Sirt6^fl/fl mice and Rag1^-/- mice were purchased from the Jackson Laboratory. Rorc^Cre mice was provided by Ju Qiu from the Shanghai Institute of Nutrition and Health in the Chinese Academy of Sciences which was purchased from the Jackson Laboratory. The mice used in this study were C57BL/6 background. Sirt6^fl/fl mice possess loxP sites flanking exons 2-3 of the Sirt6 targeted gene. Rorc^Cre mice express cre recombinase under the control of the mouse Rorc (RAR-related orphan receptor gamma; also called RORγt) promoter. Rag1^-/- mice produce no mature T cells or B cells. Sirt6^fl/fl and Rorc^Cre mice were crossed to obtain a Sirt6^ΔRorc strain. Sirt6^ΔRorc and Rag1^-/- mice were crossed to obtain a Rag1^-/-Sirt6^ΔRorc strain. All mice were bred and maintained at accredited animal facilities under specific-pathogen-free conditions in individually ventilated cages on a strict 12-hour day–night cycle with a regular chow diet. Mice used in this study were 6–8 weeks old and sex-matched, unless otherwise indicated in the text. Mice were maintained at Shanghai Jiao Tong University School of Medicine. All animal experiments were performed in compliance with the Guide for the Care and Use of Laboratory Animals, with the protocols approved by the Institutional Animal Care and Use Committee of Shanghai, China [SYXK(SH)2023-0041].

LPLs were purified as previously described (30). Briefly, mouse intestines (with Peyer’s patches removed) were opened lengthwise and cut into 2 cm pieces. Intestinal pieces were washed with 1 mM DTT/PBS on a rocking platform for 10 min and then intensely shaken for 2 min. Repeat the last operation with 30mM EDTA/PBS. The remaining intestinal pieces were further chopped into 5 mm pieces and digested with RPMI media containing 100U/ml of Collagenase VIII (Sigma-Aldrich, C-2139) and 150ug/ml of DNase I (Sigma-Aldrich, DN25) at 37 °C in a 5% CO₂ incubator for 90min. The preparations were filtered with a 70µm cell strainer and subjected to density gradient centrifugation using 40% and 80% Percoll solutions (GE Healthcare) at 2500rpm for 20min. Lymphocytes were collected from the interface and washed once.

C. rodentium strain DBS100 was cultured overnight in Luria-Bertani (LB) broth shaking at 37°C. Adult male Sirt6^fl/fl and Sirt6^ΔRorc littermates or Rag1 ^-/- Sirt6^fl/fl and Rag1 ^-/- Sirt6^ΔRorc littermates (6-8 weeks) were orally gavaged with 10¹⁰ colony-forming units (CFU) of C. rodentium in 0.2 mL PBS. At five days post-inoculation, Feces were collected, weighed, homogenized, serially diluted and plated to determine the CFU. At six days post-inoculation, mice were sacrificed, and the colons were collected and divided into 3 pieces. One piece was flash-frozen for RNA analysis and the second was processed for flow cytometry analysis, and the third was fixed with 4% paraformaldehyde for histological analysis.

Sirt6^fl/fl and Sirt6^ΔRorc mice at the age of 6-8 weeks were oral administration of 2.5% DSS (MP Biomedical) in the drinking water for 7 days, followed by 3 days of normal drinking water. Body weight, the presence of occult or gross blood per rectum, and stool consistency were determined daily. Changes in body weight were indicated as loss of baseline body weight as a percentage. The scoring system described by Kim et al. was used to determine the disease activity index (DAI) (31). The DAI is the combined score of weight loss compared to initial weight, stool consistency, and bleeding. Scores are defined as follows: weight loss: 0 (no loss), 1 (1-5%), 2 (5-10%), 3 (10-20%), and 4 (>20%); stool consistency: 0 (normal), 1 (soft), 2(very soft) and 3 (Watery); and bleeding: 0 (no blood), 1 (hemoccult positive), 2 (hemoccult positive and visual pellet bleeding), and 3 (gross bleeding, blood around anus). DAI can be scored daily during the duration of the DSS treatment. Colons were collected and divided into 2 pieces at 10 days after DSS treatment. One piece was processed for flow cytometry analysis, and another one was fixed with 4% paraformaldehyde for histological analysis.

