Promoting intestinal antimicrobial defense and microbiome symbiosis contributes to IL-22-mediated protection against alcoholic hepatitis in mice

Background The hepatoprotective effect of interleukin 22 (IL-22) has been reported in several models of liver injuries, including alcohol-associated liver disease (ALD). However, the intestinal role of IL-22 in alcoholic hepatitis remains to be elucidated. Methods Intestinal IL-22 levels were measured in mice fed with alcohol for 8 weeks. IL-22 was then administered to alcohol-fed mice to test its protective effects on alleviating alcoholic hepatitis, focusing on intestinal protection. Acute IL-22 treatment was conducted in mice to further explore the link between IL-22 and the induction of antimicrobial peptide (AMP). Intestinal epithelial cell-specific knockout of signal transducer and activator of transcription 3 (STAT3) mice were generated and used for organoid study to explore its role in IL-22-mediated AMP expression and gut barrier integrity. Results After alcohol feeding for 8 weeks, the intestinal levels of IL-22 were significantly reduced in mice. IL-22 treatment to alcohol-fed mice mitigated liver injury as indicated by normalized serum transaminase levels, improved liver histology, reduced lipid accumulation, and attenuated inflammation. In the intestine, alcohol-reduced Reg3γ and α-defensins levels were reversed by IL-22 treatment. IL-22 also improved gut barrier integrity and decreased endotoxemia in alcohol-fed mice. While alcohol feeding significantly reduced Akkermansia, IL-22 administration dramatically expanded this commensal bacterium in mice. Regardless of alcohol, acute IL-22 treatment induced a fast and robust induction of intestinal AMPs and STAT3 activation. By using in vitro cultured intestinal organoids isolated from WT mice and mice deficient in intestinal epithelial-STAT3, we further demonstrated that STAT3 is required for IL-22-mediated AMP expression. In addition, IL-22 also regulates intestinal epithelium differentiation as indicated by direct regulation of sodium-hydrogen exchanger 3 via STAT3. Conclusion Our study suggests that IL-22 not only targets the liver but also benefits the intestine in many aspects. The intestinal effects of IL-22 include regulating AMP expression, microbiota, and gut barrier function that is pivotal in ameliorating alcohol induced translocation of gut-derived bacterial pathogens and liver inflammation.


