IL-22 initiates an IL-18-dependent epithelial response circuit to enforce intestinal host defence

IL-18 is emerging as an IL-22-induced and epithelium-derived cytokine which contributes to host defence against intestinal infection and inflammation. In contrast to its known role in Goblet cells, regulation of barrier function at the molecular level by IL-18 is much less explored. Here we show that IL-18 is a bona fide IL-22-regulated gate keeper for intestinal epithelial barrier. IL-22 promotes crypt immunity both via induction of phospho-Stat3 binding to the Il-18 gene promoter and via Il-18 independent mechanisms. In organoid culture, while IL-22 primarily increases organoid size and inhibits expression of stem cell genes, IL-18 preferentially promotes organoid budding and induces signature genes of Lgr5+ stem cells via Akt-Tcf4 signalling. During adherent-invasive E. coli (AIEC) infection, systemic administration of IL-18 corrects compromised T-cell IFNγ production and restores Lysozyme+ Paneth cells in Il-22−/− mice, but IL-22 administration fails to restore these parameters in Il-18−/− mice, thereby placing IL-22-Stat3 signalling upstream of the IL-18-mediated barrier defence function. IL-18 in return regulates Stat3-mediated anti-microbial response in Paneth cells, Akt-Tcf4-triggered expansion of Lgr5+ stem cells to facilitate tissue repair, and AIEC clearance by promoting IFNγ+ T cells. IL-22 induces IL-18 expression by intestinal epithelial cells. Authors show here that IL-18 is a key barrier maintenance factor during adherent-invasive E. coli invasion, inducing expression of anti-microbial genes in Paneth cells via Stat3, prompting IFNγ expression in T cells and triggering intestinal Lgr5+ stem cell expansion via Tcf4.

I mpaired host defence to intestinal pathogens has been implicated to associate with increased susceptibility to inflammatory disorders including inflammatory bowel disease (IBD) 1 . Certain Gram-negative bacteria, such as Salmonella typhimurium, Yersinia enterocolitica, or Adherent-invasive Escherichia coli (AIEC), are known to cause intestinal symptoms. Among them, Yersinia and AIEC are described as a pathogenic trigger for IBD and high prevalence of AIEC in patients with Crohn's disease (CD) has been reported [2][3][4] . In animal models, persistent infection with AIEC in the gut is associated with chronic inflammation and fibrosis, which CD8 + or IFNγ-producing immune cells are shown to be protective 5 . While Th1 and Th17-mediated immune responses are involved in the pathogenesis of Crohn's disease and in AIEC-infected mice 5,6 , convincing genetic evidence are still needed to elucidate how the frontline epithelial cells and their corresponding biological products, such as cytokines or antimicrobial peptides, are involved in AIEC host defence.
IL-22 is an ILC3 (innate lymphoid cell type 3) or Th17 signature cytokine and is generally known as a tissue protective cytokine which exclusively targets epithelial lineages 7,8 . IL-22 has a key role in epithelial barrier mainly because of its capability to induce epithelial production of anti-microbial and anti-inflammatory mediators, as well as to promote epithelial regeneration or wound healing 8 . As such, both ILC3 and Th17 cells are implicated in IBD pathogenesis 6,7 . Clinical relevance of IL-22 to IBD is well-established and IL-22 therapy is considered as a promising strategy for IBD treatment 8,9 . In IL-22-mediated signalling cascade, IL-18 is recently emerging as an important IL-22-induced and epithelium-derived cytokine, whose functionality is completely different from those established roles for IL-18 in inflammasome in myeloid cells 10 . For example, epithelial IL-18 not only relays IL-22-IL-22R signalling for host defence but is also required for IL-22 production during intestinal helminth infection 11 . Both protective and detrimental functions of IL-18 in the gut have been reported. Epithelial inflammasome-derived IL-18 was shown to activate an antimicrobial program which subsequently regulates microbial community to prevent intestinal inflammation 12 . A recent study reports that IL-22 and IL-18 promote intestinal epithelial migration for a rapid turnover leading to the protection against rotavirus infection 13 . In contrast, epithelial IL-18-IL-18R signalling is described to inhibit Goblet cell maturation which consequently might exacerbate colitis 14 . Regarding Paneth cells, while one study reports that IL-22stimulated organoids do not lead to enhanced Paneth cell frequency or Stat3 phosphorylation 15 , new evidence show that IL-22 promotes the production and secretion of Lysozyme, or the induction of specific markers (Lysozyme, MMP7), from Paneth cells in organoids 16,17 . In addition, IL-22R signalling in Paneth cells contributes to their maturation and full host protection against Salmonella infection 18 . Nevertheless, a role for IL-18 in Paneth cells or epithelial stem cells has not been studied. It is known that, in the milieu of IL-12, IL-18 is a potent IFNγ inducer in Th1 cells, suggesting that the IL-18-IFNγ axis could potentially contribute to the pathogenesis of Crohn's disease 19 . Supporting this notion, levels of IL-18 in serum or mucosal epithelial biopsies were significantly higher in CD patients and Il-18 gene polymorphism is linked to increased susceptibility to Crohn's disease 20,21 . While the interplay between IL-22 and IL-18 appears cross-regulated during inflammation or host defence, evidence of a specific role for IL-18 in epithelial barrier, with respect to what are definitive target cell types of IL-18 during CD-related host defence, is missing.
Here we show, using a clinical isolate of AIEC from CD patients, a coordinated response of IL-22 and IL-18 to intestinal AIEC infection. While these two cytokines show integrated functions in regulating Paneth cells and Goblet cells, they surprisingly exert distinct regulatory functions and pathways towards epithelial stem cells. At the molecular level, we identify the requirement of Stat3 in epithelial transcriptional induction of IL-18 by IL-22, of Akt-Tcf4 in IL-18-mediated transcriptional activation of Lgr5, and of Stat3 in IL-22/IL-18-induced Paneth cell functionality. A full coordinated IL-22-initiated IL-18 response circuit to enforce mucosal host defence against Crohn's AIEC is proposed and discussed.