Cells were resuspended in PBS and stained with fixable viability dye (BD Biosciences) at room temperature for 15 min to exclude dead cells. After washing with PBS, Cells were resuspended in FACS buffer and incubated with Fc Block (Biolegend,1:50) for 5 min at 4°C to block Fc receptors. For surface staining, cells were incubated with antibody cocktails for 30 minutes at 4°C in the dark. For intracellular cytokine staining, cells were stimulated with 50ng/ml of PMA (Sigma-Aldrich) and 500ng/ml of ionomycin (ENZO) or 10ng/ml of IL-23 for 4 hours and 2 µg/ml of Brefeldin A (Biolegend) was added for the last 2 hours. After surface staining, cells were fixed and permeabilized using the Foxp3/Transcription Factor Staining Buffer Set (eBioscience), and stained with antibody for 45 minutes at 4°C in the dark. For cell sorting, ILC3s were identified as Lin^-CD90^hiCD45^int live lymphocytes using a FACSAriaIII (BD Biosciences) (32–35). Lineage markers include CD3, Gr-1, B220, CD11b, and CD11c. All reagents used were listed in Supplementary Table S1 and all antibodies used were listed in Supplementary Table S2 . Flow cytometry data were acquired using LSRFortessaX20 (BD Biosciences) and analyzed with Flowjo software (version 10).

Sorted ILC3s were cultured in 96-well flat-bottom plates in RPMI 1640 complete medium containing 10% fetal bovine serum, 2mM L-glutamine, 50μM β-mercaptoethanol, 1% MEM NEAA, 1% penicillin/streptomycin (all from Gibco or Sigma-Aldrich), in the presence of 10ng/ml IL-7 (Peprotech) (36).

The parental MNK cell lines are derived from NKR^-P1B⁺ fetal thymocytes. Among them, MNK3 is an IL-7-responsive subline that expresses Rorγt and produces high levels of IL-22 in response to IL-23 and IL-1β stimulation (37). The MNK3 cell line was previously described as an in vitro system to study ILC3 functionality (37). MNK3 cells were cultured in Dulbecco’s modified Eagle’s medium-high glucose medium with 10% fetal bovine serum, 2mM GlutaMAX, 1mM sodium pyruvate, 55μM β-mercaptoethanol, 10mM HEPES, 100μg/ml streptomycin, and 1% penicillin/streptomycin (all from Gibco or Sigma-Aldrich). MNK3 cells were cultured in the presence of 10ng/ ml mouse IL-7 (Peprotech) and 10ng/ml mouse IL-15 (Peprotech).

Both ILC3s and MNK3 were cultured at 37°C in 5% CO2 humidified air.

For inhibitor treatment, sorted ILC3s and MNK3 were treated with OSS-128167 (100μM, Selleck) for 18 hours in vitro and then stimulated with IL-23 at 10ng/ml for another 4hours.The percentage of IL-22⁺ ILC3s or MNK3 was determined by flow cytometry.

Total RNA was isolated with Trizol (Invitrogen). RNA concentration was detected by Nanodrop Spectrophotometer (Nanodrop Technologies). cDNA was synthesized using PrimeScript™ RT reagent Kit (Takara). Quantitative real-time reverse transcription qRT-PCR was performed using SYBR™ Select Master Mix (Applied Biosystems™) on a ViiA7 Real-Time PCR System. Results were normalized as hprt through the ΔΔCt method. Primer sequences were shown in the Supplementary Table S3 .

Flush the colon with 10 mL of ice-cold phosphate-buffered saline (PBS) with a 10 mL syringe equipped with a gavage needle to remove the feces and blood until the eluate is completely clear. Cut it open longitudinally. Place the intestinal segment with the luminal side facing upward. Proceed with Swiss rolling the colon (38). Then fix the tissue with 4% paraformaldehyde overnight at 4°C for paraffin embedding. For staining, cut 4um thick sections using a microtome. The sections were stained with hematoxylin staining solution for 5 min, washed with tap water, differentiated with differentiation solution, and returned to a blue solution. The slices were dehydrated with 85% and 95% gradient alcohol for 5 min respectively, and stained with eosin staining solution for 5 min. Sections were observed and photographed using the OLYMPUS BX53 system.

