and Sox9 (Fig. 2C). Moreover, ISC expansion was accompanied by increased Ki67+ proliferating cells in the middle region of the crypts, presumably the TA zone ( Fig. 2A). Crypts from Ccn1 ∆Lgr5 mice with TM formed twice as many colonies in 3D cultures as those from vehicle-treated mice when seeded equally, demonstrating the presence of more functional ISCs (Fig. 2D). We also assessed the composition of fully differentiated epithelial cells. Judging from enzymatic staining of alkaline phosphatase (AP), TMtreated Ccn1 ∆Lgr5 mice showed a marked reduction in enterocytes throughout the entire villi ( Most CCN1 functions are mediated through integrin receptors, primarily α v β 3 /α v β 5 and α 6 β 1 , in a cell-type-specific manner 19,20,21 . Remarkably, Ccn1 D125A/D125A knock-in mice (crossed with Lgr5 EGFP-CreERT mice to mark the Lgr5+ ISCs), which express a mutant CCN1 with a single amino acid change (CCN1-D125A) that is unable to bind integrins α v β 3 /α v β 5 (Ref. 19 ), also exhibited increased Lgr5+ ISCs, Paneth cells (Lyz1), and goblet cells (Muc2), but a reduction in enterocytes in the jejunum by immunofluorescence staining, AP staining, and qPCR analysis ( Fig. 2G, H), phenocopying Ccn1 ∆Lgr5 mice with TM treatment. Together, these results indicate that Ccn1 plays a role in both ISC proliferation and differentiation during homeostasis, most likely through integrins α v β 3 /α v β 5 .

Curtailed Notch and enhanced Wnt signaling with Ccn1 deletion in Lgr5+ ISCs
The compound phenotypes of Ccn1 ∆Lgr5 mice with TM suggest the involvement of multiple signaling pathways, among which Notch and Wnt signaling are known to regulate ISC fate decision and proliferation, respectively 5 . The presence of the Notch1 intracellular domain (N1-NICD) indicated active Notch signaling in the crypt base of vehicle-treated Ccn1 ∆Lgr5 mice, which was greatly diminished with TM-mediated Ccn1 deletion by immunofluorescence staining and immunoblotting (Fig. 3A, B). However, NICD of another Notch paralog Notch2 was unaffected ( Fig. 3B). Despite their critical involvement in Notch activation of ISCs 31 , the expression of Notch ligand Delta-like 1 (DLL1) or DLL4 was not affected (Fig. 3B). Instead, another ligand Jagged 1 (Jag1) was greatly reduced in the jejunum of Ccn1 ∆Lgr5 mice with TM ( Fig. 3B), suggesting that Jag1 deficiency may cause the reduced Notch activation. Notch ligand-receptor interaction is regulated by Notch receptor glycosylation 13, 32 . Especially, the Fringe family (Manic, Lunatic, and Radical) of N-acetylglucosaminyltransferases glycosylate the Notch receptors, thereby potentiating DLL-induced signal but diminishing responsiveness to Jag 33,34 .
Interestingly, the crypt of Ccn1 ∆Lgr5 mice with TM exhibited a marked reduction in the expression of Manic Fringe (MFNG) ( Supplementary Fig. S3A), which might alter the glycosylation status of Notch1 receptor and hence Notch ligand preference from DLLs to Jag1 (Ref. 35 ). Consequently, reduced Notch activation in Ccn1 ∆Lgr5 mice with TM led to decreased expression of Notch target genes, including Hes1, Hey1, and HeyL, resulting in de-repression of Atoh1 (Fig. 3C), which is consistent with the preferential secretory cell differentiation observed in Ccn1 ∆Lgr5 mice with TM ( Fig. 2E).
Wnt signaling was also affected upon Ccn1 deletion in Lgr5+ ISCs. Ccn1 ∆Lgr5 mice treated with TM showed markedly stronger signals of non-phosphorylated, active β-catenin by immunohistochemistry, indicating increased Wnt signaling (Fig. 3D). Immunoblot analysis confirmed higher levels of β-catenin (both total and active forms) in the jejunum of Ccn1 ∆Lgr5 mice with TM (Fig. 3E). Accordingly, several Wnt target genes (Axin2, Cd44, Ephb2, and Myc) were upregulated (Fig. 3F) and the expression of Wnt ligand Wnt3a, but not the non-canonical ligands Wnt5a and Wnt5b, was increased (Fig. 3G). Importantly, the expression of Dkk1, a secreted Wnt antagonist that inhibits canonical Wnt signaling by competitive binding to coreceptors LRP5/6 (Ref. 36 ), was greatly reduced (Fig. 3G), possibly contributing to enhanced Wnt signaling. Taken together, these results show that CCN1 regulates both Notch and Wnt signaling, thereby influencing fate decision and proliferation of ISCs in the mouse intestine.

