key: cord-0295765-y6w2hhzd
authors: Sposito, Benedetta; Mambu, Julien; Gwilt, Katlynn Bugda; Spinelli, Lionel; Andreeva, Natalia; Galland, Franck; Naquet, Philippe; Mitsialis, Vanessa; Thiagarajah, Jay R; Snapper, Scott B; Broggi, Achille; Zanoni, Ivan
title: Type III interferons induce pyroptosis in gut epithelial cells and delay tissue restitution upon acute intestinal injury
date: 2022-03-04
journal: bioRxiv
DOI: 10.1101/2022.03.04.482997
sha: afa919bf962352b73affff7e97ff061ccdb71a50
doc_id: 295765
cord_uid: y6w2hhzd

Tissue damage and repair are hallmarks of the inflammatory process. Despite a wealth of information focused on the mechanisms that govern tissue damage, mechanistic insight on how inflammatory immune mediators affect the restitution phase is lacking. Here, we investigated how interferons influence tissue restitution after damage of the intestinal mucosa driven by inflammatory or physical injury. We found that type III, but not type I, interferons serve a central role in the restitution process. Type III interferons induce the upregulation of ZBP1, caspase activation, and cleavage of gasdermin C, and drive epithelial cell death by pyroptosis, thus delaying tissue restitution. We also found that this pathway is transcriptionally regulated in IBD patients. Our findings highlight a new molecular signaling cascade initiated by the immune system that affects the outcome of the immune response by delaying tissue repair and that may have important implications for human inflammatory disorders.

Introduction.

The immune system evolved to protect the host from external or internal threats, as well as to maintain homeostasis of the organs and tissues. The strong interrelationship between these two functions of the immune system is best exemplified during the restitution phase that follows mucosal damage, occurring as a consequence of an immune response. The skin, the lungs, the gut, and other mucosae are constantly exposed to microbial and/or physical perturbations and harbor multiple immune and non-immune cells that sense the presence of hostile environmental or endogenous factors and mount a defensive response. The causative agent of this response, the response itself, or both, may lead to tissue damage. Tissue damage sensing by tissueresident as well as newly recruited cells initiates a complex cascade of cellular and molecular processes to restore tissue functionality and homeostasis, or to adapt to persistent perturbations (Meizlish, Franklin et al. 2021 ).

The gastrointestinal tract represents an ideal tissue to explore the mechanisms underlying the exquisite balance between tissue damage and repair orchestrated by the immune system. In the intestine, immune cells, epithelial cells, and commensal microbes are in a dynamic equilibrium. A monolayer of highly specialized epithelial cells separates the gut lumen from the underlying lamina propria. The interplay between microbiota-derived inflammatory cues and the host cells in the intestine profoundly impacts the biology of the gut, both during homeostasis, inflammation, and damage responses. The lamina propria hosts a large variety of immune and non-immune cells that detect alterations in the functioning as well as in the integrity of the epithelial barrier and mount an immune response. The fine equilibrium between the microbiota, the epithelial barrier, and the immune system is lost during inflammatory bowel diseases (IBDs).

IBDs are a group of heterogeneous diseases, whose pathogenesis is associated with genetic and environmental factors, that are characterized by a dysregulated immune response Fiocchi 2011, Roda, Chien Ng et al. 2020) . Along with a heightened inflammatory response, IBDs are characterized by the breach of the intestinal barrier and a defective repair response that compromises mucosal homeostasis. Therefore, the ability of immune mediators to influence epithelial repair has an important impact on the pathogenesis of IBDs. Indeed, the promotion of mucosal healing has been recognized as a major therapeutic challenge for the management of IBDs (Pineton de Chambrun, Peyrin-Biroulet et al. 2010) .

We previously showed that a group of interferons (IFNs), known as type III IFNs or IFN-λ (Kotenko, Gallagher et al. 2003 , Sheppard, Kindsvogel et al. 2003 , Prokunina-Olsson, Muchmore et al. 2013 , limits inflammation in a mouse model of colitis by dampening the tissue-damaging functions of neutrophils (Broggi, Tan et al. 2017) . IFN-λ, as type I IFNs, plays potent anti-microbial roles, but, in contrast to type I IFNs, also preserves gut functionality by limiting excessive damage . The limited damage is largely explained by the fact that the expression of the IFN-λ receptor (IFNLR) is mainly restricted to epithelial cells and neutrophils. In contrast, type I IFNs act systemically and play potent inflammatory activities on immune and nonimmune cells thanks to the broad expression of the type I IFN receptor (IFNAR). The local activity of IFN-λ at mucosal tissues, thus, limits the extent of activation of immune cells, preventing excessive tissue damage, while preserving the anti-microbial functions of IFN-λ ).

Although we and others have shown that IFN-λ limits intestinal tissue damage, the involvement of this group of IFNs during tissue restitution of the gut is more controversial. Indeed, IFN-λ and type I IFNs have been shown to function in a balanced and compartmentalized way to favor re-epithelization by acting, respectively, on epithelial cells or immune cells resident in the lamina propria (McElrath, Espinosa et al. 2021) . IFN-λ has been also proposed to facilitate the proliferation of intestinal epithelial cells via STAT1 signaling (Chiriac, Buchen et al. 2017) and to partially enhance gut mucosal integrity during graft versus host disease (Henden, Koyama et al. 2021 ). On the other hand, though, IFN-λ and/or the IFNLR were found to be upregulated in IBD patients (Chiriac, Buchen et al. 2017 , Gunther, Ruder et al. 2019 . Also, systemic and prolonged overexpression of IFN-λ in mice favored the death of Paneth cells, a group of cells that can facilitate epithelial cell regeneration by acquiring stem-like features (Schmitt, Schewe et al. 2018) , and by regulating the balance of epithelial growth factors in the stem cell niche (Sato, van Es et al. 2011 ).

