Myeloid-specific A20-deficient mice (A20^myel-KO) develop spontaneous arthritis and show signs of systemic inflammation, including splenomegaly (12). Despite their pro-inflammatory phenotype, A20^myel-KO mice do not develop spontaneous intestinal inflammation ( Supplementary Figures 1A–C ), as previously reported (18). To study intestinal mucosal immunity in more detail, we performed immune phenotyping of the colonic lamina propria in A20^myel-KO and wild-type (WT) littermate control mice by flow cytometry and showed that myeloid-specific A20 deletion leads to the expansion of various myeloid immune cell subsets, including eosinophils, neutrophils, monocytes, and macrophages, compared with WT littermate controls ( Figures 1A, B ). In addition, colonic lamina propria T-cell subsets in A20^myel-KO mice are characterized by increased infiltration of CD3⁺ and CD4⁺ T cells and, more specifically, by an expansion of Th17 (CD4⁺RORγt⁺) and T regulatory (CD4⁺Foxp3⁺) cells ( Figures 1C, D ). The lack of spontaneous inflammation in the colon, despite increased immune cell infiltration, can be attributed to the high levels of immunosuppressive T regulatory cells. Similar immune cell composition is also observed in A20^myel-KO spleens ( Supplementary Figures 1D–K ). In order to obtain more insights into myeloid immune cell function, we performed bulk RNA sequencing on sorted myeloid cells (CD11b⁺) from colon lamina propria ( Figures 1E–H ; Supplementary Figure 2F ) and joint (combined ankle and knee synovium) tissue ( Supplementary Figures 2A–E ) from A20^myel-KO and WT littermates in steady state. The top upregulated genes in colonic A20^myel-KO CD11b⁺ cells are involved in IFN signaling and classical macrophage differentiation (M1), including Cxcl9, Gbp5 (19), Saa3, Gzma, and Nos2 ( 20) ( Figures 1E, F ). Gene Ontology (GO) pathway analysis revealed that the expression profiles of colonic myeloid cells of A20^myel-KO mice are associated with macrophage tolerance and cytokine pathways including IL-12, IL-23, IL-22, IL-17, IFN-γ, and TLR4 signaling ( Figure 1G ). Gene scoring analysis using an “M1 signature calculation” (21) indeed showed a significantly higher M1 score in the A20^myel-KO colonic myeloid compartment compared with WT colonic myeloid cells ( Figure 1H ). RA patients show an enhanced M1/Th1 profile which is characterized by increased TLR signaling and secretion of inflammatory cytokines such as TNF, IL-1β, IL-18, and IFN-γ (22). Accordingly, analysis of A20^myel-KO synovial myeloid cells (CD11b⁺) revealed transcriptional upregulation of M1-related genes, including different guanylate-binding protein (Gbp) family members (typically induced by IFN-γ) (23), IFN-γ-inducible GTPases (Gm4841) (24), acute phase response markers (such as Saa3) (25), and TLR1 pathway genes (26). Moreover, GO pathway analysis revealed that positively regulated pathways in the synovial myeloid compartment of A20^myel-KO mice include IFN signaling, proinflammatory cytokine signaling (IL-1, IL-6, IL-17, IL-23), NF-κB and inflammasome signaling, along with a significantly upregulated M1 gene signature ( Supplementary Figures 2A–E ). Together, these data indicate that A20 deletion in myeloid cells induces M1 priming in both intestinal and synovial macrophages.