Results
AIEC infection triggers an early response of stem cells and Paneth cells. Pathogenic microbes in the intestine have been implicated as a trigger for IBD 4,22 . Adherent-invasive E. coli (AIEC) is abundantly identified in the ileum mucosa of CD patients and a clinical isolate of AIEC has been shown to cause chronic inflammation and fibrosis in mice 3,5 . As functionality of stem cell-mediated epithelial regeneration and Paneth celldirected anti-microbial peptide (AMP) production is critical for epithelial barrier against mucosal infection 23 , we asked how Crohn's AIEC initiates the immune response of these two frontline cell types. As noted, oral gavage with an AIEC isolate NRG857c caused a rapid epithelial apoptosis in ileal epithelium 5 , marked by an increase of active Caspase-3 + crypts at the early stage of infection ( Fig. 1a, b), and a corresponding epithelial regeneration evidenced by an increase of CD24 -/low Ki67 + proliferating crypts which mostly represent transit-amplifying (TA) cells (Figs. 1c, 2b) 24,25 . Correlated to TA cell response, the stem cell response, marked by an increase of CD24 -/low Lgr5 + crypts 24 , or mRNA expression of stem cell marker Lgr5, Ascl2, and Olfm4 26 , was also upregulated at the early stage ( Fig. 1d-f and Supplementary Fig. 1a-b). This indicates that epithelial regeneration is also rapidly and actively triggered as a repair host defence against invasive and destructive property of AIEC. Furthermore, the response of Paneth cells (PC), which was marked by the missing CD24 + Lysozyme + subset in ileum crypts of Paneth cell-deficient (Defa6-Cre + Rosa26-LSL-DTA or PC Δ ) mice and by the PC marker Lysozyme or Cryptdin (Fig. 1g-h and Supplementary Fig. 1a, c) 24,27 , was also actively triggered at the early stage. As noted, PC-deficient mice were more susceptible to AIEC infection (Fig. 1i). Therefore, mucosal infection of Crohn's AIEC causes an early and robust response program of epithelial stem cells and Paneth cells, whose full functionality might be important to prevent chronic inflammation or fibrosis 5 .

IL-22-Stat3 axis upregulates Paneth cells in response to AIEC.
IL-22 is associated with IBD because of its capability to promote regeneration and anti-microbial function of epithelial barrier 8 . As such, we asked how IL-22 is involved in AIEC host defence. As noted, IL-22 is abundant in the lamina propria (LP) compartments ( Supplementary Fig. 1d). At the early stage of infection, IL-22 was rapidly induced in the ileum LP (Fig. 2a), indicating that IL-22 initiates an innate response to AIEC which is known to target the ileum. Next, at the steady state or during AIEC infection, loss of Il-22 in mice caused a significant decrease in Ki67 + proliferating TA cells (Fig. 2b-c), which is consistent to previous studies where injection of IL-22 into mice increases epithelial proliferation and expands TA compartments 15,17 . Of note, Ki67 + proliferating crypts are also reduced in epithelium-specific Stat3 conditional knockout (Vil-Cre + Stat3 f/f ) mice 28 , indicative of a role for IL-22-Stat3 signalling in epithelial regeneration during mucosal infection. We next tested the controversial role of IL-22 in the regulation of Paneth cells. While one study shows undetectable IL-22R expression and IL-22-induced Stat3 phosphorylation in Paneth cells 15 , a recent study provides genetic evidence that IL-22R signalling in Paneth cells is required for maturation and full protection against Salmonella typhimurium 18 3a) and analysis of ileum lamina propria in AIEC-infected Il-22 −/− mice showed an over 50% reduction of IFNγ-producing CD8 + T cells (Fig. 3b, Supplementary Fig. 2a), indicating that IL-22 is functionally linked to the induction of IFNγ + CD8 + T cells during AIEC infection. Among potential upstream regulators of IFNγ, IL-18, in the presence of IL-12, is a potent IFNγ inducer in CD8 + T cells 29 . Indeed, we found that IL-18 and IL-12 robustly and preferentially promote IFNγ + CD8 + T cells in mesenteric lymphocytes (Supplementary Fig. 2b-d), raising the possibility that IL-18 might relay IL-22 signalling to downstream IFNγ response for AIEC host defence. Supporting this notion, IL-22 has been shown to maintain epithelial homeostasis of IL-18 and the IL-22-IL-18 axis contributes to host defence against helminth 11,29 . As noted, IL-18 expression in Il-22 −/− ileum crypts was reduced at the steady state or during AIEC infection (Fig. 3c-d). Furthermore, as there was no difference in IL-12 expression in ileum lamina propria between wildtype and Il-22 −/− mice ( Supplementary Fig. 2e), it is likely that IL-22, but not IL-12, upregulates IL-18 leading to IFNγ response. We also found that epithelium-specific Stat3 conditional deletion in mice also causes a decrease of IL-18 in ileum crypts (Fig. 3c), suggesting that the IL-22-Stat3 signalling is linked to downstream IL-18-IFNγ cascade. To gain more insights, ileum organoids were in vitro stimulated with recombinant IL-22 or IL-18, and analyzed for IL-18 upregulation. The results showed that IL-22 induces IL-18 in a Stat3-dependent manner but IL-18 itself does not induce IL-18 ( Fig. 3e-f). As noted, epithelial Stat3 deletion caused more reduction of IL-18 after IL-22 stimulation, compared to Paneth-cell Stat3 deletion. This might be due to the fact that IL-22R is more expressed in CD24 low Lysozymeepithelial subset than in Lysozyme + Paneth cells (Fig. 3g, Supplementary Fig. 2f).