CCN1-integrins α v β 3 /α v β 5 controls proliferation and differentiation of ISCs in organoids
Intestinal organoids retain the intestinal stem cell hierarchy and serve as an ideal system to study regulation and signaling mechanisms 37 . We cultured intestinal crypt organoids from Several lines of evidence indicated that CCN1 functions through integrins α v β 3 /α v β 5 .

CCN1 restricts ISC proliferation through the Src-YAP-Wnt axis
CCN1 add-back reversed Lgr5+ ISC expansion even in the presence of Bay11-7082 (Fig. 4A), suggesting that CCN1 regulation of ISC proliferation is independent of the NF-κB/Jag1/Notch pathway. Integrin-mediated signaling has been shown to activate FAK as a key signaling mediator, which recruits and forms a kinase complex with the non-receptor tyrosine kinase Src 38 . Importantly, the selective Src inhibitor dasatinib blocked the CCN1 add-back effect only on ISC proliferation (Lgr5+GFP and Ki67), but not differentiation (Fig. 5A, B), which contrasted with the FAK inhibitor (GSK2256098) that blocked both proliferation and differentiation effects of CCN1 add-back (Fig. 4A, B). These findings suggest that integrins α v β 3 /α v β 5 signaling bifurcates into distinct pathways to regulate ISC proliferation and differentiation. Immunoblot analysis confirmed that Src (Y419) is activated in control organoids but not in organoids with Ccn1 deletion (Fig. 5C). CCN1 add-back reactivated p-Src (Y419), which was blocked by SB273005 or dasatinib (Fig. 5C), demonstrating CCN1 activation of Src via integrin signaling to regulate ISC proliferation.
Among the signaling molecules that Src regulates in intestinal epithelial cells is YAP, a transcriptional co-activator important for tissue regeneration and tumorigenesis in the intestine 39,40 . Importantly, YAP is known to inhibit Wnt signaling and suppress ISC proliferation 41 . Src has been shown to activate YAP through direct phosphorylation of YAP (Y357) 42 or indirectly by phosphorylating, thus inactivating, the upstream negative regulator large tumor suppressor homolog 1 (Lats1) 43 . In parallel with the activation status of Src (Y419), both total YAP and activated p-YAP (Y357) were reduced upon Ccn1 deletion but increased by CCN1 add-back ( Fig. 5C), whereas the levels of active p-Lats1 (T1079) and inactive p-YAP (S127) phosphorylated by Lats1/2 showed the opposite trend. Moreover, SB273005 or dasatinib 13 blocked CCN1 add-back-induced YAP activation (Fig. 5C), demonstrating that CCN1-integrins α v β 3 /α v β 5 signaling leads to Src-mediated YAP activation. Consistently, immunohistochemistry showed nuclear-enriched YAP expression in the crypt base of control Ccn1 ∆Lgr5 mice, but TM treatment greatly reduced YAP signals in the entire crypts of Ccn1 ∆Lgr5 mice ( Supplementary Fig.   S6A). Immunoblot analysis confirmed that active p-YAP (Y357) decreased upon TM treatment, while inactive p-YAP (S127) increased ( Supplementary Fig. S6B). YAP involvement was further examined using Super-TDU 44 3G). As Dkk1 is a YAP target gene 45 , we hypothesized that CCN1 regulation of Wnt signaling might be through YAP-dependent Dkk1 expression. Indeed, CCN1 add-back to 4-OHT-treated Ccn1 ∆Lgr5 organoids rescued Dkk1 expression, but Super-TDU efficiently blocked CCN1-induced Dkk1 expression (Fig. 5E), resulting in accumulation of β-catenin (Fig. 5F). SB273005 and dasatinib also blocked Dkk1 induction upon CCN1 add-back ( Supplementary Fig. S7B, C).
Together, these results identify a novel signaling pathway involving CCN1 engagement of integrins α v β 3 /α v β 5 , leading to Src-mediated activation of YAP, which downregulates Wnt signaling in ISCs through Dkk1.