Introduction
Alcohol is a major causal factor for the burden of disease worldwide.Heavy and prolonged alcohol drinking damages multiple organs, especially the liver (1,2).Alcohol-associated liver disease (ALD) covers a spectrum of liver disorders involves reversible steatosis, hepatitis, fibrosis, to irreversible cirrhosis and eventually liver failure (1,2).Excessive alcohol drinking leads to the translocation of gut microbiota-derived pathogen associated molecular patterns (PAMPs), such as lipopolysaccharides (LPS), to the bloodstream, which reach the liver to elicit an inflammatory cascade and the development of alcoholic hepatitis (AH) (3,4).Early studies have observed changes in the compositions of gut microbiota (dysbiosis) in both patients with AH and animal models of AH (5)(6)(7).Furthermore, transplantation of gut microbiota from AH patients to mice directly induced severe liver damage strongly suggest a causal link between mediators derived from alcoholperturbed gut microbiota and disease progression (8).Therefore, investigation regarding the mechanisms of alcohol-induced gut dysbiosis and the underlying molecular targets are necessitated for prevention and therapy purposes.
Intestinal antimicrobial peptides (AMPs) are a variety of small peptides produced by intestinal epithelial cells (IECs) and specialized AMP-secreting cells named Paneth cells.They act as the first line of defense to protect the host against infections caused by bacteria, viruses, fungi, or parasites (9,10).Besides having a direct antimicrobial activity, some AMPs exhibit the ability to regulate host immune response and thereby indirectly eliminate infection (11,12).The critical role of AMPs in innate immunity is increasingly recognized in AH.Regenerating islet-derived protein 3b (Reg3b) and Reg3g were found to be reduced in the intestines of mice after alcohol consumption, and overexpression of Reg3g restricted alcohol-induced PAMP translocation and alleviated liver damage (13).Recently, our group reported that a-defensins, another type of AMPs, that are exclusively produced by Paneth cells are significantly decreased in a mouse model of AH (6).We further revealed the therapeutic potential of AMPs in treating AH as synthetic human a-defensin 5 (HD5) treatment corrected alcohol-perturbed gut microbiota, alleviated alcohol-induced gut barrier disruption and endotoxemia, and ultimately improves AH in mice.Despite the emerging attention to AMPs in orchestrating gut microbiota symbiosis and combatting AH, it remains unclear how alcohol alters the expression and secretion of AMPs.
Interleukin-22 (IL-22) is a cytokine produced by type 3 innate lymphoid cells (ILC3) and T cells in the intestine (14,15).It is critically involved in proliferation, regeneration, and cellular defense (16).One of the working mechanisms for IL-22 in maintaining a healthy gut is via the induction of several AMPs, including Reg3 family, lipocalin-2, S100 proteins, and b-defensins (16,17), which are all primarily produced by IECs.Importantly, intestinal IL-22 levels were reduced in the chronic-binge alcohol feeding (NIAAA) mouse model (18).In that study, the authors constructed a Lactobacillus reuteri (L.reuteri) strain overexpressing IL-22 and found that IL-22 overexpression induced Reg3g and reduced alcohol-induced liver damage.The significance of IL-22 at the gut-liver-axis in AH was also suggested by another study, which demonstrated that toll-like receptor 7 (TLR7) agonist-mediated protection was abolished by IL-22 deficiency in a mouse model of AH (19).However, the specific AMP targets of IL-22 still remain largely unknown, especially whether IL-22 instructs the antimicrobial function of Paneth cells and/or directly affects Paneth cell AMP production.Recently, Paneth cells were found to express IL-22 receptor, IL-22RA1, and mice with Paneth cellspecific Il22ra1 deficiency were susceptible to bacterial infection (20).As a result, IL-22 can modulate gut microbiota by indirectly shaping their compositions via AMP production (21) or by directly promoting the growth of commensal communities (22,23).It is, however, still unclear whether and how IL-22 orchestrates gut microbiota in mice exposed to alcohol.
IL-22 interacts with the IL-22 receptor complex named IL-22RA1 and IL-10R2, to phosphorylate and activate the signal transducer and activator of transcription (STAT) pathway, including STAT1, STAT3, and STAT5 (24).Emerging evidence suggest that epithelial STAT proteins participate in host immune defense against bacterial or viral infections as well as tissue regeneration (25-27).Citrobacter rodentium infection is observed with IL-22 elevation, epithelial STAT3 activation, and the induction of Reg3g (28).In our previous study, we found that alcohol consumption deactivated intestinal STAT3 in mice, which could be reversed by treatment with IFN-g (29).
In this study, we determined the role of IL-22 in reversing the detrimental effects of alcohol on the intestine and liver.We aimed to explore the effects of IL-22 in improving alcohol-impaired intestinal AMP production, gut microbiota symbiosis, epithelial barrier, and the subsequent PAMP translocation and liver damage.A mouse model with IEC-specific knockout of STAT3 was used to investigate the role of STAT3 in IL-22-mediated AMP production in vivo and in vitro.

Chronic alcohol feeding and IL-22 treatments
Twelve week-old male WT C57BL/6J mice were subjected to Lieber-DeCarli liquid diets for 8 weeks as previously described (30).Briefly, alcohol-fed mice (AF; n = 6) were given alcohol liquid diet and control mice were pair-fed (PF; n = 6) an isocaloric control lipid diet.Food grade ethanol was purchased from Sigma-Aldrich (St. Louis, MO) and added to the diet at 4.00, 4.14, 4.28, or 4.42% for every quarter (two-week interval).IL-22 intervention was introduced to AF mice through intraperitoneal injection of recombinant mouse IL-22 (Biolegend, San Diego, CA; Cat.#576208; n = 6) at 1 mg/kg every other day for the last 2 weeks of alcohol feeding.
To test the time effects of IL-22 on AMPs, male WT mice were intraperitoneally treated with either recombinant mouse IL-22 at 1 mg/kg or same volume of saline for 1, 3, or 8 h (n = 6 per group).Tissue samples were collected after isoflurane anesthesia in strict accordance with the protocols approved.

Organoid culture and treatments
Small intestinal crypts were isolated from PF and AF mice as well as from Stat3 fl/fl and Stat3 IEC-/-mice following a method reported by Sato et al. (31) and cultured in Matrigel (Corning, Corning, NY).Morphology of organoids was examined under light microscope during culture.After 6 days of growth, organoids were incubated with 100 ng/ml IL-22 for 10 h and harvested for gene expression and protein localization analysis by relative PCR and immunofluorescence staining, respectively.