We next assessed whether type-2 immune activation can suppress M1 macrophage polarization and type-1 immunity upon A20 deficiency and evaluated the sensitivity of A20^myel-KO mice to an intestinal parasitic worm (helminth) infection since helminth clearance critically depends on type-2 immunity (27). For this purpose, adult A20^myel-KO and WT littermate control mice were inoculated with 200 embryonated T. muris eggs by oral gavage and were sacrificed for analysis 21 days later ( Figure 2A ). While WT mice induce robust type-2 immune responses and clear T. muris from the intestine to re-establish homeostasis within 21 days (28), A20^myel-KO mice show defects in worm clearance and still contain approximately 150–200 T. muris parasites in their cecum 21 days post-infection ( Figure 2B ). A20^myel-KO mice display persistent infection with T. muris and intestinal inflammation characterized by severe immune cell infiltration, tissue damage, and edema ( Figure 2C ). We also show that GATA3 and IL-4, two prototype type-2 immunity markers, are expressed in lower levels in A20^myel-KO colon lysates from T. muris-infected mice compared with WT littermates ( Figure 2D ). Restimulation of mesenteric lymph node (mLN) cells 21 days post-infection (p.i.) with anti-CD3/CD28 induces the production of type-2 cytokines, such as IL-4, IL-5, and IL-13, in cells derived from WT mice but not in A20^myel-KO mLN cells ( Figure 2E ). WT-derived mLN cells are also characterized by higher levels of CD4⁺GATA3⁺ cells ( Figure 2F ). To further evaluate the induction of type-2 immunity in response to T. muris infection, we quantified T. muris-specific serum IgG1 (type-2-associated) and IgG2c (type-1-associated) immunoglobulins. We found IgG1 to be significantly increased in the serum of WT mice after T. muris infection but not in the serum of A20^myel-KO mice. In contrast, T. muris-specific type-1-associated IgG2c immunoglobulins are increased in A20^myel-KO mice but absent in WT mice ( Figure 2G ). We also show increased CD19 and IgG1 levels in WT colon sections after T. muris infection, which is not the case in A20^myel-KO colon tissue, in line with our findings for serum IgG1 ( Figure 2H ). Immune profiling of colonic lamina propria immune cells after T. muris infection (21 days p.i.) confirmed the inability of A20^myel-KO mice to mount an effective Th2 response, in contrast to WT littermates. We detected increased immune cell (CD45+) infiltration in A20^myel-KO colonic lamina propria, with enhanced levels of neutrophils, monocytes, and macrophages, compared with WT controls ( Figures 3A, B ). In the T-cell compartment, WT mice exhibit significantly elevated levels of CD4⁺GATA3⁺ (Th2) cells, compared with A20^myel-KO mice. In contrast, A20^myel-KO mice show a skewed Th1/Th17 response, as CD4⁺Tbet⁺ and CD4⁺RORγt⁺ cells are shown to be increased ( Figures 3C, D ). Similar findings are observed in the spleen, confirming a systemic immune response ( Supplementary Figures 4A, B ). It has been reported that the phenotype of the recruited T regulatory cells (Tregs) is indicative of the associated adaptive responses and that Tbet expression in T regulatory cells supports their homeostasis and activity in a type 1-mediated inflammatory context (29). We observe that WT-derived Tregs express higher levels of GATA3, while A20^myel-KO-derived Tregs express higher levels of Tbet, supporting Th2 and Th1 polarization, respectively ( Supplementary Figures 4D , 5 ). Multiplex imaging on colon sections confirmed the increased number of colonic CD45 cells in WT tissue after infection with T. muris; however, immune cell infiltration is even more increased in colons of infected A20^myel-KO mice. Similar recruitment patterns are observed for CD11b⁺ myeloid cells and CD64⁺ macrophages ( Figure 3E ). Finally, we performed RNA sequencing on sorted CD11b⁺ cells from the colon of WT and A20^myel-KO mice after T. muris infection. Samples were isolated and processed 10 days after infection (D10) in order to investigate myeloid cell responses during the peak of infection and anti-helminth immune response. Interestingly, we observe a totally different expression profile for the sorted CD11b⁺ cells of both genotypes ( Supplementary Figure 4C ). In the A20 knockout myeloid cells, we again observe the upregulation of markers related to classical macrophage activation such as Nos2, CD300e (30), Bmx (31), and Plekhg4 ( Figure 4A ). GO analysis confirms the activation of pathways involved in IFN and TLR4 signaling and IL-1β, IL-12, and IL-22 signaling, which are all related to macrophage polarization toward classical activation ( Figure 4B ). In contrast, in WT mice, we observe the upregulation of markers related to alternative macrophage differentiation such as CD163 (32), Nmb (33), and Cpeb1 (34). GO analysis supports this notion and indicates upregulation of pathways related to IRF4 targets, PPARγ activation, and IL-4 and GATA3 signaling ( Figures 4A, B ). In order to further evaluate the difference in macrophage polarization, we compared the gene expression profiles of WT and A20^myel-KO mice to existing gene signature datasets of macrophage gene expression (21, 35). This confirms that WT mice show enhanced expression of genes involved in alternative macrophage activation (M2-like), whereas A20^myel-KO mice show a clear polarization toward classical macrophage activation (M1) upon T. muris infection ( Figures 4C, D ), as observed in steady state. Moreover, gene set enrichment analysis (GSEA) confirmed the induction of an M1 phenotype and the downregulation of an M2-like phenotype in A20^myel-KO mice compared with WT mice ( Figure 4E ). Collectively, these data clearly indicate that upon T. muris infection, WT myeloid cells shift to an alternatively activated state, supporting proper type-2 immune polarization, pathogen clearance, and resolution of inflammation. In contrast, A20-deficient myeloid cells are highly polarized toward classical activation already in steady state, and this immune polarization is not altered upon T. muris infection. In conclusion, loss of A20 prevents myeloid cell polarization toward an M2-like state and subsequently impedes proper type-2 immune activation, which prevents parasite clearance and leads to persistent infection and inflammation upon helminth infection.