CFU (AIEC-d6)
As Stat3 is a transcription factor, we next asked whether IL-22 regulates Stat3 binding to the Il-18 promoter for transcriptional upregulation. To this end, we analyzed and identified five (marked as P1~P5) STAT consensus binding sites within the mouse Il-18 promoter ( Supplementary Fig. 3a). Next, chromatin immunoprecipitation (ChIP) assay, with a specific anti-phospho-Stat3 Tyr705 (active form of Stat3) antibody for immunoprecipitation, was performed in IL-22-stimulated ileum crypts, to determine whether and where IL-22 promotes phospho-Stat3 binding to those putative STAT binding sites ( Supplementary  Fig. 3b). As noted, we found that IL-22 induces phospho-Stat3 binding to the P3, P4, and P5 sites within the Il-18 promoter (Fig.  3h). We also analyzed the human Il-18 promoter and performed the ChIP assay in human colon epithelial HT-29 cells.  Fig. 4a), suggesting that epithelium-derived IL-18 might have a more important role over myeloid cell-derived IL-18 which is enriched in the lamina propria. Similar to an early induction of IL-22 in ileum lamina propria, IL-18 was also rapidly upregulated in ileum crypts at the early stage (Fig. 4a). The same induction kinetics provides a functional link that IL-22-initiated IL-18 is triggered as an innate response. Notably, in ileum organoids, both AIEC and IL-22 can directly induce IL-18 ( Supplementary Fig. 4b). IL-18, like IL-22, is also capable of inducing Paneth cell-related AMP (Itln1, Ang4) and Paneth cell-specific AMP (Lysozyme, Cryptdin), all in a Stat3-dependent manner ( Fig. 4b-c, Supplementary Fig. 4c-d).
As noted, recombinant IL-22 or IL-18 promoted Ki67 + organoid proliferation in a Stat3-dependent manner (Fig. 6a) (Fig. 6d). Regarding morphological changes of organoids, we found that IL-22 shows a major effect to promote Stat3mediated organoid size but IL-18 primarily promotes organoid budding in a Stat3-independent manner (Fig. 6e). While one study suggests that IL-22 promotes organoid size, but not   Fig. 7a). We further confirmed that IL-22 promotes organoid size independent of Paneth cells (or Stat3 in Paneth cells) and that IL-18 promotes organoid budding in a dose-dependent manner and the induction partially depends on Paneth-cell Stat3 (Fig. 6f, Supplementary Fig. 7b). As Paneth cells provide crucial niche factors to support nearby stem cells and an organoid bud develops into a crypt-like structure 32    inhibits stem cell genes (by Ascl2/Olfm4 mRNA) slightly better in the absence of IL-18 (because IL-22-induced IL-18 promotes stem cells in the same culture), it still promotes organoid budding and size in Il-18 −/− organoids ( Fig. 6g-h). These discrepancies could be due to the limitation of organoid culture where many growth factors are added for optimal growth condition that may directly affect the function of IL-22 or IL-18. For example, recombinant R-spondin 1 is added to provide enhanced Wnt signalling for stemness 33 and Jagged-1 peptide (a Notch agonist) is used to enhance Notch signalling 34  IL-18-Akt-Tcf4 axis upregulates Lgr5 + stem cells. While our gain-of-function studies in IL-22 or IL-18-stimulated organoids support a role for IL-22 and IL-18 in epithelial proliferation (Fig.  6), loss-of-function study in genetic knockout mice is also crucial to elucidate the in vivo function of IL-22-Stat3 or IL-18 signalling in barrier homeostasis. To this end, we explored how loss of Il-22, Il-18, or epithelial Stat3 in mice affects stem cells at the steady state or during AIEC infection. As noted, all knockout mice showed compromised Lgr5 + stem cells in the context of homeostasis or AIEC host defence (Fig. 7a-c, Supplementary Fig. 8a-b), providing in vivo evidence that the IL-22-Stat3 axis or IL-18 signalling is a key regulator of stem cells. By immunofluorescence analysis of littermate mice, we also found that Olfm4 + crypt base columnar (CBC) stem cells are greatly reduced in both Il-22 −/− and Il-18 −/− crypts at the steady state or during AIEC infection (Fig. 7d) 26,35 . Furthermore, correlated to mRNA data and organoid budding (Fig. 6d-e), we confirmed, by flow cytometry analysis, that IL-18 promotes Lgr5 + stem cells independently of Stat3 in ileum organoids (Fig. 7e). Therefore, while Stat3 activation has been shown to contribute to Lgr5-β-catenin signalling for stemness and IL-18 indeed activate Stat3 in crypts (Fig. 4d) 36,37 , our results exclude a role for IL-18-Stat3 signalling in Lgr5 + stem cell upregulation. To further elucidate the underlying mechanism, we asked how IL-18 is involved in Wnt-mediated signalling for stemness, which is induced via the Lgr5-Frizzled-Lrp5/6 complex leading to β-catenin/Tcf4 activation 32 . T-cell factor 4 (Tcf4) is a key transcription factor downstream of Wnt signalling and is indispensable for epithelial stem cell compartments in vivo 38,39 .
We analyzed the mouse Lgr5 promoter and identified six (marked as P1~P6) putative TCF4 consensus binding sites ( Supplementary  Fig. 8c). Further analysis of IL-22 or IL-18-stimulated ileum organoids by the ChIP assay, with a specific anti-Tcf4 antibody for immunoprecipitation, revealed that IL-18, but not IL-22, specifically promotes Tcf4 binding to the P1 region within the Lgr5 promoter (Fig. 7f). Next, as Akt activation has been linked to the survival of Lgr5 + organoids 40 , we tested in colon epithelial CMT93 cells by Western blotting and found that IL-18 robustly induces Akt phosphorylation at Ser 473 and a corresponding induction of Lgr5. Inhibition of Akt activity with an inhibitor MK-2206 abolished IL-18-induced Akt activation and reduced Lgr5 expression as well (Fig. 7g) 41 , indicating that the IL-18-Akt-Lgr5 pathway is acting on stemness induction. To further confirm this, a ChIP assay in IL-18/Akt inhibitor-treated CMT93 cells was performed. The result showed that the Akt blockade indeed abolish the IL-18-induced binding of Tcf4 to the P1 region within the Lgr5 promoter (Fig. 7h), supporting that it is the IL-18-Akt-Tcf4-Lgr5 axis that is crucial for stem cell upregulation.  