A regulatory loop between CCN1 and YAP during ISC homeostasis
CCN1 activates YAP through integrin-mediated activation of Src (Fig. 5). Interestingly, Ccn1 itself is a YAP target gene and frequently used as a YAP transcriptional readout 46 (Fig. 6A, B).
To determine whether YAP may act through CCN1, we examined the effect of CCN1 add-back in Yap ∆Lgr5 organoids with 4-OHT. 4-OHT-treated Yap ∆Lgr5 organoids also exhibited increased Lgr5+GFP and Ki67 as well as secretory cell lineage markers Lyz1 and Muc2 by immunohistochemistry and qPCR analysis (Fig. 6C, D). Importantly, CCN1 add-back to these organoids only promoted the reversal of differentiation phenotype with a reduction in Lyz1 and Muc2 expression similar to those in control organoids, but Lgr5+GFP and Ki67 levels were still elevated (Fig. 6C). Corresponding changes in the expression of marker genes were shown by qPCR analysis, along with an increase in the Notch target Hes1, but not the YAP target Dkk1, upon CCN1 add-back (Fig. 6D). Consistently, immunoblot analysis detected that CCN1 addback increased Jag1 and N1-NICD, but failed to block β-catenin accumulation without YAP (Fig.   6E). These results imply that CCN1 still can activate Notch signaling for ISC differentiation independent of YAP, but YAP is indispensable for CCN1 to control ISC proliferation. By contrast, a soluble Dkk1 protein efficiently lowered the expression of Lgr5+GFP, Ki67, and a Wnt target Axin2 in 4-OHT-treated Yap ∆Lgr5 organoids (Fig. 6F, G), thus confirming that YAP acts through Dkk1 to inhibit Wnt signaling and Lgr5+ ISC proliferation. Interestingly, the elevated expression of Lyz1 and Muc2 in 4-OHT-treated Yap ∆Lgr5 organoids was also reduced by the treatment of Dkk1 (Fig. 6G), suggesting that Wnt inhibition may influence Notch signaling in the context of Yap deletion. Together, these results indicate that the loss of Ccn1 expression upon Yap deletion may be responsible for the increased secretory cell differentiation, and CCN1 regulation of ISC proliferation depends on YAP in a regulatory loop ( Fig. 7 and Supplementary   Fig. S10).  (Figs. 1, 2). Especially, CCN1 regulation of ISC homeostasis is through direct binding to two α v integrins, α v β 3 and α v β 5 , through a non-RGD sequence located in its von

DISCUSSION
Willebrand factor type C (vWC) domain 54 . Ccn1 D125A/D125A knock-in mice expressing a mutant CCN1 unable to bind integrins α v β 3 /α v β 5 phenocopy mice with Ccn1 deletion in Lgr5+ ISCs (Fig.   2G, H). These in vivo results, together with organoid experiments using the CCN1-D125A mutant protein and the α v β 3 /α v β 5 antagonist SB273005 (Fig. 4), further establish that CCN1 acts through integrins α v β 3 /α v β 5 to regulate proliferation and differentiation of ISCs. Several integrins are reported to be enriched at the intestinal crypt base, suggesting their potential role in ISC regulation 55 . Indeed, conditional deletion of β 1 integrins in intestinal epithelial cells caused crypt hyperplasia and dysplasia, although specific ECM molecules involved were not identified 56 . By contrast, our study uncovers both CCN1 and integrins α v β 3 /α v β 5 as previously unknown regulators of ISCs and their specific activities and mechanisms of action in ISCs homeostasis. CCN1 is involved in ISC fate decision by regulating Notch signaling, whose blockage results in skewed differentiation into secretory cells 15 Fig. S3A) might render Notch1 activation dependent upon Jag1 (Figs. 3, 4).