In keeping with a possible detrimental role of IFN-λ during an inflammatory response at mucosal surfaces, we and others have recently demonstrated that IFN-λ delays the proliferation of lung epithelial cells in murine models of persistent viral infections (Broggi, Ghosh et al. 2020 , Major, Crotta et al. 2020 . Also, that IFN-λ production in the lower respiratory tract of COVID-19 patients is associated with increased apoptotic and decreased proliferative transcriptional programs, and characterizes SARS-CoV-2-infected individuals with severe-to-critical outcomes (Sposito, Broggi et al. 2021) . Whether IFN-λ plays similar roles in the intestine, and the molecular mechanisms initiated by this group of IFNs to exert their functions during gut restitution, remain unknown.

Here, by exploiting conditional knock-out mice that do not respond to IFN-λ only in intestinal epithelial cells or in neutrophils, ex vivo transcriptomics, and biochemical assays, as well as intestinal organoids in vitro, we dissected the role of IFN-λ during tissue repair secondary to either an inflammatory insult or to radiation damage. Our data reveal a new molecular cascade initiated by IFN-λ that culminates in the activation of ZBP1 and of gasdermin C (GSDMC), in the induction of pyroptosis and results in delayed gut restitution.

We previously showed that, in the acute inflammatory phase of the dextran sulfate sodium (DSS) model of colitis, IFN-λ signaling in neutrophils dampens reactive oxygen species production and neutrophil degranulation, and thus restrains intestinal damage (Broggi, Tan et al. 2017) . To assess the involvement of IFN-λ during the restitution phase of the DSS colitis model, we injected, or not, recombinant (r)IFN-λ in mice after DSS-induced inflammation has peaked. We confirmed that rIFN-λ administration upregulated interferon-stimulated genes (ISGs) in the colon of DSStreated mice ( Figure S1A ). Mice administered rIFN-λ, but not vehicle controls, showed persistent weight loss, reduced colon length, and prolonged tissue damage as measured by histology ( Figure 1A-C) . These data suggest that IFN-λ delays tissue restitution in mice encountering colitis.

Next, we tested whether the endogenous IFN-λ, which is produced during colitis development (Broggi, Tan et al. 2017) , also affects the restitution phase. After the peak of the inflammatory process induced by DSS administration, mice were treated with a blocking antibody directed against IFN-λ and compared to mice treated with DSS, in the presence or absence of rIFN-λ. Our data demonstrated that inhibition of endogenous IFN-λ facilitates tissue restitution as measured by increased weight gain and colon lengthening (Figure 1D, E) . Notably, we found that ISG levels in epithelial cells were significantly decreased in mice treated with the anti-IFN-λ antibody ( Figure S1A ), suggesting that IFN-λ, rather than type I IFNs, plays a major role in driving gene transcription during the repair phase of colitis. To directly test the involvement of type I IFNs in the restitution phase of DSS-induced colitis, we either blocked type I IFN signaling using an anti-IFNAR antibody, or added rIFNβ, 7 days after DSS administration. In keeping with a key role of IFN-λ in regulating mucosal epithelial responses, none of the treatments aimed at targeting type I IFNs affected tissue repair (Figure 1F, G) . Accordingly, ISG levels in colonocytes were not altered under these experimental conditions compared to control mice ( Figure S1B ).

While intestinal epithelial cells are the major effector cell type during mucosal restitution, other cells, including immune cells, can participate in modulating tissue repair. Since intestinal epithelial cells and neutrophils are the two cell types that respond to IFN-λ in the gut of mice (Broggi, Tan et al. 2017) , we used conditional knock out mice that do not express the IFNLR either in intestinal epithelial cells (Vil CRE Ifnlr1 fl/fl mice) or neutrophils (Mrp8 CRE Ifnlr1 fl/fl mice). Ifnlr1 fl/fl (WT) littermates were used as controls. In contrast to WT littermates, administration of rIFN-λ to Vil CRE Ifnlr1 fl/fl mice did not delay tissue restitution as measured by weight change (Figure 1H ).

Vil CRE Ifnlr1 fl/fl mice in which intestinal epithelial cells do not respond to IFN-λ showed a faster recovery as measured by a significant increase in colon length, regardless of the presence or absence of rIFN-λ ( Figure 1I) . In contrast to Vil CRE Ifnlr1 fl/fl mice, Mrp8 CRE Ifnlr1 fl/fl behaved similarly to their WT counterpart, in the presence or absence of rIFN-λ ( Figure S1C ). These data demonstrate that, in contrast to the acute inflammatory phase of colitis, epithelial cells, not neutrophils, are the major responders to endogenous, as well as exogenous, IFN-λ and that IFNλ signaling in epithelial cells delays tissue restitution.

Repair of the gut epithelial monolayer is a complex process, and the regenerative capacity of intestinal stem cells (ISCs) plays a critical role (Blanpain and Fuchs 2014) . To target ISCs and assess the direct involvement of IFN-λ during gut restitution, we employed a well-characterized model of epithelial damage resulting from exposure to ionizing radiations (Kim, Yang et al. 2017 ).