Fig. 9a). However, the percentage of Muc2 + Goblet cells in both knockout mice was reduced at the steady state, as was also observed in Muc2 mRNA levels in ileum epithelial cells isolated from naïve or AIEC-infected mice (Fig. 8a, Supplementary Fig. 9b). Correlated to mRNA levels, by immunofluorescence and H&E staining, we confirmed that loss of Il-22 or Il-18 in mice causes a homeostatic reduction of Muc2 + Goblet cells (Fig. 8b and Supplementary Fig. 9c, upper panels). As noted, the reduction at the steady state was not due to aberrant apoptosis of Goblet cells by TUNEL staining ( Supplementary  Fig. 9d, upper panels). Intriguingly, AIEC infection promoted drastic mucin secretion in wild-type mice but this process was almost abolished in Il-22 −/− or Il-18 −/− mice ( Fig. 8b and Supplementary  Fig. 9c, lower panels). Further analysis of IL-18-injected Il-22 −/− mice confirmed that IL-18 indeed promote mucin secretion from Goblet cells during AIEC infection (Fig. 8c). In contrast, IL-22 injection into Il-18 −/− mice failed to promote mucin secretion,   which is reminiscent of the results that IL-22 injection also failed to restore Paneth cell or IFNγ response in Il-18 −/− mice during AIEC infection (Fig. 5e-f). Again, our results argue that it is actually IL-22induced epithelial IL-18 that is upregulating Goblet cells for AIEC host defence. Furthermore, in ileum organoids, it appears that both IL-22 and IL-18 promote Muc2 mRNA levels in a Stat3-dependent manner and that IL-22 induces Muc2 mRNA independent of epithelial IL-18 (Fig. 8d, Supplementary Fig. 9e). Of note, while there was increased cell apoptosis by TUNEL staining in both Il-22 −/− and Il-18 −/− mice during AIEC infection, the percentage of Caspase-3 + UEA1 + Goblet cells was not increased in Il-18 −/− mice ( Supplementary Fig. 9d, f). Overall, these results indicate that IL-18 regulates mucin production and secretion but not survival of Goblet cells. Taken together, integrated and differential regulation of epithelial barrier by IL-22 and IL-18, mostly based on organoid culture, is summarized (Fig. 8e). We conclude that mucosal AIEC infection triggers an early IL-22 response cascade which subsequently initiates an IL-18 response circuit. This IL-18-mediated circuit promotes IL-22-redundant and non-redundant immunity by boosting proliferating stem/TA cells, anti-microbial Paneth cells, mucinproducing Goblet cells, and IFNγ-producing T cells.

Discussion
Intestinal microbiota dysbiosis is an important factor in the etiology of Crohn's disease (CD) and high prevalence of adherent-invasive Escherichia coli (AIEC) in the mucosa of CD patients has been consistently reported 44,45 . Both IL-22 and IL-18 contribute to active CD and represent a promising therapeutic potential 9,19,21,29 . Here, our work provides clinical relevance of these two cytokines to host defence against Crohn's AIEC, by revealing a crucial role of the IL-22-IL-18 axis and a coordinated IL-22-initiated IL-18 response circuit at the frontline barrier. The IL-22 response cascade and IL-18 response circuit exert redundant and non-redundant immunity for host defence. At the cellular level, we show that IL-18 regulates Ki67 + proliferating transit-amplifying (TA) cells, Lgr5 + stem cells, and anti-microbial Paneth cells at the steady state and during AIEC host defence, providing the first genetic evidence that IL-18 is a bona fide regulator of stem cells and Paneth cells. At the molecular level, we reveal that IL-22 directs a pleiotropic IL-18 response circuit, which subsequently deploys an IL-18-Stat3-AMP (targeting Paneth cells) pathway, an IL-18-Akt-Tcf4-Lgr5 (targeting stem cells) pathway, and an IL-18-IFNγ (targeting CD8 + T cells) pathway, to boost innate immunity and connect innate to adaptive immune responses. Our rescue experiments in mice further provide in vivo evidence, showing a direct link of the IL-22-IL-18-IFNγ axis and the IL-22-IL-18-Paneth cell axis to AIEC host defence. As IL-22/IL-18-mediated therapeutic strategy is under development for treating inflammatory diseases and cancers 29 , our work here provides new insights of how the interplay of IL-22 and IL-18 contributes to cross-regulated immune response.
Increased IL-18 levels in serum and mucosal biopsies are correlated to Crohn's disease and IL-18 upregulation is reported to be a feature of CD 19,20 . As a potent IFNγ inducer in T cells, IL-18 could be an ideal target for blocking inflammatory Th1mediated IFNγ response in CD patients 19 . Intriguingly, mucosal biopsies of IBD patients show more abundant IL-18 expression in epithelial cells over lamina propria mononuclear cells 46 , suggesting that the function of IL-18 in epithelial cells might be underestimated, especially in wound healing and anti-microbial immunity which are key processes for tissue recovery and host defence. In this regard, we reveal that IL-18 activates Akt-Tcf4 to promote Lgr5 + stem cell expansion or tissue regeneration, which is highly relevant to CD pathogenesis and Lgr5-mediated intestinal tumorigenesis 47 . As a transcription factor, Tcf4 is a key regulator downstream of Wnt/β-catenin signalling 39 . In the gut, Tcf4 controls mucosal homeostasis and immunity as it directly binds to and activates the promoters of stem cell marker Lgr5 and Paneth-cell marker α-defensin 39,48 . Global deletion of Tcf4 in mice causes absence or abnormality of proliferating crypt compartments leading to neonatal lethality within 24 hr 38 , phenotypically similar to global knockout of Lgr5 49 . Conditional Tcf4 ablation in gut epithelium causes loss of proliferating crypt cells, as well as gradual depletion of stem cell compartment and nearby Paneth cells 50 . These facts explain why Tcf4 is a pronounced risk factor for Crohn's disease 51 , as is also evidenced by a high correlation of genetic variants of Tcf4 and the decrease of Paneth cell-specific α-defensin HD5 to CD patients 52,53 . Therefore, the IL-18-Akt-Tcf4-Lgr5 pathway may represent a novel and key repair mechanism in the epithelium to prevent CD pathogenesis. In addition, while activation of the IL-18-IFNγ axis in T cells appears detrimental to the mucosa, we found the IL-18-mediated IFNγ response in CD8 + T cells is also essential for clearance of AIEC 5 , which is highly associated with ileum mucosa in CD patients 3 . Together, our results support a protective role for IL-18 in host defence, where IL-18 is promoting IFNγ + T cells with IL-12, stem cells via Tcf4, and Paneth cells via Stat3. As such, concerns should be taken when the IL-18 blockade, which can suppress epithelial barrier function and T-cell-mediated host defence, is developed for treating mucosal inflammation such as Crohn's disease.