Biochemical and histopathological analysis of liver damage
Liver damage caused by alcohol consumption were analyzed by serum transferase levels (alanine aminotransferase/ALT and aspartate aminotransferase/AST), liver histology, and hepatic lipid accumulation.Serum ALT and AST levels were quantified by Thermo Fisher Scientific Infinity Reagents (Waltham, MA), respectively.For histological examination of the liver, paraffinembedded liver sections were stained with hematoxylin and eosin (H&E; Dako-Agilent, Santa Clara, CA).Quantification assays of triglycerides (TG) and free fatty acids (FFA) in the liver were conducted using commercial kits from Abcam (Waltham, MA; Cat.#ab65336 and ab65341) per the manufacturer's instructions.Hepatic neutral lipid droplets were examined by borondipyrromethene (BODIPY) staining.Liver cryostat sections were incubated with 1 µg/ml BODIPY 493/503 (Thermo Fisher Scientific) for 20 min at room temperature followed by DAPI (4′,6-diamidino-2-phenylindole; Thermo Fisher Scientific) counterstaining after routine fixation and permeabilization.

Lipopolysaccharide quantification
Lipopolysaccharide (LPS) levels in the blood were determined using a Limulus amoebocyte lysate (LAL) method-based assay kit from Thermo Fisher Scientific, whereas LPS in the livers of mice was quantified using EndoLISA Endotoxin Detection Kit (BioVendor, Asheville, NC; Cat.#609033) as recommended by the manufacturer's protocol.The levels of serum endotoxin were expressed in endotoxin units (EU) per milliliter, whereas the levels of hepatic endotoxin were expressed as EU per milligram.

Sequencing and analysis of the bacterial 16S amplicons
Mouse cecal contents DNAs were extracted using DNeasy PowerLyzer PowerSoil kit (Qiagen, Germantown, MD; Cat.#12855) and sequenced for the V4 region of the 16s rRNA genes on Illumina MiSeq platform (Illumina Inc., San Diego, CA) (6).The acquired 16S rRNA MiSeq data were analyzed by Mothur software (v.1.39.5) (32), quality-filtered, aligned against SILVA v132 database (33), clustered into operational taxonomic units (OTU), and classified against the Ribosomal Database Project (34).Alpha diversity was calculated by observed OTUs and Shannon index.Beta diversity among the samples were explored by Bray-Curtis distance matrices.Reconstruction of unobserved states (PICRUSt) was used to predict functional genes of microbiota based on taxonomy obtained from the Greengenes reference database (35).
The categories with log linear discrimination analysis (LDA) scores of > 3.0 were considered as differential signatures than better discriminate between AF and AF+IL-22 groups.Then linear discrimination analysis effect size (LEfSe) was used to explore functional genes between groups.Lastly, the bacterial correlations coefficient in PF and AF samples with and without IL-22 treatment were calculated based on the relative abundance of each genus using Spearman.Cytoscape 3.9.0 was used for co-occurrence network building and analyzed network density and network centralization.

Gene expression analysis
RNAs of ileal or liver tissue were isolated using TRIzol reagent (Thermo Fisher Scientific) and reverse transcribed using TaqMan Reverse Transcription Reagents (Thermo Fisher Scientific; Cat.#N8080234) as recommended by the manufacturer.Relative PCR was performed using SYBR green PCR master mix (Qiagen; Cat.#330523) on a QuantStudio 5 real-time RT-PCR system.Samples were normalized to the general housekeeping gene Rn18s and calculated as relative fold change to controls (36).Primers summarized in Table 1 were synthesized by Integrated DNA Technologies (Coralville, CA).

Immunofluorescence
The levels of IL-22 in mouse ileum after alcohol exposure, hepatic inflammation status with or without IL-22 treatment, and gut barrier function in mouse ileum and isolated organoids were examined by immunofluorescence staining.Cryostat sections of mouse ileum or liver tissues were incubated with anti-IL-22 (Genetex, Irvine, CA; Cat.#GTX18498), anti-F4/80 (BD Biosciences, Franklin Lakes, NJ; Cat.#565409), anti-MPO

Gene
Genebank Accession No.

Statistics
All data are expressed as mean ± standard deviation (SD).Statistical analysis was carried out using two-tailed Student's t-test or one-way ANOVA followed by a post hoc Tukey test where appropriate using GraphPad Prism software (La Jolla, CA).*P < 0.05, **P < 0.01, ***P < 0.001.