In this model, radiation induces widespread epithelial cell death in the small intestine, with a particularly dramatic effect on cycling ISCs that reside at the bottom of the small intestinal crypt.

Cell death is followed by repair of the damaged epithelial crypts and return to homeostasis. Three to four days after radiation injury, during the peak of the repair response, crypt regeneration was assessed in WT mice, WT mice administered exogenous rIFN-λ, or Ifnlr1 -/mice. We found that mice that received rIFN-λ showed reduced regeneration of the crypts, while Ifnlr1 -/mice had an increased number of crypts (Figure 2A) . Notably, in the small intestine of irradiated mice, similarly to what we observed in the colon of mice exposed to DSS, endogenous IFN-λ, but not type I IFN, signaling caused the delay in tissue restitution ( Figure 2B) . Similarly, ISG induction in epithelial cells was dependent on IFN-λ, rather than type I IFNs ( Figure S2A ).

Next, we assessed the nature of the cell types that respond to IFN-λ in the irradiated small intestine. When Vil CRE Ifnlr1 fl/fl mice and WT littermates were used, we found that the number of crypts three days post-radiation was significantly increased in Vil CRE Ifnlr1 fl/fl mice compared to WT mice ( Figure 2C, S2B) . We also demonstrated that exogenous rIFN-λ does not affect the number of crypts in Vil CRE Ifnlr1 fl/fl mice, while delaying tissue restitution in WT littermates (Figure 2C, S2B) .

In keeping with a key role for epithelial cells, but not neutrophils, in responding to IFN-λ during tissue restitution, we found that Mrp8 CRE Ifnlr1 fl/fl mice didn't show significant differences compared to their WT littermates ( Figure S2C ). No differences were measured in the number of crypts of non-irradiated mice regardless of their capacity to respond, or not, to IFN-λ ( Figure S2D ). Finally, we followed over time irradiated Vil CRE Ifnlr1 fl/fl mice or WT littermates, treated or not with rIFN-λ.

We found that WT mice irradiated and treated with rIFN-λ lost significantly more weight than irradiated WT mice, and all died ( Figure 2D, E) . Notably, WT littermates lost significantly more weight compared to Vil CRE Ifnlr1 fl/fl mice, treated or not with rIFN-λ ( Figure 2D ). In contrast, Vil CRE Ifnlr1 fl/fl mice treated or not with rIFN-λ showed a very similar behavior (Figure 2D, E) .

Overall, these data demonstrate that epithelial cell regeneration and tissue restitution in the small intestine of irradiated mice is inhibited in the presence of IFN-λ. Also, that IFN-λ delays repair by acting on intestinal epithelial cells.

To determine the transcriptional programs initiated by IFN-λ to delay tissue restitution, we isolated intestinal crypts from the small intestine of Vil CRE Ifnlr1 fl/fl mice or WT littermates that have been irradiated and performed targeted transcriptomics analysis (RNAseq). In keeping with a major role of IFN-λ-dependent responses in the intestine, when we performed gene ontology (GO) enrichment analyses, IFN-signaling related pathways, as well as anti-viral or anti-bacterial pathways, were highly enriched in WT epithelial cells, compared to Vil CRE Ifnlr1 fl/fl ( Figure 3A) . In contrast, GO terms associated with cell migration and extracellular remodeling, which are linked to higher efficiency in the closure of mucosal wounds (Quirós and Nusrat 2018) , were mostly represented in epithelial cells that do not respond to IFN-λ ( Figure 3A) .

Gene set enrichment analysis (GSEA) confirmed that genes associated with IFN responses were significantly enriched in WT, compared to knock-out, epithelial cells (Figure 3B , S3A). We next assessed the relative enrichment of a previously identified colitis-associated regenerative epithelial gene-set (Yui, Azzolin et al. 2018) , as well as gene-sets associated with epithelial cell proliferation. Both gene-sets were significantly enriched when epithelial cells did not respond to IFN-λ ( Figure 3C , D, S3B, C).

To assess whether IFN-λ-dependent delayed tissue restitution is characterized by reduced cell proliferation in vivo, we administered the thymidine analog 2'-deoxy-5-ethynyluridine (EdU) two hours before mice were euthanized and measured cell proliferation in either WT littermates or Vil CRE Ifnlr1 fl/fl mice, administered or not rIFN-λ. We found that exogenous rIFN-λ reduced the number of EdU-positive cells per crypt in WT but not Vil CRE Ifnlr1 fl/fl mice ( Figure 3E) . Also, that Vil CRE Ifnlr1 fl/fl mice had a significant increased number of proliferating cells per crypts, compared to WT littermates, irrespectively of the administration of rIFN-λ ( Figure 3E ).

During tissue repair that follows radiation damage (Metcalfe, Kljavin et al. 2014) or colitis (VanDussen, Sonnek et al. 2019) , specialized ISCs drive re-epithelialization by massively proliferating. Therefore, the decreased number of proliferating epithelial cells we observed in WT mice may reflect the lack of reparatory ISCs that proliferate. To assess whether endogenous IFN-λ affected the cellular composition of the small intestine in WT or Vil CRE Ifnlr1 fl/fl mice that were irradiated, we used CIBERSORTx (Newman, Steen et al. 2019 ) and deconvoluted our bulk RNAseq data based on single-cell RNAseq data previously published (Haber, Biton et al. 2017 ).