The Th1-associated cytokine IFN-γ is a very potent antagonist of type-2 polarization and is induced by upstream myeloid-derived cytokines, including IL-12 and IL-18, which, as mentioned earlier, are elevated in A20^myel-KO mice ( Supplementary Figure 2G ). Ex-vivo differentiated bone marrow-derived macrophages (BMDMs) of WT and A20^myel-KO mice were stimulated with T. muris antigens and analyzed at different time points. A20^myel-KO BMDMs express significantly higher levels of Th1 polarizing cytokines IL-12 and IFN-γ ( Figures 5A, B ) and secrete increased levels of IL-18 ( Figure 5C ) in the supernatant upon T. muris antigen stimulation. In order to evaluate the role of IL-12, IL-18, and IFN-γ in preventing an effective type-2 immune response upon T. muris infection in A20^myel-KO mice, we assessed the outcome of pharmacological inhibition of these cytokines in vivo. To this end, A20^myel-KO mice were infected with T. muris and i.p. injected with neutralizing antibodies against IFN-γ, IL-12, or IL-18 versus isotype control antibody (IgG), every 48 h for a total of 21 days ( Figures 5D, E ). Interestingly, depletion of either one of these three cytokines leads to a significant reduction in the worm counts in the cecum compared with isotype-antibody-treated A20^myel-KO mice, with the most potent reduction in anti-IFN-γ-treated mice ( Figure 5E ). We further determined the levels of IgG1 and IgG2c antibodies, as immunoglobulin markers of Th2 and Th1 responses, respectively. We observed that anti-IFN-γ and anti-IL-12 treatment had the most prominent effect in inducing a Th2-associated IgG1 antibody response, as efficient as in WT mice, although all A20^myel-KO mice retain increased levels of IgG2c ( Figures 5F, G ). Histological analysis confirms that inflammation is reduced in the colon of A20^myel-KO antibody-treated groups ( Figure 5H ). In conclusion, we show that A20^myel-KO macrophages overproduce type-1 polarizing cytokines upon recognition of T. muris antigens and that neutralization of these specific cytokines in vivo can overcome Th1 dominance and improve worm clearance in A20^myel-KO mice.