Alcohol exposure reduces intestinal IL-22 and recombinant IL-22 treatment alleviates alcohol-induced liver damage
Our previous study demonstrated that Paneth cell-produced adefensins were significantly reduced upon alcohol exposure in a mouse model of AH (6).To explore the cause of alcohol-induced AMP reduction, we first determined intestinal IL-22 levels and found that alcohol exposure significantly reduced ileal mRNA encoding for IL-22 in mice (Figure 1A).Compared to PF mice, AF mice had a significant reduction in IL-22 + staining in the ileal lamina propria (Figure 1B).The protein levels of IL-22 averaged 67.18 ± 15.32 pg/mg in PF mice and was decreased to 13.84 ± 7.45 pg/mg in AF mice (Figure 1C).To determine whether IL-22 may reverse alcohol-induced organ damage at the gut-liver axis, we then treated AF mice with recombinant mouse IL-22 every 2 days starting from day 1 of the 7 th wk of alcohol feeding till the end of the feeding.Treatment with IL-22 significantly suppressed alcohol-induced serum levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels (Figure 1D).IL-22 also slightly decreased the ratio of liver to body weight that was significantly elevated by alcohol (Figure 1E).Compared to PF group, AF group showed substantial lipid droplets accumulation (Figure 1F; arrowheads) and inflammatory cell infiltration (Figure 1F; arrows) in the liver, which were both improved by IL-22 intervention.
BODIPY 493/503 staining further confirmed that IL-22 treatment effectively reversed alcohol-induced hepatic accumulation of neutral lipid droplets (Figure 2A).Quantification of lipids showed that alcohol consumption resulted in excessive TG and FFA depositions in mouse liver, whereas IL-22 treatment reduced hepatic TG and FFA levels by 28.7% and 38.5%, respectively, compared with AF group (Figure 2B).Alcoholinduced expression of TG synthesis enzymes, Dgat1 (diglyceride acyltransferase 1) and Dgat2, were reversed by IL-22 treatment (Figure 2C).Alcohol upregulated the expressions of FA uptake transporters, Cd36 and Fatp5 (fatty acid transport protein 5), which were reduced/normalized by IL-22 (Figure 2D).IL-22 treatment did not further affect alcohol-downregulated FA binding protein Fabp1 (fatty acid binding protein 1; Figure 2D) nor FA oxidation enzymes Cpt1a (carnitine palmitoyltransferase 1a; Figure 2E).There was a light reduction in mRNAs encoding Acox1 in AF+IL-22 group compared to that in AF group (Figure 2E).Proteins involved in very-low-density lipoprotein (VLDL) secretion, such as Apob (apolipoprotein B) and Mtp (Microsomal triglyceride transfer protein), showed no difference at mRNA levels among the groups (data not shown).
The effects of IL-22 on countering alcohol-induced liver inflammation were further analyzed.Unlike AF mice, which had dramatically increased positive staining of inflammatory cells, such as F4/80 + macrophages and MPO + neutrophils, in the liver compared with PF mice, AF mice with IL-22 treatment exhibited significantly fewer hepatic inflammatory cells (Figure 3A  correlated with the immunofluorescence findings that both were significantly upregulated by alcohol and reversed by IL-22 (Figure 3B).In line with that, IL-22 treatment decreased hepatic mRNA levels of pro-inflammatory cytokines and chemokines, including Tnfa, Ccl2, and Sele, compared with AF only group (Figure 3C).To explore whether the observed anti-inflammatory effect of IL-22 is associated with bacterial-derived signaling, we measured blood and hepatic levels of LPS, a representative marker of translocated PAMPs.As shown in Figure 3D, IL-22 normalized blood and hepatic LPS levels that were significantly elevated by alcohol.Examination of LPS signaling molecules in the liver further revealed that alcohol stimulated the expression of Lbp (LPS binding protein) and Cd14 (cluster of differentiation 14), which were both inhibited by IL-22 treatment (Figure 3E).