Our deconvolution analysis revealed that, while most epithelial cell types did not present major significant differences, the ISC compartment was significantly expanded in mice that were irradiated and whose epithelial cells do not respond to IFN-λ ( Figure 3F ). In keeping with retrodifferentiation of transit-amplifying (TA) cells to replenish the ISC compartment upon ISC depletion (Wang, Chiang et al. 2019 , Ohara, Colonna et al. 2022 , TA cells were significantly decreased in the small intestine of Vil CRE Ifnlr1 fl/fl mice, compared to WT controls ( Figure 3F ). The expansion of the Lrg5 + compartment in mice that do not respond to IFN-λ was confirmed by qPCR ( Figure S3D) . We also confirmed that the major epithelial cell populations analyzed were not different under homeostatic conditions in Vil CRE Ifnlr1 fl/fl mice or WT littermates ( Figure S3E ).

Overall, these data demonstrate that IFN-λ initiates a transcriptional program that reduces tissues restitution, limits ISC cell expansion, and, thus, dampens the overall capacity of epithelial cells to proliferate.

The reduced expansion of ISC can be driven either by increased cell death of ISCs and/or TA cells, reduced proliferative programs, or both. To determine the molecular mechanisms regulated by IFN-λ to dampen tissue restitution, we identified the genes that were significantly differentially regulated in epithelial cells derived from irradiated Vil CRE Ifnlr1 fl/fl , compared to WT, mice ( Figure 4A ). As expected, multiple ISGs were among the genes significantly downregulated in cells that cannot respond to IFN-λ ( Figure 4A ). Intriguingly, Zbp1 was among these genes.

ZBP1 is a key component in the multiprotein complex PANoptosome, which encompasses effectors of several forms of cell death, and is an important regulator of cell fate . We also found that protein levels of ZBP1, as well as another ISG such as RSAD2, were upregulated in epithelial cells of the small intestine upon in vivo administration of rIFN-λ in non-irradiated WT mice ( Figure S4A ). Upregulation of these proteins was prevented in epithelial cells derived from Vil CRE Ifnlr1 fl/fl mice and was not different in the absence of rIFN-λ in the two backgrounds ( Figure S4A ). Among other genes significantly downregulated in cells that do not respond to IFN-λ there were two members of the gasdermin C (GSDMC) family. GSDMs are critical effectors of pyroptosis, a form of inflammatory cell death (Kovacs and Miao 2017) .

Compared to other GSDMs, very little is known about the functions of GSDMC, and scattered reports have involved GSDMC in the lytic death of tumor cells (Hou, Zhao et al. 2020 , Zhang, Zhou et al. 2021 , or of enterocytes during helminth infections (Xi, Montague et al. 2021) . Of note, non-irradiated mice administered with rIFN-λ do not show upregulation of the GSDMC protein ( Figure S4A) , demonstrating that additional pathways associated with irradiation and/or tissue damage and repair regulate Gsdmc gene expression and/or protein synthesis.

IBD patients present increased levels of IFN-λ and/or Ifnlr1 (Chiriac, Buchen et al. 2017 , Gunther, Ruder et al. 2019 . Prompted by our findings in irradiated mice, we assessed the expression levels of ZBP1 and GSDMC (the only GSDMC present in humans) in the biopsy derived from a cohort of IBD patients with active or inactive disease, or non-IBD controls (see Material and Methods for details). We found a significant increase in the expression of ZBP1 as well as GSDMC in patients with active IBD, compared to non-IBD controls ( Figure 4B ). Similar expression trends were observed in RNAseq datasets derived from rectal mucosal biopsies from ulcerative colitis (UC) pediatric patients (PROTECT cohort) and from ileal biopsies from Crohn's disease (CD) pediatric patients (RISK cohort) in two independent cohorts previously published (Haberman, Tickle et al. 2014 , Haberman, Karns et al. 2019 (Figure S4B ). In keeping with a key role of IFNs also in IBD patients, we found that genes regulated in response of IFNs (evaluated as mean expression of genes that belong to the "HALLMARK _IFN_ALPHA_RESPONSE" geneset (Liberzon, Birger et al. 2015) and indicated as "IFN response score") were significantly enriched in IBD patients with active disease, compared to controls or patients with inactive disease ( Figure 4B) . In support of a role of IFN-λ in modulating GSDMC expression, GSDMC levels positively correlated with the IFN response score in patients with active disease, but not in the other subjects analyzed ( Figure 4C) . These data indicate that IFN induction and upregulation of the genes that encode for ZBP1 and GSDMC are hallmarks of intestinal damage both in mice and humans.

GSDMs exert their pyroptotic function upon cleavage by caspases (Kovacs and Miao 2017) , when the N-terminal cleavage product oligomerizes to form lytic pores in the cell membrane, leading to the loss of ionic homeostasis and cell death (Broz, Pelegrín et al. 2020) .

We, thus, tested whether GSDMC-2/-3 were cleaved in epithelial cells of the small intestine upon irradiation and confirmed that irradiated, but not non-irradiated, mice not only showed increased levels of GSDMC-2/-3, but also that GSDMC-2/-3 were efficiently cleaved in WT, but not Vil CRE Ifnlr1 fl/fl , mice ( Figure 4D ). In contrast, another key effector of pyroptosis, GSDMD, was not activated.