Our study also provides a comprehensive dissection of the controversial role for IL-22 and the unexplored function for IL-18 in stem cells 15,17,54 . While accumulating evidence show that IL-22 promotes proliferative transit-amplifying (TA) compartments but suppresses Lgr5 + stem cells in vivo and in vitro [15][16][17] , there have been no reports of a role for IL-18 in stem cells. In organoid culture, we found that both IL-22 and IL-18 promote Stat3independent budding and Stat3-dependent size expansion. As an organoid bud develops into a crypt-like structure that usually contains 4~6 stem cells 32,55 , we reason that the number of organoid buds is a good indicator of stemness. Intriguingly, IL-18-induced organoid budding and increase in size are correlated to IL-18-induced mRNA upregulation of stem cell markers (Lgr5, Ascl2, Olfm4), but similar IL-22-induced morphological changes can not be attributed to IL-22-induced mRNA suppression of stem cell markers. This discrepancy immediately argues that organoids may not be a truly untouched sample after cultured in an optimized condition, which usually provides enhanced Wnt/ Notch signalling for growth that potentially may interfere with IL-22-mediated signalling 17 . However, this also explains why IL-22-mediated organoid budding and size expansion does not require IL-18, which is supported by our identification of nonredundant IL-18-Akt-Tcf4-Lgr5 pathway for stemness. In contrast, by flow cytometry and immunofluorescence analyses of fresh "untouched" crypt samples for stem cells and TA cells, we   16,17 . While an earlier study shows IL-22R is not detectable on Lysozyme + Paneth cells and therefore has no effect on Paneth cell frequency in organoids 15 , new genetic evidence in mice indicate that IL-22R signalling in Paneth cells is crucial for their maturation, which consequently might affect microbiota colonization and antibacterial immunity 18 . We extend these observations by showing that, first, both IL-22 and IL-18 indeed induce Stat3 activation in Paneth cells and both require Stat3 to promote Lysozyme + Paneth cells. Secondly, IL-18R is highly expressed in CD24 + Paneth cells but IL-22R is more enriched in CD24 -/low subsets which mostly contain stem cells and TA cells, suggesting that IL-22 and IL-18, while functionally linked, may preferentially target distinct epithelial subsets. Thirdly, different from the action mode in stem cells, IL-22 absolutely requires IL-18 to promote Paneth cells in vitro and in vivo. Lastly, in sorted Paneth cells, we show that IL-18 is able to directly promote AMP production from Paneth cells. Therefore, the identification of the IL-22-IL-18-Stat3 axis in Paneth cell functionality could impact the therapeutic strategy of targeting Paneth cell-mediated host defence in Crohn's disease (also called Paneth disease) 56 .
In addition to differential expression of the receptor, the requirement for signalling components downstream of IL-22R and IL-18R could also be different in epithelial subsets. IL-22Rassociated Jak1/Tyk2 activates Stat3 when IL-22 signalling is triggered 57 . It is unclear whether the same machinery is involved in IL-18-mediated Stat3 activation in Paneth cells but it is Akt that relays IL-18 signalling to activate Tcf4-Lgr5 in stem cells. Canonically, the IL-18 receptor complex interacts with Myd88, which is capable of inducing downstream NF-κB via IRAK1-4, or MAPK via TRAF6 10 . Therefore, the differential receptor expression pattern and requirement for different signalling components in IL-22 vs IL-18 signalling might highlight the importance and uniqueness of IL-18 response circuit, downstream of the IL-22 response cascade, in the homeostasis of epithelial barrier and during AIEC host defence. Wild-type controls were generated by intercrossing of heterozygous knockout mice. Animals were bred separately, not co-housed, and maintained in a specificpathogen-free (SPF) facility at a relative humidity 50 ± 10%, 20-26°C, and in 12 h dark/light cycles (08:00-20:00 light). Experiments were performed on mature animals (8-12 week-old) in both male and female mice, unless otherwise indicated. Animal care and experimental protocols (Protocol ID: 17-05-1092) have been approved by the Institutional Animal Care and Use Committee (IACUC) at the Institute of Biomedical Sciences, Academia Sinica. Ethical compliance has been observed in all animal studies. Dr. John T. Kung is the chairperson of IACUC and Ethics Committee in Academia Sinica, Taiwan.
AIEC infection and in vivo rescue experiment. Mice (8-12 week-old) were orally given 40 mg of ampicillin in a total volume of 200 μl one day before oral gavage with 2×10 9 colony forming units (CFU) of adherent-invasive E. coli strain (AIEC) NRG875c (O83:H1) (a gift of Dr. Brian K. Coombes 5 ) in a total volume of 200 μl per mouse. The bacteria were prepared by sharking at 37°C overnight in LB broth. The concentration of bacteria was measured by serially diluted and plated each inoculation culture to confirm the colony-forming units (CFU) administrated. Mice were euthanized at day-3 or day-6 for desired experiments, when they usually show no detectable pathology by H&E staining which is consistent with the original report 5 . Some mild alterations of Paneth cell or Goblet cell numbers can be microscopically observed during AIEC infection. Feces were collected into 1 ml cold PBS buffer and homogenized by a vortexer. The ileum, cecum, and colon were collected into a 50 ml tube containing 10 ml cold PBS buffer and homogenized using the tissue homogenizer (OMNI, #TH115) with plastic homogenizing probes (OMNI, #30750H-S). Feces and tissue homogenates were serially diluted and plated on LB agar containing ampicillin to select for AIEC NRG875c. . HRP-conjugated secondary antibodies (1:10000 dilution) for immunoblotting were from Jackson ImmunoResearch unless otherwise mentioned. The following secondary