The regulation of MFNG expression is not well understood and how it might be regulated upon
Ccn1 deletion is currently unclear. Recently, excessive Wnt signaling has been linked to MFNG deficiency, imposing Jag1-dependent Notch activation 35 . Similarly, we observed that wild-type intestinal organoids exhibited reduced MFNG expression upon the repeated treatment of CHIR99021, a glycogen synthase kinase 3 (GSK3) inhibitor/Wnt activator 60 ( Supplementary   Fig. S9). Moreover, a neutralizing anti-Jag1 antibody in combination with CHIR99021, but not by itself, diminished Notch activation and led to increased secretory cell differentiation ( Supplementary Fig. S9), reminiscent of phenotypes of Ccn1 deletion. In addition, enhanced secretory cell differentiation in organoids with loss of Ccn1 (Yap deletion; Fig. 6A), was blocked with Wnt signaling inhibition by Dkk1 treatment (Fig. 6G). Whether inhibition of Wnt signaling may restore MFNG expression and Notch signaling needs further examination. CCN1 regulates Wnt signaling through integrins α v β 3 /α v β 5 -FAK/Src-YAP axis ( Fig. 5 and Supplementary Fig. S10), modulating ISC expansion during homeostasis ( Fig. 2A, G).
Mechanistically, CCN1 engagement and clustering of integrins α v β 3 /α v β 5 lead to the recruitment of Src homology 2 (SH2) domain of Src and its conformational activation by phosphorylation at Y419 (Ref. 38 ). Active p-Src (Y419), in turn, phosphorylates YAP at Y357 (Fig. 5C), resulting in its stabilization and enhanced nuclear translocation ( Supplementary Fig. S6) Supplementary Fig. S6, 7), Ccn1 was also under the transcriptional control of YAP (Fig. 6A, B, and Supplementary Fig. S8A), as a known transcriptional target of YAP 46 . Thus, CCN1 and YAP amplify the activity of each other.
Interestingly, previous epithelial-specific Yap deletion studies in mice using a Villin Cre deleter suggested that YAP is dispensable for normal intestinal homeostasis 40,41,64  In summary, our finding of CCN1 as a novel matricellular niche factor that coordinately regulates ISC proliferation and differentiation in homeostasis through integrin-mediated signaling expands our understanding of stem cell biology. Given the importance of ISCs in injury repair and epithelial regeneration after damage, further investigation on whether CCN1 plays a role in such contexts through the regulation of ISCs is warranted and may uncover new insight into the regenerative biology of the intestine.

CCN1 proteins and reagents
Recombinant CCN1 and CCN1-D125A mutant proteins were produced using a baculovirus expression system in Sf9 insect cells and purified by ion-exchange or immuno-affinity chromatography 66 knock-in mice were obtained from Jackson Laboratory (#08875) 69 . Ccn1 flox/flox mice were constructed as described previously 70 . Yap floxl/flox mice were kindly provided by Dr. Duojia Pan 71 . Ccn1 ∆Lgr5 and Yap ∆Lgr5 mice were generated by crossing Ccn1 flox/flox or Yap floxl/flox mice with the Lgr5-EGFP-IRES-creERT2 mice. Ccn1 D125A/D125A knock-in mice were constructed as described 19 and crossed with the Lgr5-EGFP-IRES-creERT2 mice for detecting Lgr5+ ISCs in knock-in mice. All mice were housed in sterile static micro-isolator cages on autoclaved corncob bedding with water bottles. Both irradiated food and autoclaved water were provided ad libitum. The standard photoperiod was 14 hours of light and 10 hours of darkness. Both male and female mice ranging from 2 to 4 months old were used. For Cre induction, mice were daily injected (i.p.) with 100 µl of tamoxifen (Sigma-Aldrich, T5648) in corn oil at 100 mg per ml.