GSDMC is primarily cleaved by Caspase-8 (Hou, Zhao et al. 2020 , Zhang, Zhou et al. 2021 ). Indeed, the pattern of bands of cleaved GSDMC-2/-3 is compatible with the activity of Caspase-8 (Julien and Wells 2017). We thus investigated the activation of Caspase-8. Caspase-8 was not activated in WT or Vil CRE Ifnlr1 fl/fl mice in the absence of irradiation ( Figure S4A ). In contrast, epithelial cells derived from irradiated WT littermates, but not Vil CRE Ifnlr1 fl/fl mice, efficiently activated Caspase-8 ( Figure 4D ). Of note, we found that expression of CASP8 was increased in patients with active IBD compared to non-IBD controls ( Figure 4B, S4B ). In keeping with the capacity of ZBP1 to control the activation of multiple caspases (Kuriakose and Kanneganti 2018), we found that a similar pattern of activation was also true for Caspase-3 ( Figure 4D ).

Overall, these data demonstrate that IFN-λ initiates in the small intestine of irradiated mice a signaling cascade that allows the upregulation of ZBP1, the activation of Caspase-8/-3, and the induction and cleavage of GSDMC, an executor of pyroptosis. Also, that similar programs are transcriptionally upregulated in IBD patients with active disease.

To assess directly the role of the signaling cascade initiated by IFN-λ in driving cell death, we used intestinal organoids. Mouse and human small intestinal organoids seeded in the presence of rIFN-λ died between 48 and 72h from treatment ( Figure 5A , S5A). Dying cells assumed typical changes associated with pyroptosis including swelling and sudden disruption of the plasma membrane and liberation of nuclear DNA ( Figure 5A ). By using organoids derived from WT or Stat1 -/mice, we also confirmed that gene transcription induced by IFN-λ was necessary to induce cell death ( Figure 5B ). No differences were observed between the two genotypes in the absence of rIFN-λ ( Figure 5B) . Similar to what we observed in irradiated epithelial cells in vivo, we found that rIFN-λ administration to small intestine organoids profoundly diminished the level of Lgr5 expression, suggesting a defect in the maintenance and/or proliferation of ISCs ( Figure S5B ). In agreement with a reduced number of ISCs that proliferate, we also found that cell proliferation (as measured by EdU incorporation) was significantly decreased in IFN-λ-treated mouse, as well as human organoids ( Figure 5C , S5C).

To assess the involvement of ZBP1 in these processes, we derived organoids from either WT or Zbp1 -/mice. Organoids were differentiated in the presence or absence of IFN-λ and their survival and/or growth was followed for 72h. Survival and growth of organoids, derived either from the small or large intestine, differentiated from WT, but not Zbp1 -/-, mice were significantly reduced upon the administration of rIFN-λ ( Figure 5D , S5D). In agreement with the capacity of IFN-λ to activate a ZBP1/Caspase-8/GSDMC axis, organoids grown for 6 days and then administered with rIFN-λ showed ZBP1 upregulation and cleavage of GSDMC-2/-3 and Caspase-8 and Caspase-3 ( Figure 5E ). Furthermore, inhibition of caspase activity with the pan-caspase inhibitor Z-VAD-FMK protected the organoids from cell death induced by IFN-λ ( Figure 5F ). To directly assess the involvement of GSDMC in this process, we knocked down Gsdmc-2 and Gsdmc-3 in small intestine organoids and found that, similar to Zbp1 -/organoids, upon exposure to rIFN-λ, survival of organoids that do not express Gsdmc2, 3 was significantly increased compared to controls ( Figure 5G ).

To better reflect the cycles of injury and repair characteristic of IBD and mouse models of colitis, we implemented a previously described model of long-term organoid culture (Wang, Chiang et al. 2019 ). We grew organoids in a two-dimensional (2D) epithelial monolayer system and exposed their apical side to air, to obtain a self-organizing monolayer that mimics cells in homeostasis. This monolayer can then be re-submerged in medium (to elicit damage response mimicking in vivo epithelial injuries) and re-exposed to air (which induces epithelial regeneration responses) ( Figure 5H ). When we treated with IFN-λ the epithelial monolayer after re-exposure to air, the proliferative repair response was curbed, as demonstrated by the failure to incorporate EdU ( Figure 5I ).

Overall, our data demonstrate that IFN-λ signaling induces a ZBP-1-GSDMC axis that controls epithelial cell survival, and dampens the capacity of ISCs to proliferate and orchestrate tissue restitution.

In our work, we revealed that IFN-λ restrains the restitution of the intestinal mucosae secondary to either inflammatory damage or ionizing radiations toxicity. We reveal the capacity of IFN-λ to initiate a previously overlooked molecular cascade in intestinal epithelial cells that allows the induction of ZBP1, the activation of caspases, and the induction and cleavage of GSDMCs. We also found that similar pathways are transcriptionally upregulated in IBD patients with active disease. Induction of epithelial cell death via the ZBP1-GSDMC axis reduces the number of ISCs, and dampens the proliferation and restitution of epithelial cells, thus affecting the re-epithelization of the injured intestine. Finally, we revealed that IFN-λ, but not type I IFNs, are the major drivers of the delayed restitution in vivo in mouse models of gut damage.

The immune system is endowed with the capacity not only to protect against pathogen invasion but also to maintain tissue homeostasis. Fundamental to exert these activities, is the fine balance between anti-microbial functions of the immune system that can drive tissue damage, and the regenerative capacity of organs and tissues. Many cellular and molecular mediators of the immune system are involved in exerting anti-microbial and potentially damaging functions, but several can also modulate mucosal repair. Here, we focused our attention on a group of IFNs, known as type III IFNs or IFN-λ. IFN-λ activities at mucosal surfaces are essential to limit pathogen spread while reducing inflammation and immune cell infiltration ).