antibodies were used: goat antimouse IgG (#115-035-003), goat anti-rabbit IgG (#111-035-003). Western blots were scanned with ImageQuant LAS 4000 mini equipment and its software. The intensity of each band was quantified by ImageJ (1.53k) software. The following Fig. 6 Differential role of IL-22 and IL-18 in organoid culture. a Flow cytometry analysis of IL-22 or IL-18-stimulated ileum organoids, derived from the indicated mice, for Ki67 + proliferating cells. Mean fluorescence intensity (MFI) of Ki67 + organoids is shown. b Immunofluorescence analysis of ileum crypts for Ki67 + proliferating cells in the indicated naïve or AIEC-infected mice. Crypt base columnar (CBC, outlined with a solid line) stem cell and the above transit-amplifying (TA, outlined with a dashed line) compartments are indicated. c Flow cytometry analysis of ileum crypts for CD24 -/low Ki67 + proliferating cells in the indicated AIEC-infected mice. d Quantitative real-time PCR analysis of IL-22 or IL-18-stimulated ileum organoids, derived from the indicated mice, for stem cell marker Lgr5, Ascl2, and Olfm4. e Quantification of the number of buds per organoid and size in IL-22 or IL-18-stimulated ileum organoids, derived from the indicated mice. "n" indicates the number of images taken from organoids derived from four mice in each group. f Quantification of the number of buds in IL-18-stimulated wild-type ileum organoids. "n" indicates the number of images taken from organoids derived from three mice per group. g Quantitative real-time PCR analysis of IL-22-stimulated ileum organoids, derived from the indicated mice, for stem cell marker Lgr5, Ascl2, and Olfm4. h Quantification of the number of buds per organoid and size in IL-22-stimulated ileum organoids, derived from the indicated mice. "n" indicates the number of images taken from organoids derived from four mice per group. Each symbol in bar graphs represents an ileum crypt sample (c) or organoid culture (a, d, g), derived from one mouse. Data shown are representative (b) or combined (a, c-h) results from two independent reproducible experiments. Statistical significance is indicated using unpaired two-tailed t test (c), One-way ANOVA with Sidak's multiple comparisons test (f), or Two-way ANOVA with Tukey's multiple comparisons test (a, d, e, g, h). Data are presented as mean ± SD. Source data are provided as a Source Data file.   Tissue cytokine analysis by ELISA. After cleaning and cut open longitudinally, ileum fragments were further cut into 1-2 mm pieces, transferred to a 24-well plate with 100 mg ileum/1 ml/well in RPMI buffer containing 10% fetal bovine serum (FBS) (Gibco #10437-028), 100 U/ml of penicillin, 100 μg/ml of streptomycin (Gibco #15140-122), and 10 μl/ml Gentamicin (Gibco #15750-078), and incubated in a 37°C humidified 5% CO 2 incubator. After overnight (16 hr) incubation, supernatants were harvested and analyzed for protein levels by mouse IL-22 ELISA kit (ABclonal #RK00108) or mouse IL-18 ELISA kit (ABclonal #RK00104) according to instructions.
Isolation, culture, and analysis of ileal lamina propria (LP) cells. To isolate intestinal lamina propria cells, the described protocol was performed with some modifications 60 . Mice (8)(9)(10)(11)(12) week-old) were euthanized with CO 2 and the abdomen was cut open to separate~12 cm of distal small intestine (whole Ileum). Peyer's Patches and fat tissues were removed and ileum fragments were cut open longitudinally and washed with PBS. The fragments were further cut into 2 cm pieces and then placed into 50 ml falcon tubes and washed vigorously by shaking in Hank's Balanced Salt Solution (HBSS) until the supernatant was clear. Ileum pieces were incubated with 40 ml of warm depletion buffer (1 mM EDTA in 1X PBS) in 100 ml glass bottles and stirred at 500 g at 37°C for 20 min, then washed in 1X PBS by hand shaking vigorously in a 50 ml tube to ensure removal of epithelial cells. Next, ileum pieces were minced into 1 mm by a scissor in an eppendorf with 1 ml of digestion buffer [2% FBS, 15 μg/ml Liberase (Roche, #05401119001), 50 μg/ml DNase I (Sigma, #DN25) in RPMI medium (Gibco, #31800022)], then transferred to a 50 ml flask with 9 ml of digestion buffer and put for stirring at 500 rpm at 37°C for 30 min. The supernatant was filtered through a 70 μm strainer (BD Biosciences #352350) into a new 50 ml tube on ice and the rest of pieces were homogenized by a 18 G needle then again stirred in 10 ml of digestion buffer for another 30 min. The supernatant was centrifuged at 600 g for 6 min at 4°C and the pelleted LP cells were resuspended in complete RPMI medium (10% fetal bovine serum, 100 U/ml of penicillin and 100 μg/ml of streptomycin in RPMI medium) for cell counting. The resulting cells were split into three aliquots for RNA extraction, unstimulated, or stimulated culture in a V-bottom 96-well plate for flow cytometry analysis of cell composition or cytokine production. After 4 hr stimulation of IL-23 (50 ng/ml, eBioscience, #14-8231), PMA (50 ng/ml, sigma, #P8139), and Ionomycin (800 ng/ml, sigma, #I9657), the cells were briefly resuspended by pipetting before spinning down at 1,500 rpm for 5 min at 4°C and then washed once with staining buffer (2% FBS in 1X cold PBS). 50 μl staining buffer containing blocking antibody (anti-CD16/32) was added to cells for 10 min before performing surface staining for 20 min in dark at 4°C. For intracellular staining, the cells were fixed and permeabilized in BD Fixation/Permeabilization Solution (BD, #554714) for 30 min at 4°C, washed once with 200 μl of BD Perm Wash Buffer (BD, #554714), and followed by centrifugation at 1,800 rpm, 6 min at 4°C. The cells were resuspended with intracellular staining antibody in 1X Wash Buffer and stained for 30 min in dark at 4°C. After washing in staining buffer, the cells were resuspended in 200 μl staining buffer and immediately analyzed by BD LSR-II flow cytometer.