In vivo paracellular permeability assay
Ccn1 ∆Lgr5 mice at the post-injection (either corn oil or tamoxifen) day 10 were orally gavaged with fluorescein isothiocyanate (FITC)-dextran (MW 4000, 60 mg per kg; Sigma-Aldrich, #46944) 4 h before sacrifice. Blood was collected via intraorbital vein blood collection, and the fluorescence intensity in the serum was measured (excitation at 485 nm, emission at 520 nm) using a Victor3V plate reader (PerkinElmer). FITC-dextran concentrations were determined from a standard curve generated by a serial dilution.

Measurement of glucose concentrations
Ccn1 ∆Lgr5 mice at the post-injection (either corn oil or tamoxifen) day 10 were fasted overnight and treated either with saline (vehicle) or sucrose (3 g per kg) by oral gavage. After the indicated time. Blood samples collected from the tail vein were assayed with a commercial glucose meter (Contour, Bayer).

Intestinal organoid culture
Mouse intestinal organoids were established and maintained as described previously 72

RNA isolation and quantitative RT-PCR
Total RNA was purified from jejunum tissue or intestinal organoids using the GeneJET RNA purification Kit (Thermo Scientific, K0732) following the manufacturer's protocol. Total RNAs (2 µg) were reverse transcribed using Superscript Reverse Transcriptase III (Invitrogen, #18080044). Quantitative RT-PCR was performed with iCycler Thermal Cycler (Bio-Rad) using iQ SYBR Green Supermix (Bio-Rad, #1708880). PCR specificity was confirmed by agarose gel electrophoresis and melting curve analysis. A housekeeping gene (cyclophilin E) was used as an internal standard. Gene-specific primers used are listed in Supplementary Table 2.
Images were acquired using a Leica DM4000B microscope mounted with QI Click CCD digital camera (QImaging) and analyzed for measuring the length of villus and crypt using NIH ImageJ software (NIH). Alkaline phosphatase staining was performed using the NBT/BCIP Kit (Vector laboratories, SK-5400). Frozen intestinal sections (7 µm), fixed with ice-cold acetone, were probed with individual primary antibodies (listed in Supplementary Table 1) in immunofluorescence staining using Alexafluor488-conjugated goat anti-rabbit IgG or Alexafluor546-conjugated goat anti-rabbit IgG (Invitrogen, 1:500 dilution; each). DAPI (1 mg per ml) was used as a counterstain. Fluorescence images were acquired using an LSM 700 confocal microscope (Zeiss) and processed with Photoshop 2021 (Adobe).
For staining of whole-mount organoids, intestinal organoids were seeded in an 8-well glass-bottom chamber slide (Ibidi, #80827). The organoids were fixed in 4% paraformaldehyde 24 (PFA) for 10 min and washed in PBS (3×) at room temperature (RT). Following permeabilization with PBS containing 0.3% Triton X-100 (PBST) for 20 min RT and blocking in PBS containing 5% BSA (Sigma) for 1 hr at RT, organoids were incubated with primary antibodies (listed in Supplementary Table 1) in PBS containing 2% BSA for overnight at 4°C, followed by immunofluorescence staining, as described above.

In situ hybridization
Oligonucleotide probe for Ccn1 (5'-CTGCGGCTGCTGTAAGGTCTGCGCTAAACAACTCAA - For immunoblotting, immunohistochemistry, and in situ hybridization, representative images are shown. Each of these experiments was independently repeated at least three times.

Reporting summary
Further information on research design is available in the Nature Research Reporting Summary.

DATA AVAILABILITY
Data supporting the findings of this work are available within the main text and the Supplementary Information files. All data used for tabulated representation are provided in Source Data.  1, 2, 3, 7, and 8, Zo-1, -2, and -3). Data represent the mean±s.e.m (n=4 per group). * p<0.05; ** p<0.01; ns, not significant.