Given the strong Th1 immune profile of A20^myel-KO mice, we finally evaluated the sensitivity of A20^myel-KO mice and WT littermate controls to infection with a virulent strain of Salmonella enterica serovar typhimurium, as a model for type-1-dependent bacterial infection. WT and A20^myel-KO mice were infected by oral gavage with 10⁷ CFU/mouse S. typhimurium for 8 days, and disease progression was assessed. A20^myel-KO mice show increased resistance against S. typhimurium infection compared with WT littermates, as they show increased survival ( Figure 6A ). Moreover, A20^myel-KO mice are able to restrict bacterial systemic dissemination and inflammation, as confirmed by significantly lower numbers of viable S. typhimurium derived from liver and spleen tissue ( Figure 6B ) and by significantly decreased levels of TNF in the serum ( Figure 6C ). Histological analysis of spleen tissue of WT and A20^myel-KO mice showed a decrease of the lymphoid tissue (white pulp) over red pulp in WT compared with A20^myel-KO spleens ( Figure 6D ). Moreover, the liver of infected WT mice shows increased immune cell infiltration and necrotic areas compared to the liver of A20^myel-KO ( Figure 6E ). Although we do not observe differences in body weight and temperature between the two genotypes ( Supplementary Figure 6C ), A20^myel-KO mice are able to clear the Salmonella infection more effectively compared to WT littermates, which can be attributed to their enhanced type-1 immunity of A20^myel-KO mice.

A20^myel-KO mice, with a myeloid-cell-specific deletion (lysozyme M-Cre) of the A20/Tnfaip3 gene, were previously described (13). Mice were housed in individually ventilated cages at the specific pathogen-free animal facility of the Medical Research Building II at Ghent University Hospital. For our experiments, both males and females were used, evenly distributed over the experimental groups, as we observe no difference in response to T. muris and S. typhimurium between both sexes. All mice experiments were conducted according to the national Belgian Law (14/08/1986 and 22/12/2003, Belgian Royal Decree 06/04/2010) and European (EU Directives 2010/63/EU, 86/609/EEG) animal regulations.

Six- to eight-week-old RaG2^−/− mice were infected with 200 Trichuris eggs by oral gavage. After 32–35 days, the mice were sacrificed, and the cecum and proximal colon were isolated. The cecum and colon were cut open to expose the lumen, and the worms were placed in a 10-cm Petri dish containing PBS and, with the help of forceps, were shaken gently to remove feces. The cecum was placed into another Petri dish containing Dulbecco’s modified Eagle medium (DMEM) with 500 U/ml of penicillin and 500 μg/ml of streptomycin (P/S), and by using smooth curved forceps, the worms were gently pulled out and placed in a well of a 6-well plate containing 5 ml of DMEM + P/S. This process was repeated for all the infected mice. The plate was placed in a plastic container with a damp paper towel to maintain humidity and incubated at 37°C for 4 h. After 4 h, the worms were removed using smooth curved forceps and split into two new wells of the six-well plate containing 5 ml of DMEM + P/S. The content of the original well was harvested and placed in a 15-ml falcon tube and spun at 3,000 rpm, and the was supernatant removed and reserved as a 4-h T. muris excretory and secretory antigen. This supernatant was used for mLN restimulation and BMDM treatment. The egg-containing pellet was resuspended in autoclaved tap water. The 6-well plate was placed back at 37°C overnight. Thereafter, the worms were removed, and the content of the wells was harvested into a falcon tube and spun at 3,000 rpm for 5 min. The supernatant was collected and reserved as a 24 h T. muris excretory and secretory antigen. This supernatant was used for Trichuris Ag-specific antibody isotype ELISAs. The remaining egg-containing pellet was resuspended in autoclaved tap water, and the eggs were embryonated at room temperature prior to storage at 4°C for future use.

Infection with T. muris was performed by oral gavage of 200 embryonated T. muris eggs/mouse to 12-week-old WT and A20^myel-KO mice. After 21 days, the mice were sacrificed by cervical dislocation for further analysis. For the cytokine neutralization experiments, A20^myel-KO mice were infected with T. muris eggs as described earlier and treated with antibodies against IFN-γ, IL-12, and IL-18 (BioXcell, Lebanon, USA,, #BE0237) (100 μg). The antibodies were administered i.p. every second day for 21 days in total [IFN-γ (1 mg), IL-12 (1 mg), IL-18 (100 μg)]. Antibodies against IFN-γ and IL-12 were kindly provided by Dr. Louis Boon.