Statistical analyses were performed using the GraphPad Prism 8.0 program. Unless otherwise noted, the Kolmogorov-Smirnov test was performed for normality tests. Statistical significance was tested using two-tailed unpaired Student’s t-test as parametric tests or Mann-Whitney test as nonparametric tests. For two independent variables, the differences were measured by two-way ANOVA. Data from these experiments are presented as the mean values ± SEM. P values are noted in each figure.

To investigate the role of SIRT6 in gut ILC3s, we crossed Sirt6^fl/fl mice with Rorc^Cre mice to knock out SIRT6 in ILC3s (Sirt6^ΔRorc ) ( Supplementary Figure S1 ). Genotyping of mice was performed by PCR ( Supplementary Figure S1A ). The efficiency of SIRT6 deletion in ILC3s exceeded 75% ( Supplementary Figure S1B ). We found that deletion of SIRT6 did not affect body weight and colon length in mice ( Supplementary Figures S1C, D ). We then examined the effect of SIRT6 on the ILC3 compartment in the small intestine. The Gating strategy of ILC3 is shown in Figure 1A as previously described (6, 32). Neither the frequency nor the absolute number of ILC3s was affected in Sirt6^ΔRorc mice compared with Sirt6^fl/fl control mice ( Figures 1B, C ). The composition of ILC3 subsets, including NKp46⁺, CCR6⁺, and CCR6⁻NKp46⁻ had no change in Sirt6^ΔRorc mice as well ( Figures 1D-G ). Given that ILC3s exert immune function through the production of effector cytokines, we examined the expression of IL-22, IL-17A, GMCSF, and IFN-γ by ILC3s. Intestinal ILC3s were treated with activating cytokine IL-23. We observed that SIRT6-deficient ILC3s produced more IL-22, while IL-17A, GMCSF, and IFN-γ were unaffected ( Figures 2A-H ). Together, these data suggest that SIRT6 suppresses IL-22 production by ILC3s in the gut.

We next determined the role of SIRT6 in ILC3s during C. rodentium infection. Mice were challenged by oral gavage with 10¹⁰ CFU of C. rodentium. Compared to Sirt6^fl/fl littermate controls, Sirt6^ΔRorc mice controlled bacterial infection more efficiently, as indicated by longer colon length and reduced bacterial load (CFU counts) in feces ( Figures 3A, B ). Consistently, Sirt6^ΔRorc mice exhibited reduced tissue inflammation characterized by decreased expression of pro-inflammatory cytokines including Il-6, Tnf, and Il-1β, compared to the Sirt6^fl/fl littermate controls ( Figure 3C ). As IL-22 has a crucial role in the early phase of host defense against C. rodentium, we examined the expression of IL-22 in ILC3s and T cells in the large intestine by flow cytometry ( Figures 3D-F ). As expected, infection with C. rodentium induced the expression of IL-22 in ILC3s. On day 6 post-infection, ILC3s from Sirt6^ΔRorc mice produced more IL-22 than Sirt6^fl/fl littermate controls ( Figures 3D,E ). Since SIRT6 was deleted in both ILC3s and T cells in Sirt6^ΔRorc mice, IL-22 production by T cells was assessed during C. rodentium infection. We did not observe significant differences in IL-22-expressing T cells between Sirt6^fl/fl and Sirt6^ΔRorc mice ( Figures 3D, F ). Since IL-22 protects epithelial function by stimulating the secretion of antimicrobial peptides, we next analyzed the expression of the genes encoding the antimicrobial peptides. The expression of RegIIIβ and RegIIIγ were increased in the Sirt6^ΔRorc mice ( Figure 3G ). Consistently, Sirt6^ΔRorc mice exhibited reduced tissue inflammation, as indicated by less immune cell infiltration in H&E staining ( Figure 3H ). Thus, SIRT6 plays a critical role in regulating ILC3 function during C. rodentium infection.