Recombinant IL-22 orchestrates gut microbiota favoring the enrichment of Akkermansia to combat alcohol-induced gut dysbiosis
The observed hepatic protective effects of IL-22 on restraining PAMP translocation motivated us to explore how IL-22 impacts the gut microbiota in terms of alcohol intoxication.Cecal microbiota was analyzed by metagenomic sequencing of the bacterial 16S rRNA gene.Alpha-diversity analysis showed a trend of elevation in microbial richness (P = 0.253 for observed OTUs index and P = 0.175 for Shannon) in AF mice compared with PF mice, while IL-22 treatment to AF mice had the highest levels of both alpha-diversity indexes (Figure 4A).Both alcohol and IL-22 treatment altered the structure of microbial communities as determined by b-diversity analysis.As shown in Figure 4B, Bray-Curtis distance matrices clearly separated all 3 groups into different clusters (ANOSIM test, P < 0.001, r = 0.930).We then compared the bacterial abundance between groups at the family and genus levels (Figure 4C).At the family level, Verrucomicrobiaceae is the family that showed the largest changes when comparing AF+IL-22 group with AF group.Its abundance was 3.87 ± 1.28% in PF mice, reduced to 2.25 ± 0.96% in AF mice, and dramatically increased to 17.00 ± 5.20% in AF+IL-22 mice (Figure 4C left, colored in orange).Meanwhile, IL-22 significantly enriched alcohol-inhibited Bacteroidaceae and reduced alcohol-increased Streptococcaceae.At the genus level, a similar trend was observed that Akkermansia, which belongs to Verrucomicrobiaceae, was largely expanded in the gut after IL-22 treatment (Figure 4C right, colored in purple).
The phylotypes in gut microbiota between AF and AF+IL-22 mice were further evaluated by LEfSe analysis.The cladogram showed a compelling effect of IL-22 on inhibiting a panel of bacteria in phyla of Firmicutes, Actinobacteria, and Proteobacteria, such as Deferribacteraceae, Erysipelotrichaceae, Coriobacteriaceae, Bifidobacteriaceae, Micrococcaceae, Corynebacteriaceae, and Desulfovibrionaceae when treated to AF mice (Figure 5A; Red, IL-22-decreased; Green, IL-22-increased).AF +IL-22 mice were also characterized by preponderant Verrucomicrobia and Bacteroidetes abundance.Metagenomic analysis of functional bacterial genes defined 32 dominant Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways that were markedly different between AF and AF+IL-22 groups, among which 12 KEGG pathways were more abundant in AF group and 20 KEGG pathways were more abundant in AF+IL-22 group (Figure 5B).Genes mapping for microbial energy metabolism, including oxidative phosphorylation, citric acid cycle, and pyruvate metabolism, were more frequent in AF+IL-22 mice compared with AF mice.On the other hand, bacterial genes involved in triggering host inflammatory responses, such as bacterial motility proteins, flagellar assembly, bacterial chemotaxis, and two   4C), we further analyzed sub-networks related to Akkermansia.The numbers of genera that were directly correlated with Akkermansia were 8, 5, and 7 for PF, AF, and AF +IL-22 groups, respectively (Figure 5C; Blue, Akkermansia; Green, bacteria directly correlated with Akkermansia), suggesting that IL-22 drives gut microbiota toward a beneficial direction in the host favoring the improvement of alcoholic hepatitis.

Recombinant IL-22 treatment improves alcohol-impaired AMP production, gut barrier, and intestinal STAT3 signaling
We then analyzed intestinal IL-22RA1 to validate if recombinant IL-22 could directly activate its receptor and elicit intestinal protection.Compared to PF group, AF group showed reduced intestinal IL-22RA1 + immunofluorescence staining intensity, whereas IL-22 treatment restored the signal strength (Figure 6A).Similarly, the expression of Il22ra1 was decreased after alcohol exposure and upregulated by IL-22 treatment (Figure 6B), indicating that recombinant IL-22 reached to the intestine and activated its receptor to induce local responses.Given the fact that host AMPs critically regulate gut microbiota, we next determined whether the ability of IL-22 in re-establishing gut microbiome was related to AMP regulation.Consistent with a previous report (18), intestinal Reg3b and Reg3g were both inhibited by alcohol and significantly increased by IL-22 treatment (Figure 6C).Moreover, since our group recently reported that another important type of AMPs, Paneth cell a-defensins, were also decreased by alcohol (6), we determined the changes of a-defensins and found that IL-22 treatment also effectively reversed alcohol-downregulated expression of a-defensins, including Defa4, Defa5, and Defa24, as well as their activation factor, Mmp7 (matrix metalloproteinase 7; Figure 6C).Subsequent immunofluorescence staining of DEFA5 revealed that alcohol-reduced distribution of DEFA5 in the small intestinal crypts were restored by IL-22 treatment (Figure 6D).Gut barrier integrity was determined by examining tight junction protein, ZO-1.Compared to PF mice which displayed a continuous circumferential distribution of ZO-1 along the apical part of the intestinal epithelium, reduced fluorescent intensity and disassembly of ZO-1 were found in AF mice, which was markedly improved in AF +IL-22 mice (Figure 6E).Thus, IL-22 treatment reversed alcoholinduced intestinal damage, at least partially via regulating AMP production and epithelial barrier integrity.