For infection, we used Salmonella typhimurium, virulent strain SL1344. Culture: Salmonella from glycerol stock was grown in LB for 1 h at 37°C. Next, the liquid culture was plated and grown overnight (ON) on blood agar plates at 37°C to acquire single colonies. The day before infection (D-1), a) a colony of Salmonella from the plate was inoculated in 5 ml of LB and grown ON at 37°C in a shaking incubator at 180 rpm and b) the mice were treated with streptomycin to improve colonization efficiency (25 mg/mouse). On the day of infection (D0), food and water were removed from the cage for 4 h. The ON liquid culture was diluted 100× and 500× in LB and grown for 3 h at 37°C. OD600 was measured to determine the concentration and make the appropriate dilutions so that mice were infected with 10⁷ CFU Salmonella/200 μl. Water and food were returned to the mice and the animals were monitored for 8 days for body weight change, temperature, and condition. In the end, the mice were sacrificed, and the organs were processed for further analysis. To evaluate bacterial recovery in the spleen, colon, and liver, we used supplemented MacConkey plates [MacConkey agar (CM0007B; Oxoid, Hampshire, United Kingdom), 6.8% sodium thiosulfate (Sigma-Aldrich), 0.8% ferric ammonium citrate (Sigma-Aldrich)].

Isolation of colonic lamina propria of WT and A20^myel-KO mice was performed as described previously (44). The isolated cells were used for extracellular and intracellular staining for representative markers of T and myeloid cells, in two separated panels (1 × 10⁶ cells per sample) and then analyzed by flow cytometry using a five-laser BD LSRFortessa analyzer. For the analysis of the T-cell compartment, the cells were stained with 7-AAD (1/1,000) for live/dead separation and then extracellularly with the following antibodies: anti-CD3 APC (eBioscience, San Diego, USA; 17-0031-83; 1/100), anti-CD4 APC-Cy7 (BD, New Jersey, USA; 552051; 1/200), and anti-CD8 V500 (BD; 560776; 1/100). The cells were then fixed/permeabilized using the Foxp3 Transcription Factor Staining Buffer Set (eBioscience; 00-5523-00). For the intracellular staining, the following antibodies were used: anti-Foxp3 Alexa Fluor 488 (eBioscience; 53-5773-82; 1/100), anti-T-bet PE-Cy7 (eBioscience; 25-5825-82; 1/100), anti-RORγt BV421 (BD; 562894; 1/100), and anti-GATA3 PE (eBioscience; 12-9966-42; 1/100). For the analysis of the myeloid compartment, the cells were stained with Fixable Viability Dye eFluor 506 (eBioscience; 65-0866-14; 1/300) for live/dead separation and then only extracellularly with the following antibodies: CD19 (eBioscience; 15-0193-82; 1/400), CD3 (eBioscience; 15-0031-82; 1/200), NK1.1 (BioLegend; 108716; 1/200)/PE-Cy5, anti-CD45 Alexa Fluor 700 (eBioscience; 56-0451-82; 1/800), anti-Ly6G PercpCy5.5 (BD; 560602; 1/200), anti-Ly6C APC (eBioscience; 17-5932-80; 1/200), or anti-GR-1 where mentioned (eBioscience; 45-5931-80; 1/200), anti-Siglec-F BUV395 (BD; 740280; 1/200), anti-CD11b BV605 (BD; 563015; 1/600), anti-CD64 BV711 (BioLegend; 139311; 1/100), anti-F4/80 Biotin (eBioscience; 13-4801-82; 1/100), Streptavidin BV421 (BioLegend; 405226; 1/1,000), anti-CD11c PE-eFluor 610 (eBioscience; 61-0114-82; 1/300), and anti-MHC Class II APC-eFluor 780 (eBioscience; 47-5321-80; 1/800). The flow cytometry data were analyzed using FlowJo software and a sequential gating strategy. The graphs showing the percentages of myeloid cells represent the proportions of the parent population. In the tSNE plots for T cells, 10,000 CD3⁺ cells/genotype were exported, and in total, 20,000 cells were analyzed. In the tSNE plots for myeloid cells, 20,000 CD45⁺ cells (lineage excluded: CD19⁺, Nk1.1⁺, and CD3⁺ fraction)/genotype were exported, and in total, 40,000 cells were analyzed. The samples were then concatenated and analyzed with the tSNE plugin of FlowJo (10.6.1) (iterations = 1,000; perplexity = 30). Following dimensional reduction, coordinates for each tSNE dimension (i.e., tSNE1 and tSNE2) in the two-dimensional plots were determined and integrated as novel parameters. The gating strategies for myeloid and T-cell analysis are included in Supplementary Figures 2H and 3 .