IFN-λ and type I IFNs regulate very similar transcriptional programs, but the limited expression of the IFNLR restricts the activity of IFN-λ to epithelial cells, hepatocytes, neutrophils and few other cell types and, thus, reduces the extent of the inflammation . The limited number of cells that respond to IFN-λ signaling, and the reduced capacity of IFN-λ, compared to type I IFNs, to activate IRF1 (Forero, Ozarkar et al. 2019) allow to preserve the functionality of mucosal tissues during an immune response. Although the protective functions of IFN-λ in the gut, and in general at mucosal surfaces, are well known ), much less is known about the functions of this group of IFNs during the healing phase that follows intestinal tissue damage. Our data reveal the unique capacity of IFN-λ, compared to type I IFNs, to negatively affect tissue restitution in the intestine, possibly opening new ways of therapeutic intervention for individuals that encounter tissue damage such as IBD patients or subjects exposed to radiation therapies. healing. Lrg5 + ISCs support normal cell turnover as well as injury-induced restitution (Metcalfe, Kljavin et al. 2014) . When ISC are depleted by radiation, or by immune-mediated tissue-damaging events, TA cells retro-differentiate and acquire new stem-like properties in the small, as well as in the large, intestine (Wang, Chiang et al. 2019 , Ohara, Colonna et al. 2022 . These cells then proliferate to allow the re-epithelization of the damaged tissue. We and others previously described the capacity of IFN-λ to dampen lung epithelial cell proliferation (Broggi, Ghosh et al. 2020 , Major, Crotta et al. 2020 and to instruct anti-proliferative transcriptional programs in the lung of patients infected with SARS-CoV-2 (Sposito, Broggi et al. 2021) . Nevertheless, our new findings in the gut suggest that decreased proliferation assessed at transcriptional and cellular levels is due to augmented cell death, possibly occurring in newly generated ISCs or in TA cells.

If similar processes also take place in the lung, it remains an open question that will require further investigation.

Another interesting observation we made is that exogenous administration of IFN-λ does not induce caspase or GSDMC activation in vivo in the absence of tissue damage, although it induces ZBP1 upregulation at the transcriptional as well as protein level. ZBP1 is a Z-DNA binding protein, and is part of the PANoptosome, a multiprotein complex that governs the cell fate (Place, Lee et al. 2021) . PANoptosis is a form of cell death that encompasses pyroptosis, apoptosis, and necroptosis. ZBP1 can interact directly or indirectly with proteins that regulate cell death and drive the activation of apoptotic caspases 8, 3 and 7, the necroptosis effector MLKL, or pyroptosis effectors Casp-1, 11 and GSDMD. So far, GSDMC was not associated with ZBP1 and/or PANoptosis. Our data highlight the existence of a ZBP1-GSDMC axis that appears to be the preferential pathway of cell death that is active during cycles of intestinal epithelial damage and restitution. Upregulation of ZBP1 alone is not sufficient to trigger the full activation of the PANoptosome, which is consistent with our inability to detect toxic effects of IFN-λ in the absence of inflammation or tissue damage. Conversely, ZBP1 can be activated both by binding microbialderived nucleic acids (Kuriakose, Zheng et al. 2018 , Muendlein, Connolly et al. 2021 or by binding host-derived Z-DNA following oxidative damage of the mitochondria (Szczesny, Marcatti et al. 2018) . It is, thus, possible that in vivo, under tissue-damaging conditions, either microbiota-or host-derived DNA becomes available to induce the assembly and activation of the PANoptosome downstream of ZBP1. In contrast to our in vivo data, administration of IFN-λ to murine or human intestinal organoids induces the ZBP1-GSDMC axis and drives cell death. We, thus, speculate that under our in vitro experimental conditions a "tissue damage" signal, e.g. Z-DNA from dying cells that are differentiating in vitro, is available, thus making additional signals unnecessary.

Linked to the above-mentioned observations, we also found that the ISC compartment is These data, together with our new findings, suggest that the compartmentalization of IFN-λ signaling at homeostasis preserves the functions of ISCs and the normal turnover of gut epithelial cells.

Our models of intestinal damage either of the colon, in the DSS-colitis model, or of the small intestine, in the radiation model, highlight the centrality of IFN-λ and its capacity to delay tissue restitution. Two previous studies suggested that IFN-λ may play an opposite role and favor tissue restitution during colitis (Chiriac, Buchen et al. 2017 , McElrath, Espinosa et al. 2021 .

Nevertheless, both studies were performed by inducing colitis in total Ifnlr1 -/mice and thus they suffer the confounding activity of IFN-λ on neutrophils. The absence of IFN-λ signaling in neutrophils potentiates tissue damage (Broggi, Tan et al. 2017) , making it hard to compare the tissue restitution phase with WT mice that start from a different level of damage. Indeed, we always administered or blocked IFNs in our colitis model after the peak of the inflammatory phase.

Alternatively, we used mice deficient for the IFNLR only in epithelial cells. Total knock-out mice were solely used in the radiation model in which damage and/or inflammation are not driven by neutrophils but by the ionizing radiations. The compartmentalized activity of IFN-λ in different cell types appears to be, thus, a key feature of this group of IFNs.