IL-22 directly regulates intestinal AMP production through STAT3 signaling
Intestinal STAT3 signaling was examined to explore its role in IL-22-induced intestinal protection in ALD.Alcohol suppressed STAT3 phosphorylation in the small intestine without affecting total STAT3 protein levels as indicated by Western blot, whereas recombinant IL-22 reactivated intestinal STAT3 (Figure 7A).To characterize the regulation of IL-22 on intestinal AMPs and STAT3 signaling, we performed an acute study in which single dose of recombinant IL-22 was given to WT mice, and AMPs and STAT3 were measured at different time points post treatment.As illustrated in Figure 7B, acute IL-22 treatment prompted fast and robust gene expression related to intestinal AMPs, including Reg3g, Defa24 and Mmp7, as early as 1 h post IL-22 treatment and lasted for up to 8 h.The expressions of AMPs were highest at 1 h compared to 3 h and 8 h.Meanwhile, intestinal STAT3 was significantly phosphorylated and activated by IL-22 at 1 h post treatment and returned to baseline by 8 h (Figure 7C).
Next, we generated IEC-specific STAT3 knockout mice and established corresponding small intestinal organoids to further explore the involvement of STAT3 in IL-22-mediated AMP IL-22 directly induces intestinal AMP expression via activating STAT3 regardless of microbiota.Recombinant mouse IL-22 was treated to AF mice through intraperitoneal (i.p.) injection at 1 mg/kg every other day for the last 2 wk in an 8-wk alcohol feeding experiment.Frontiers in Immunology frontiersin.orgexpression and to exclude the involvement of gut microbiota.Treatment with IL-22 to small intestinal organoids strikingly stimulated the expression of Reg3g, Defa24, and Mmp7 in WT organoids.In Stat3 IEC-/-organoids, the basal levels of Reg3g and Defa24 were lower than those in WT organoids, and IL-22-induced expression of Reg3g, Defa24 and Mmp7 were all inhibited (Figure 7D).Therefore, we concluded that IL-22 regulates intestinal AMP expression through STAT3 signaling, and the induction is independent of gut microbiota.

IL-22 stimulates intestinal epithelium differentiation that is impaired by alcohol
Lastly, we investigated the effects of IL-22 on intestinal epithelial cell differentiation in mice exposed to alcohol and in cultured organoids with or without STAT3.In the crypts of small intestine of AF mice, the number of Ki67 + cell was lower than that in PF mice, whereas that was dramatically raised in AF+IL-22 mice, suggesting an upregulated epithelial proliferation induced by IL- IL-22 stimulates IEC differentiation in a STAT3-depedent manner in mice and cultured organoids.WT mice were subjected to an 8-wk alcohol feeding, and IL-22 was i.p. injected at 1 mg/kg for the last 2 wk of feeding.8A).NHE3 is widely used as a functional marker of differentiated IECs (38), and examination of ileal NHE3 revealed reduced fluorescent intensity of NHE3 + staining at the apical part of IECs in AF mice compared with PF mice, whereas the strongest NHE3 + staining were observed in AF+IL-22 mice (Figure 8B).Small intestinal organoids isolated from PF and AF mice showed an inhibitory effect of alcohol on organoid budding time in culture (Figure 8C).IL-22 treatment to cultured organoids directly stimulated the expression of Nhe3 in WT organoids up to 12fold, which was completely blocked by IEC-specific STAT3 deletion (Figure 8D).Furthermore, immunofluorescence staining of NHE3 confirmed that IL-22-induced epithelial NHE3 expression is STAT3-dependent (Figure 8E).Immunofluorescence staining of NHE3 showed that IL-22 stimulated NHE3 expression in WT organoids.Compared to the levels of NHE3 + staining in WT organoids, NHE3 + staining intensities were much lower in Stat3 IEC-/-organoids with or without IL-22 treatment (Figure 8E).Taken together, besides its role in the regulation of intestinal AMPs, IL-22 also promotes intestinal epithelium differentiation that is inhibited by alcohol intoxication, which represents another line of protection in the intestine to prevent systemic PAMP translocation induced by alcohol.