Intestinal crypts isolation for oganoid culture. To isolate the crypts, the described protocol was performed with some modificatins 61 . The clean ileum fragments were cut into 2 cm pieces and then placed into 50 ml falcon tubes and washed vigorously by shaking in Hank's Balanced Salt Solution (HBSS) until the supernatant was clear. To dissociate the crypts from ileum, fragments were placed in a 50 ml tube containing 10 ml of 30 mM EDTA in PBS buffer and shaked at 100 rpm in 37°C for 5 min. The PBS-EDTA buffer was aspirated, replaced with cold wash buffer [1X PBS pH 7.4, 1X penicillin/streptomycin, 50 μg/ml Gentamicin, 0.1% bovine serum albumin (BSA)], and sharked vigorously with a vortexer 4 times (2 s at a time) to remove villi and epithelial debris. After that, the supernatant was aspirated and 10 ml cold wash buffer was added to the fragments and vortexed for 8 times (2 s at a time). The supernatant containing crypts was collected and transferred to a new 50 ml falcon tube pre-coated with 5% FBS. This step was repeated and each successive fraction was collected and an aliquot was examined under a phase-contrast microscope for the presence of intact intestinal crypts and lack of villi. The fractions with intact ileum crypts were pooled together and filtered through a 70 μm cell strainer (BD Biosciences #352350) into a 50 ml FBS-precoated tube to remove any debris. To remove single cell contamination from the heavier epithelial crypts, the crypts were pelleted by centrifuged at 80 g for 5 min at 4°C. The supernatant was aspirated and crypt pellet was resuspended in 1-2 ml wash buffer, 100 μl was placed on a petri dish and the crypts were counted under a phase-contrast microscope.
Ileum organoid culture. To culture ileum organoids, a total of~500 freshly isolated ileum crypts per well were mixed with 50 μl basement membrane matrix growth factor reduced Matrigel (BD Biosciences #356231) containing 1 μM Jagged-1 peptide (Notch Agonist) (AnaSpec #AS-61298) and plated in 24-well plates. The Matrigel was polymerized for 10 min at 37°C incubator and 500 μl WENR growth media was added on top of the Matrigel. The ENR growth medium was prepared by mixing basal culture medium (BCM) [advanced DMEM/F-12 supplemented with 100 U/ml penicillin/100 μg/ml streptomycin, 50 μg/ml Gentamicin, 2 mM GlutaMAX, 10 mM HEPES (Gibco #15630-080), 1 mM N-acetyl-L-cysteine (Sigma-Aldrich #A9165), 1X B-27 Supplement (Gibco #17504-044), 1X N-2 Supplement (Gibco #17502-048), and 1% BSA (Gibco #15260-037)] with WNR conditioned medium in a 1:1 ratio, and 50 ng/ml murine Epidermal Growth Factor (mEGF) (Gibco #PMG8041). To maximize early growth of developing primary organoids from crypts, the WENR growth medium was supplemented with 10 μM Rho-associated protein kinase (ROCK) inhibitor, Y-27632 (Sigma-Aldrich #Y0503), 0.5 μM transforming growth factor (TGF)-β type I receptor inhibitor, and A-83-01 (R&D systems #2939) for the first 2 days of the culture. The culture medium was replaced every 2 days with ENR growth medium. For mRNA analysis Fig. 7 IL-18-Akt-Tcf4 signalling upregulates Lgr5 + stem cells. a-c Flow cytometry analysis of CD24 -/low ileum crypts, isolated from the indicated uninfected or AIEC-infected mice at d6, for Lgr5 + stem cells. An isotype antibody was included to indicate the specificity of the Lgr5 antibody. d Immunofluorescence analysis of ileum crypts for Olfm4 + stem cells in the indicated uninfected or AIEC-infected mice at d6. Crypt base columnar (CBC, outlined with a solid line) stem cell and the above transit-amplifying (TA, outlined with a dashed line) compartments are indicated. e Flow cytometry analysis of IL-18-stimulated ileum organoids, derived from the indicated mice, for Lgr5 + stem cells. f Chromatin immunoprecipitation (ChIP) and PCR analyses of IL-22 or IL-18-stimulated wild-type ileum organoids for the binding of Tcf4 to the mouse Lgr5 promoter regions (P1~P6). Fold induction is calculated based on PCR signal strength of Tcf4-bound Lgr5 promoter P1 region before and after IL-22 or IL-18 stimulation. g Western blot and flow cytometry analyses of IL-18/Akt inhibitor-stimulated CMT93 cells for Akt phosphorylation and Lgr5. Quantification and the ratio (pAkt to total Akt, Lgr5 to β-actin) of protein bands are indicated. h ChIP and PCR analyses of IL-18/Akt inhibitor-stimulated CMT93 cells for the binding of Tcf4 to the mouse Lgr5 promoter regions (P1 and P5). Fold induction is calculated based on PCR signal strength of Tcf4-bound Lgr5 promoter P1 region before and after stimulation. Each symbol in bar graphs represents a well of CMT93 cell culture (g), an ileum crypt sample (a-c) or organoid culture (e, f) derived from one mouse. Data shown are representative (d, f-h) or combined (a-c, e) results from two independent reproducible experiments. Statistical significance is indicated using unpaired two-tailed t test (a-c), One-way ANOVA with Sidak's multiple comparisons test (f, g), or Two-way ANOVA with Tukey's multiple comparisons test (e). Data are presented as mean ± SD. Source data are provided as a Source Data file.
of stimulated organoids, some wells were treated with 50 ng/ml recombinant mouse IL-22 protein (eBioscience #14-8221-63), 100 ng/ml recombinant mouse IL-18 protein (R&D #9139-IL), or both, for 24 hr on day-5. For organoids morphology observation or protein analysis with FACS or immunofluorescence, the treatment wells were given recombinant IL-22 or IL-18 as above at the initial stage of organoid culture, and the medium was replaced every 2 days with fresh ENR growth medium contained fresh IL-22 or IL-18.
Organoid observation and isolation from the Matrigel. To observe the morphology of organoids, the 24-well plates containing organoids were observed by a phase-contrast microscope. Images were obtained with a Leica model DMi8 imaging system under 10x, 20x objectives with Leica Application Suite X software. The number of organoid buds was determined by counting new buds except for the original one. The size of the organoid was calculated from the longest length within the original bud. To remove the organoid from Matrigel, the growth medium was  removed and Matrigel containing 6 days old organoids were washed with cold PBS twice. About 500 μl Cell Recovery Solution (CRS) (Corning #354253) was added to each well and the Matrigel was scrapped into an ice-cold 50 ml falcon tube. The well was additionally rinsed with CRS once and cell suspension was transferred to the same tube and shaken at 100 rpm on ice for 1 hr until the Matrigel was completely dissolved. After the organoids were released from the Matrigel, the tube was centrifuged at 300 g for 5 min at 4°C and the supernatant was discarded. The organoid pellet was resuspended in 10 ml wash buffer and transferred to a 15 ml falcon tube and centrifuged at 20 g for 5 min at room temperature to remove any dead cells.