Overall, our data unveiled a new axis between IFN-λ, ZBP1 and GSDMC that governs tissue restitution in the gut and open new perspectives to future therapeutic interventions.

IZ is supported by NIH grants 1R01AI121066, 1R01DK115217, 1R01AI165505 and contract no. 

with Dunn correction for multiple comparisons was performed. ns= not significant (p > 0.05); *p < 0.05;**p < 0.01; ***p < 0.001; ****p < 0.0001.

(A) Mouse small intestinal organoids were seeded from freshly isolated crypts and allowed to grow for 48h. Organoids were then treated with 200ng/ml of rIFN-λ in the presence of 1ug/ml propidium iodide (PI) and imaged every 12h over 72h. Organoids were seeded in transwells and grown to confluence. The apical side was then exposed to air up to 14 days, which favored differentiation of a homeostatic monolayer. Organoids were then submerged for 7 days to induce damage responses. After 7 days they were re-exposed to air to stimulate repair responses. Concomitantly with re-exposure to air, organoids were treated with 200ng/ml of rIFN-λ for 3 days. Šidak correction for multiple comparison. (C) Unpaired t test. ns= not significant (p > 0.05); *or §p < 0.05;**or § §p < 0.01; ***or § § §p < 0.001; ****or § § § §p < 0.0001.

WT mice were treated with 2.5% DSS for 7 days. Upon DSS withdrawal mice were injected i.p. with either 50 µg kg-1day-1 of rIFN-λ or 12.5 mg kg -1 day -1 of anti-IFN-λ2,3 antibody for five days. 

To induce colitis, mice were given 2.5% (w/v) dextran sulfate sodium (DSS, Affymetrix) in drinking water for 7 days and were then administered water for 7 days. Where indicated in the figure legends, mice were received daily intraperitoneal injections of 50 mg kg -1 day -1 rIFN-λ or rIFN-β, and, to deplete endogenous IFN-λ or to block type I IFN signaling, mice received daily intraperitoneal injections of 12.5 mg kg -1 day -1 of anti-IFN-λ2-3 or anti-IFNAR1 antibody respectively. Body weight, stool consistency and the presence of blood in the stool were monitored daily. Weight change was calculated as percentage of initial weight. Crypt extraction. Small intestines were longitudinally cut and rinsed in PBS. Mucus was washed away by incubation with 1mM DTT at 4°C for 5 minutes. The tissue was moved to 10mM EDTA, 1% FBS, 1% sucrose at 37°C for 5 minutes. Samples were vortexed and small intestine fragments were moved to a new tube with 10mM EDTA, 1% FBS, 1% sucrose at 37°C for 10 minutes.

Samples were vortexed. The supernatant was filtered through a 70uM strainer and kept on ice.

Small intestine fragments were moved to a new tube with 10mM EDTA, 1% FBS, 1% sucrose at 37°C for 10 minutes. Samples were vortexed and the supernatant was combined with the previous fraction. The isolated crypts were resuspended in Trizol for RNA extraction and in RIPA Buffer with protease and phosphatase inhibitors for Western Blot analysis.

Samples were collected in Trizol (Thermo Scientific) and RNA was isolated using phenolchloroform extraction. Purified RNA was analyzed for gene expression by qPCR on a CFX384 Volcano plots were created using the R package EnhancedVolcano (v 1.12). The differentially expressed genes with an adjusted p-value lesser that 0.1 and a log2 fold change greater than 1.5

were selected for downstream analysis. Functional enrichment analysis in Gene Ontology was performed using the R package ClusterProfiler (v 4.2) with the Biological Process terms and

Benjamin-Hochberg multi-test correction with 5% of FDR threshold. Geneset Enrichment Analysis (GSEA) of hallmarks was performed using the R package fgsea (v 1.20) using the hallmark genesets (v 7.4) from the Broad Institute MSigDB or custom genesets. Leading edges of the different selected genesets were selected to build heatmaps of their expression in the different conditions and samples. The R package ComplexHeatmap (v 2.10) was used to plot the heatmaps. We used CIBERSORTx (Newman, Steen et al. 2019) to estimate the abundances of epithelial cell types using using bulk gene expression data as an input and scRNAseq signature matrices from single-cell RNA sequencing data to provide the reference gene expression profiles of pure cell populations. The scRNAseq signature matrix used to deconvolute RNAseq dataset from small intestine crypts was taken from (Haber, Biton et al. 2017 excluded from the analysis, and IBD samples labeled as macroscopically or microscopically inflammation were categorized as "Active" with the rest as "Inactive". For the PROTECT cohort, samples lacking histology scores were excluded from the analysis and all other IBD samples were categorized as "Active" if they had a Histology Severity Score (for chronic and active acute neutrophil inflammation) > 1 and "Inactive" if they had a Histology Severity Score of 0-1 (Boyle, Collins et al. 2017) . Group comparisons between healthy controls, inactive and active IBD were performed using non-parametric t-testing (Wilcoxon test) and p values reported.