Since Rorc-driven Cre recombinase is expressed in both T cells and ILC3s, we investigated whether T cells were affected in Sirt6^ΔRorc mice. Sirt6^ΔRorc mice did not show any defects in the small intestine T cells. Neither the frequency nor the absolute number of RORγt⁺ T cells (gating on Gr-1^-B220^-CD11b^-CD11c^-CD45⁺CD3⁺RORγt⁺) was affected in Sirt6^ΔRorc mice compared with Sirt6^fl/fl control mice ( Supplementary Figures S2A, B ). Furthermore, the composition of RORγt⁺ T cells subsets, including RORγt⁺ Th17, RORγt⁺ Treg, and RORγt⁺CD4^- T cells had no change in Sirt6^ΔRorc mice ( Supplementary Figures S2A, C ). LPLs from the small intestine were stimulated with PMA/ionomycin and followed by intracellular staining for cytokines production. The expression of IFN-γ, IL-4, and IL-17A by CD3⁺ T cells showed no differences between Sirt6^fl/fl and Sirt6^ΔRorc mice ( Supplementary Figures S3A-F ). By contrast, IL-22-expressing ILC3s were largely increased despite normal IL-22 production by T cells in Sirt6^ΔRorc mice ( Supplementary Figures S3G-I ). Thus, the T cell compartment in the gut was not affected in Sirt6^ΔRorc mice.

To completely rule out the impact of SIRT6-deficient T cells on ILC3s, we crossed Sirt6^ΔRorc mice to Rag1^−/− mice lacking the adaptive immune system. Consistent with the findings in Sirt6^ΔRorc mice, SIRT6-deficient ILC3s had elevated IL-22 production in Rag1^-/-Sirt6^ΔRorc mice. Next, mice were infected with C. rodentium by gavage for 6 days. As expected, Rag1^-/-Sirt6^ΔRorc mice had longer colon length ( Figure 4A ), reduced bacterial load ( Figure 4B ), increased IL-22-producing ILC3s ( Figures 4C, D ), and enhanced expression of RegIIIβ, RegIIIγ ( Figure 4E ). Together, these data support the conclusion that SIRT6 regulates ILC3s in a cell-intrinsic manner during C. rodentium infection.

To further determine the role of SIRT6 in regulating ILC3 function during gut inflammation, Sirt6^ΔRorc mice were treated with 2.5% DSS to induce acute colitis ( Figure 5A ). Compared to their littermate controls, Sirt6^ΔRorc mice exhibited less body weight loss, lower DAI scores, and longer colon length ( Figures 5B-D ). Moreover, Sirt6^ΔRorc mice exhibited decreased inflammation, as indicated by fewer infiltrated immune cells in H&E staining ( Figure 5E ). These data suggested that the deletion of Sirt6 in ILC3s protects mice from DSS-induced colitis. At the cellular level, we observed enhanced IL-22⁺ ILC3s in Sirt6^ΔRorc mice ( Figure 5F ). However, levels of IL-17 were comparable between SIRT6-deficient ILC3s and control ILC3s ( Figure 5G ). Together, these data indicate that SIRT6 negatively regulates ILC3 function during gut inflammation.

We also found that the mRNA level of IL-22 was increased in SIRT6-deficient ILC3s, suggesting that SIRT6 regulates IL-22 expression at the transcriptional level ( Figure 6A ). To further investigate whether SIRT6 could directly suppress ILC3 function in vitro, a cell-permeable SIRT6-specific inhibitor, OSS-128167, was used to treat ILC3 culture in the presence of IL-23. After treatment with OSS-128167, ILC3s produced more IL-22 ( Figures 6B, C ). In addition, we used the ILC3-like MNK3 cell line, which mimics the phenotype and effector function of murine primary intestinal ILC3s. Similar to the findings observed in ILC3s, in vitro OSS-128167 treatment enhanced IL-22 production ( Figures 6D, E ). These data suggest that SIRT6 directly inhibits IL-22 production in ILC3s.