Crypt killing assay. Intestinal crypts were isolated as previously described 30 . Briefly, mice were euthanized, ileum or colon fragments were isolated, opened longitudinally, and immediately washed with ice-cold PBS. The fragments were immersed in PBS with 30 mM EDTA for 20 min on ice. After removal of EDTA buffer, the fragments were vigorously suspended by using a 3 ml dropper with cold PBS and briefly centrifuged at 200 g for 3 min. After discarding the supernatant, the resulting sediment, mostly containing the villi, was resuspended with PBS. After further vigorous resuspension, centrifugation, and filter through a 70 μm cell strainer (BD Bioscience), the supernatant was enriched for crypts. Isolated crypts were manually counted under a microscope. About 2000 crypts from each sample were in vitro stimulated with 100 ng/ml IL-22, 100 ng/ml IL-18, or both, to release anti-microbial peptides (AMP) at 37°C for 30 min. AMP-containing supernatants were then incubated with 1000 CFU of live adherent-invasive E. coli (AIEC) for 90 min, and the AIEC titer after AMP-mediated killing were determined by CFU assay. The percentage of AIEC killing by crypts was normalized to those without cytokine stimulation.
Chromatin immunoprecipitation (ChIP). For ChIP assay in human HT-29 cells, 1.5×10 7 cells were stimulated with or without 50 ng/ml IL-22 for 15 min at 80% cell confluency. For ChIP assay in mouse CMT93 cells, 1.5×10 7 cells were stimulated with 100 ng/ml IL-18 and/or 5 μM MK-2206 (Akt inhibitor) for 24 h at 80% cell confluency. Cross-linking were performed with 1% formadehyde for 10 min at room temperature, followed by a quenching of formadehyde using 0.125 M glycine. Cells were washed with PBS 3 times. Nuclei were isolated in 400 μl of cell lysis buffer (10 mM Tris, pH 8.0, 10 mM NaCl, 0.2% NP-40) containg 1X phosphatase and protease inhibitor cocktail (MedChemExpress, HY-K0022 and HY-K0010) in the cold room (4°C) with 30 min rotation. Nuclei were then centrifuged at 13,200 rpm in a refrigerated microcentrifuge and were lysed in 100 μl of nuclear lysis buffer (50 mM Tris, pH 8.1, 10 mM EDTA, 1% SDS, 1X phosphatase and protease inhibitor cocktail) in the cold room with 2 hr rotation. 1 ml of immunoprecipitation (IP) dilution buffer (20 mM Tris, pH 8.1, 2 mM EDTA, 150 mM NaCl, 1% Triton X-100, 0.01% SDS) was added and the chromatin was sheared with 14-16 pulse at 40% amplitude using an Ultrasonic Processor (110 V) with a 3 mm microtip probe (Misonix, XL-2020), with tubes immersed in cold water. Each pulse cycle is comprised of 5 sec sonication followed by 15 sec cool down. The chromatin was sheared to an average size of 200-1000 base pairs. Pellets were centrifuged at 13,200 rpm at 4°C and the supernatant was mixed with 40 μl of packed protein A/G agarose beads (Santa Cruz Biotechology #sc-2003) for preclean process before immunoprecipitation. Supernatant was then split into three aliquots after centrifuged at 3,000 rpm and 700 μl of IP dilution buffer was added to each qliquot. 2.5 μg of antibody were added and the tubes were rotated overnight at 4°C. 40 μl of packed protein A/G agarose beads were added and the samples were rotated for 2 hr at 4°C. The beads were washed with IP wash I buffer (20 mM Tris, pH 8.1, 2 mM EDTA, 50 mM NaCl, 1% Triton X-100, 0.1% SDS) three times for 10 min each at 4°C, followed by one wash with IP wash II buffer (10 mM Tris, pH 8.1, 1 mM EDTA, 0.25 M LiCl, 1% NP-40, 1% deoxycholic acid) for 10 min at 4°C. Three washes with cold TE buffer (50 mM Tris, pH 8.0, 10 mM EDTA) were performed. After eluting twice using 200 μl of fresh elution buffer (100 mM sodium bicarbonate, 1% SDS) for 10 min at room temperature, 16 μl of 5 M NaCl were added and the cross-links were reversed at 65°C overnight. The DNA was purified using PCR purification kit (Qiagen, #28106). 50 μl of each ChIP product was recovered and 1 μl was used for PCR. PCR was performed using following conditions: 95°C for 30 sec, 55°C for 30 sec, and 72°C for 30 sec, for 35 cycles. Primer sequences are listed in the section of Supplementary Methods. After electrophoresis analysis, the intensity of each PCR product band was quantified by ImageJ (1.53k) software. For ChIP assay in mouse ileum organoids, one 24-well plate of organoid culture was used. Organoids were isolated from the ileum and cultured (600 crypts for a well) for 5 days before stimulation with 50 ng/ml IL-22 or 100 ng/ml IL-18 for 24 hr. Stimulated organoids were isolated from the the Matrigel and the ChIP assay was performed as described above. Transcription factor binding sites on the Il-18 or Lgr5 promoter (human and mouse) were predicted by database JASPAR. The PCR primers for ChIP assays were designed by Primer3web.
Statistics. All statistical analyses were performed using GraphPad Prism, v8.02 software. The results are expressed as the mean ± S.D. Statistical significance (p value) is indicated. Unpaired two-tailed t test, One-way ANOVA with Sidak's multiple comparisons test, or Two-way ANOVA with Tukey's multiple comparisons test is used for the analysis and indicated in each figure legend. The definition of symbols or sample numbers (n) is provided in each figure for statistics analysis.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability
The Source Data, containing all raw data presented in each figure of the main manuscript and supplementary information, as well as all uncropped Western blots and DNA gel blots, are provided with this paper. Transcription factor binding sites, illustrated in Supplementary Fig. 3a, 3c, and 8c, are predicted by the open-access database JASPAR (https://jaspar.genereg.net). The PCR primers used for all ChIP assays are designed by Primer3web (https://primer3.ut.ee/). All relevant primer sequences are listed in Supplementary information. Source data are provided with this paper.