Discussion
Altered gut microbial ecosystem is associated with metabolic and immune disorders, including alcoholic hepatitis, obesity, and metabolic-dysfunction-associated steatotic liver disease (39).Upon alcohol intoxication, the interplay between the host and gut microbiota is perturbed by mechanisms that remain incompletely understood.IL-22 is increasingly recognized in promoting intestinal homeostasis (14,16,17), and previous studies reported that short-term alcohol exposure reduces intestinal IL-22 levels in rodents (18,40).The study demonstrates the protective effects of IL-22 in both the intestine and the liver against alcohol toxicity in an 8week alcohol feeding mouse model (summarized in Figure 9).We found that intestinal IL-22-STAT3 signaling is an important regulator for intestinal AMP production, gut microbiota symbiosis, and epithelial barrier function that are disrupted by alcohol.The study provided experimental evidence that IL-22 treatment is effective in alleviating alcohol-induced PAMP translocation and hepatic inflammation.
IL-22 has been proposed as a promising therapeutic target in treating many gastrointestinal diseases (41).The direct hepatoprotective effects of IL-22 have been investigated in multiple models of liver injuries, including T cell-mediated hepatitis (42), acetaminophen-induced liver injury (43,44), fatty liver disease (45)(46)(47), and AH (48).A previous study revealed that IL-22-mediated hepatoprotection in ALD is involved in the stimulation of proliferation, inhibition of apoptosis, and prevention of fatty acid uptake in hepatocytes (48).Our current study indicated that IL-22 downregulates hepatic expression of several lipid metabolism genes involved in TG synthesis and fatty acid uptake, which may involve both direct effect of IL-22 to the liver and indirect signals from the gut-derived microbial products, such as LPS.In fact, LPS has been reported to directly and/or synergistically induce lipid disorder and promote fatty liver (49,50).An earlier study found that suppression of LPS binding protein improves obesity-associated fatty liver (51).
Our study revealed a profound protective role of IL-22 in the intestine following alcohol exposure via regulating AMP production (both Reg3 and a-defensins), promoting the growth of commensal bacteria, and enhancing gut barrier function.Hendrikx et al. reported an intestinal protective effect of IL-22 against AH using bioengineered bacteria overexpressing IL-22 (18).The authors found that treatment with IL-22-producing L. reuteri upregulated intestinal Reg3g levels and alleviated alcohol-induced liver injury.Administration of IL-22 was also protective in a mouse model of ethanol and burn injury (40).Although the whole story for IL-22mediated beneficial effects on the intestine is not completely understood, our findings provide more comprehensive knowledge supporting the feasibility of IL-22 in clinical use for AH treatment.
One of the important findings in the study is that IL-22 directly regulates the expression and activation of a-defensins.Previous studies investigating the effects of IL-22 on AMP induction mostly focused on a few preselected AMPs, such as Reg3 (18).In the present study, in addition to the Reg3 family, we also measured adefensins and their activation factor, MMP7, in mouse models of chronic alcohol feeding with IL-22 treatment, in acute IL-22 administration, and in small intestinal organoids.We observed that IL-22-mediated a-defensin expression is independent of gut microbiota.Unlike most AMPs that require bacterial stimulation for robust production, a-defensins are constitutively expressed even in germ-free mice (52), which suggests a fundamental role of adefensins in intestinal homeostasis.However, the regulatory mechanisms of a-defensin expression are still unclear.By implementing the IEC-specific STAT3 knockout mouse model, we identified that IL-22 induces a-defensin expression through STAT3.
IL-22 impacts not only pathogenic bacteria to prevent infection, but also commensal gut microbiota to orchestrate microbial symbiosis (22,23).In IL-22RA1 deficient mice, a dramatic depletion of Lactobacillaceae, Bifidobacteriaceae, and Ruminococcaceae families and an overgrowth of opportunistic pathogen Enterococcus spp.were detected (22).To the best of our knowledge, this study is the first report showing that IL-22 induces robust bloom of commensal Akkermansia spp. in the gut.Akkermansia spp. is commonly found in the gut of human and mouse, ranging 2-5% of the microbial community (53).We found that the bacterium makes up to 17% of the total cecal microbiota after IL-22 administration.Akkermansia spp. is more abundant in healthy subjects than in patients with obesity and diabetes (54), inflammatory bowel diseases (55), and ALD (37).Accumulating evidence from animal and human studies suggest that Akkermansia has a promising potential for the prevention and treatment of diseases, including ALD (37).The precise mechanism by which IL-22 exerts Akkermansia spp.expansion is obscure.One proposed mechanism could be a better nourished microenvironment provided by IL-22 that promotes the growth of Akkermansia.IL-22 is known to regulate many mucus-associated genes, such as mucin-1 (MUC1), MUC3, MUC10, and MUC13 (16), whereas Akkermansia spp. is a mucin-degrading bacterium (56).Another possibility is the regulatory effect of AMPs, especially a-defensins.Besides direct bacterial killing, AMPs have been recently recognized to possess immunoregulatory effects, such as neutralizing pathogenic toxins, modulating immune cell differentiation, and initiating adaptive immunity (10).We previously reported that synthetic HD5 treatment to AF mice induced Akkermansia spp. in mouse cecum for over 10 fold (6), a phenomenon similar to what was observed in IL-22-treated AF mice.Interestingly, another group reported that HD5 treatment to non-treated WT mice also dramatically increased Akkermensia spp.without altering microbial diversity (57).A third mechanism would be a secondary effect of IL-22-improved gut microenvironment that is dampened by alcohol.However, this hypothesis cannot explain why the levels of Akkermansia spp. in alcohol plus IL-22 group were further elevated.Therefore, future studies are needed to explore the causal-effect mechanism of how IL-22 promotes Akkermansia spp.expansion.Our findings that IL-22 treatment favors Akkermansia spp.provide new insights into IL-22-mediated protection against AH at the gut-liver axis.
Although it has been reported that IL-22 regulates gut barrier function, the precise functional mechanisms of IL-22 in the intestine remain largely unknown.Here, we report that NHE3 is a direct target of IL-22 via STAT3 signaling, which is impaired by alcohol exposure.NHE3 is used as a marker for differentiated epithelium (38).It plays an essential role in maintaining sodium and water homeostasis in the intestine yet the regulatory mechanisms of NHE3 are not well-defined (58).It is worth mentioning that NHE3 is essential for gut microbial symbiosis as mice deficient in NHE3 have microbial dysbiosis and develop spontaneous colitis (59,60).Hence, we believe that IL-22mediated NHE3 induction not only indicates a welldifferentiated gut barrier that prevents the penetration of PAMPs but also provides another mechanism that orchestrates gut microbiota.The crosstalk between IL-22, epithelial NHE3, gut microbiota, and even AMPs in the pathogenesis of AH warrants further investigation.
It remains largely unknown how alcohol reduces intestinal IL-22 levels.IL-22 is mainly produced by T helper type 17 (Th17), Th22, and innate lymphoid type 3 (ILC3) cells.Other potential IL-22-producing cells include gd T cells, natural killer cells, and neutrophils (14)(15)(16).IL-22-expressing ILC3 cells were reported to be significantly reduced without affecting the levels of IL-22 repressor, IL-25, in mice after alcohol exposure (18).In the same study, intestinal levels of indole-3-acetic acid (IAA), which is an aryl hydrocarbon receptor (AhR) ligand that regulates IL-22 expression, was reduced by alcohol feeding (18).More work is warranted to explore the cellular and molecular mechanisms of intestinal IL-22 reduction in the pathogenesis of ALD.In conclusion, our data indicate that administration of IL-22 promotes intestinal innate defense via regulating AMP production, orchestrating gut microbiota symbiosis, and stimulating IEC differentiation, all of which contribute to a profound effect on inhibiting the translocation of gut-derived PAMPs and subsequent hepatic inflammation.The study strongly supports the potential clinical utility of IL-22 targeting both the intestine and liver in ALD treatment.Persistently upregulated IL-22 can promote inflammation and accelerate tumor growth (61, 62).Elucidating factors determining whether IL-22 is protective or inflammatory is of high priority and necessitates further investigation.