For the sorting of CD11b⁺ cells from WT and A20^myel-KO mice, the cells were stained with anti-CD11b BV605 (BD; 563015; 1/600) and CD19 (eBioscience; 15-0193-82; 1/400), CD3 (eBioscience; 15-0031-82; 1/200), and NK1.1 (BioLegend; 108716; 1/200)/PE-Cy5 as lineage markers ( Supplementary Figures 6A, B ).

The mesenteric lymph nodes of WT and A20^myel-KO mice were harvested post-mortem. The tissue was dissociated with a plunger through a 70-μm filter. mLN cells were resuspended in RPMI 1640 medium supplemented with fetal calf serum and penicillin/streptomycin and treated with 3 μg/ml of a-CD3 and 5 μg/ml of a-CD28, both soluble for 48 h. After culture, the cells were processed for RNA isolation, cDNA synthesis, and quantitative PCR, and the supernatants were used for cytokine measurement.

Bone marrow was isolated from the tibia and fibula bones of WT and A20^myel-KO mice. The samples were treated with ACK for 3 min at room temperature to remove red blood cells. The cells were then cultured in Petri dishes in 10 ml of RPMI 1640 medium containing 10% fetal calf serum, penicillin/streptomycin, and 40 ng/ml of m-CSF for a total of 7 days. On day 3 (D3), the cells were supplemented with 1 ml of 400 ng/ml m-CSF. On D5, the medium was replaced with a fresh solution containing 40 ng/ml of m-CSF. On D7, the differentiated cells were replated and cultured with 4 h T. muris antigen for different time points, i.e., 30 min, 1 h, 2 h, 4 h, 6 h, 24 h, 48 h, and 72 h. At the end of the treatment, the cells were processed for RNA isolation, cDNA synthesis, and quantitative PCR, and the supernatants were used for cytokine measurement.

Tissues and cells were lysed using RLT and β-mercaptoethanol and TissueLyser II (Qiagen, Tegelen, Netherlands, for tissues). Total RNA was isolated using the RNeasy Mini Kit (Qiagen), according to the manufacturer’s instructions. The synthesis of cDNA was performed using QuantiTect^® Reverse Transcription Kit (Qiagen). For qPCR, SensiFAST SYBR No-ROX (Bioline, London, United Kingdom) and specific primers for GATA3 (Fw GGCAGAAAGCAAAATGTTTGCT; Rev TGAGTCTGAATGGCTTATTCACAAAT), IL-4 (Fw TGGACTCATTCATGGTGCAG; Rev AACATGGGAAAACTCCATGC), IFN-γ (Fw TCTGGAGGAACTGGCAAAAG; Rev GCTGATGGCCTGATTGTCTT), and IL-12 (Fw CCTGGGTGAGCCGACAGAAGC; Rev CCACTCCTGGAACCTAAGCAC) were used on LightCycler 480 (Roche, Basel, Switzerland). The reactions were performed in triplicates, and the results were analyzed with qbase+ software. As housekeeping genes, GAPDH, Actb, and Ubc were used.

The plates for the ELISA were coated overnight at 4°C with diluted Trichuris antigen in 0.05 M of carbonate bicarbonate buffer. After washing 5× with PBST 0.05%, the plate was incubated with blocking buffer (3% BSA+PBST 0.05%). The diluted serum was then added to the plate and incubated for 90 min at room temperature (RT). Later, the plate was washed 5× with PBST 0.05% and incubated with biotinylated antibody G1 (Serotec MCA336B) or G2A (BD Bioscience 553388) for 60 min at RT. After five washes with PBST 0.05%, the plate was incubated with Streptavidin-POD conjugate (Roche, 11089153001) for 1 h at RT. Next, the peroxidase substrate ABTS (2,2′-azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt, Sigma-Aldrich, A1888) was added, and the absorbance was determined at 405 nm, with reference to 490 nm.