Organoid culture and stimulation. Mouse intestinal spheroids were derived and maintained as previously described (Miyoshi and Stappenbeck 2013 For experiments with 2D organoids in Air-Liquid Interface (ALI), we followed a previously described protocol (Wang, Chiang et al. 2019) . Briefly, cultured mouse small intestinal organoids were dissociated in single cells and seeded on polycarbonate transwells, with 0.4μM pores (CORNING). Initially, cells were seeded in the presence of 50% L-WRN media with 10μM Rock inhibitor Y-27632 in both the lower and the upper chamber. After 7 days, the media was removed from the upper chamber to create an ALI. Cells were maintained in these conditions for 14 days to establish a homeostatic monolayer. The ALI culture was then resubmerged with 200μL 50% L-WRN medium, for 7 days and re-exposed to air for 3 days in the presence or absence of rIFN-λ, as indicated in the figure legends. After 3 days from re-exposure to air, cells were pulsed with 10 μM EdU for 2 hours, fixed in 10% formalin and stained for EdU incorporation. Samples were examined using a Zeiss LSM 880 confocal microscope (Carl Zeiss) and data were collected with fourfold averaging at a resolution of 2100 × 2100 pixels. The percentage of EdU-positive-cells was calculated as the ratio of the number EdU-positive foci and DAPI-positive foci.

cytofluorimetry. GSDMC knockdown (Gsdmc2, 3 KD ) stable cell lines were produced using commercially designed lentivirus particles targeting mouse Gsdmc2 (NM_001168274.1) and software or using custom scripts in R (v 4.1.1) and details are indicated in figure legends.

Throughout the paper statistical significancy is defined as follows: ns, not significant (p > 0.05); *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. 

Stem cell plasticity. Plasticity of epithelial stem cells in tissue regeneration

Histologic Correlates of Clinical and Endoscopic Severity in Children Newly Diagnosed With Ulcerative Colitis

Type III interferons disrupt the lung epithelial barrier upon viral recognition

Type III interferons: Balancing tissue tolerance and resistance to pathogen invasion

IFN-lambda suppresses intestinal inflammation by non-translational regulation of neutrophil function

The gasdermins, a protein family executing cell death and inflammation

Activation of Epithelial Signal Transducer and Activator of Transcription 1 by Interleukin 28 Controls Mucosal Healing in Mice With Colitis and Is Increased in Mucosa of Patients With Inflammatory Bowel Disease

Ulcerative colitis

Differential Activation of the Transcription Factor IRF1 Underlies the Distinct Immune Responses Elicited by Type I and Type III Interferons

Interferon Lambda Promotes Paneth Cell Death Via STAT1 Signaling in Mice and Is Increased in Inflamed Ileal Tissues of Patients With Crohn's Disease

A single-cell survey of the small intestinal epithelium

Ulcerative colitis mucosal transcriptomes reveal mitochondriopathy and personalized mechanisms underlying disease severity and treatment response

Pediatric Crohn disease patients exhibit specific ileal transcriptome and microbiome signature

IFN-lambda therapy prevents severe gastrointestinal graft-versus-host disease

PD-L1-mediated gasdermin C expression switches apoptosis to pyroptosis in cancer cells and facilitates tumour necrosis

Caspases and their substrates

Epithelial Regeneration Following Radiation-Induced Gut Injury

IFN-lambdas mediate antiviral protection through a distinct class II cytokine receptor complex

Gasdermins: Effectors of Pyroptosis

ZBP1: Innate Sensor Regulating Cell Death and Inflammation

IRF1 Is a Transcriptional Regulator of ZBP1 Promoting NLRP3 Inflammasome Activation and Cell Death during Influenza Virus Infection

The Molecular Signatures Database (MSigDB) hallmark gene set collection

Type I and III interferons disrupt lung epithelial repair during recovery from viral infection

Critical role of interferons in gastrointestinal injury repair

Tissue Homeostasis and Inflammation

Lgr5+ stem cells are indispensable for radiation-induced intestinal regeneration

In vitro expansion and genetic modification of gastrointestinal stem cells in spheroid culture

ZBP1 promotes LPS-induced cell death and IL-1beta release via RHIMmediated interactions with RIPK1

Determining cell type abundance and expression from bulk tissues with digital cytometry

Adaptive differentiation promotes intestinal villus recovery

Clinical implications of mucosal healing for the management of IBD

PANoptosis in microbial infection

A variant upstream of IFNL3 (IL28B) creating a new interferon gene IFNL4 is associated with impaired clearance of hepatitis C virus

Contribution of Wound-Associated Cells and Mediators in Orchestrating Gastrointestinal Mucosal Wound Repair

Crohn's disease

Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett's epithelium

Paneth cells constitute the niche for Lgr5 stem cells in intestinal crypts

Paneth Cells Respond to Inflammation and Contribute to Tissue Regeneration by Acquiring Stem-like Features through SCF/c-Kit Signaling

IL-28, IL-29 and their class II cytokine receptor IL-28R

The interferon landscape along the respiratory tract impacts the severity of COVID-19

Severity of SARS-CoV-2 infection as a function of the interferon landscape across the respiratory tract of COVID-19 patients

Mitochondrial DNA damage and subsequent activation of Z-DNA binding protein 1 links oxidative stress to inflammation in epithelial cells

A homeostatic interferon-lambda response to bacterial microbiota stimulates preemptive antiviral defense within discrete pockets of intestinal epithelium

L-WRN conditioned medium for gastrointestinal epithelial stem cell culture shows replicable batch-to-batch activity levels across multiple research teams

Differential roles of interferons in innate responses to mucosal viral infections

Long-Term Culture Captures Injury-Repair Cycles of Colonic Stem Cells

worm-induced type 2 immunity

YAP/TAZ-Dependent Reprogramming of Colonic Epithelium Links ECM Remodeling to Tissue Regeneration

The metabolite alpha-KG induces GSDMCdependent pyroptosis through death receptor 6-activated caspase-8