4 IL- 22
FIGURE 4 IL-22 treatment increases mouse cecal bacteria diversity and shifts the composition that are perturbed by alcohol.(A) Alpha-diversity measurements of observed number of OTUs and Shannon index.*P < 0.05.(B) PCoA plot showing dissimilarity in bacterial community structures based on Bray-Curtis distances.(C) Barplot showing the cecal bacterial compositions at family level (left) and genus (right) level, respectively.PF, pair-fed; AF, alcohol-fed.

5
FIGURE 5 Administration of IL-22 results in distinct cecal microbial and functional variations other than that associated with alcohol intoxication in mice.(A) Cladogram taxonomic abundances of bacteria.Legend of prominent taxons is shown at bottom.(B) Bacterial gene functions predicted based on 16S rRNA gene sequences using the PICRUSt algorithm and annotated from KEGG databases.Red: bacterial genera enriched in AF mice; Green, bacterial genera enriched in AF+IL-22 mice.(C) Bacterial co-occurrence network revealed the key role of Akkermansia spp. in formation of bacterial correlation network in intestinal microbiota.Akkermansia spp. is highlighted as blue circle, while those genera directly correlated with Akkermansia spp.are colored as green.Solid line, positive correlation; dash line, negative correlation.
fl/fl ; Stock No. 016903) were purchased from the Jackson Laboratory (Bar Harbor, ME).Intestinal epithelial cell (IEC)-specific STAT3 (Stat3 IEC-/-) mice were generated by breeding corresponding floxed mice with Villin997-Cre mice (the Jackson Laboratory; Stock No. 004586).Wild type (WT) C57BL/6J mice (Stock No. 000664) were also obtained from the Jackson Laboratory.All mice were kept in ventilated cages under specific pathogen-free condition.Animal experiments were approved by the Institutional Animal Care and Use Committee of North Carolina Research Campus (Protocol #21-008).

TABLE 1
Primer sequences used for qPCR analysis.