Cytokine levels were determined in mouse blood serum and in supernatants of ex-vivo cultured mesenteric lymph nodes and bone marrow-derived macrophages. Mouse blood was collected post-mortem by cardiac puncture and stored in clot activator tubes containing serum separating silica particles.

IL-4 (171G5005M), IL-5 (171G5006M), and IL-13 (171G5012M) were quantified using the Bio-Plex Pro kit (Bio-Rad, California, United States) on the Bio-Plex 200 system (Bio-Rad) according to the manufacturer’s instructions. IFN-γ, IL-12, and IL-18 were quantified using ELISA (R&D Systems, Minneapolis, United States).

Mouse colon and cecum tissues were fixed in formalin (10%; neutral buffered; Sigma-Aldrich, Missouri, United States). Samples were dehydrated, embedded in paraffin, sectioned at 5 μm, stained with hematoxylin and eosin, and examined by light microscopy. For combined Alcian Blue and periodic acid Schiff staining, dewaxed sections were dehydrated and incubated in Alcian Blue for 20 min. The sections were subsequently washed with PBS and treated with 1% periodic acid for 15 min followed by incubation in Schiff’s reagent for 15 min. Sections were counterstained with Mayer’s hematoxylin for 30 s, washed, and dehydrated before mounting with DePeX.

The following samples were prepared for MACSima imaging: colon—WT colon steady state, A20^myel-KO colon steady state, WT colon T. muris, A20^myel-KO colon T. muris. The tissues were mounted in OCT compound, and sections of 8 μm thickness were prepared with a cryostat. The sections were later fixed with acetone at −20°C for 10 min and washed 3× with PBS, and staining proceeded. Thirty-two markers in total were assessed with the MACSima technology which were coupled to PE or FITC fluorophores. According to the basic principle of the technology, validated antibodies by Miltenyi enable the identification of multiple markers at the same time. The MACSima imaging system acquires fluorescence images of two markers/cycle of the desired predefined ROIs. To complete the cycle, the fluorescence signal is erased, and the process restarts automatically. The images were later analyzed with MACSiQView software.

Sorted CD11b⁺ cells from the colon lamina propria and from the (ankle and knee) joint synovium were analyzed with bulk RNA sequencing using a QuantSeq mRNA Library Prep on 37 samples and sequencing on the NextSeq500 SR-75 by NXTGNT (Faculty of Pharmaceutical Sciences, Ghent University). The quality of the raw data was checked using FastQC (version 0.11.9) and MultiQC (version 1.12). The trimmed data (Trimmomatic version 0.39) were mapped to the mouse reference genome (Assembly: GRCm39, GCA_000001635.9) using STAR (version 2.7.10a). BAM files were created using Samtools (version 1.9) and counted with featureCounts from the Subread package (version 2.0.3).

The quality of the experimental design was assessed using PCA and sample-to-sample distance to determine outlier samples. DESeq2 (version 1.38.2) was used to perform differential expression analysis. P-values were adjusted for multiple testing by DESeq2 using Benjamini–Hochberg correction. Genes were labeled as differentially expressed when the adjusted P-value <0.05. Functional enrichment was performed using the fgsea (v1.24.0) package to find enriched pathways. A pre-ranked gene list was generated by ordering the shrunk differential expression results according to log2FC. Enrichment results were filtered using a normalized enrichment score (NES) cutoff >1.

Gene scoring was performed by comparing the expression data of our dataset with published expression data for in-vitro M1 classically activated (IFN-γ+LPS) macrophages and M2 alternatively activated (IL-4) macrophages described by Orecchioni M. et al. (21) This gene list consists of 710 genes identified to be upregulated in classically activate macrophages and 806 genes to be upregulated in alternatively activated macrophages. The RNA sequencing datasets are available at public repositories ( Supplementary Data ).

Data were analyzed using GraphPad Prism 9.4.1 software. The results are expressed as means ± s.e.m. We used the unpaired Student’s t-test for two-parameter comparisons or one-way ANOVA for multiple comparisons. Differences with P <0.05 were considered statistically significant and denoted as *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.