Uhrf1-Mediated Tnf-α Methylation Controls Proinflammatory Macrophages in Experimental Colitis Resembling Inflammatory Bowel Disease This information is current as of September 30, 2021. Shanshan Qi, Yongkui Li, Zheng Dai, Mengxi Xiang, Guobin Wang, Lin Wang and Zheng Wang J Immunol published online 14 October 2019 http://www.jimmunol.org/content/early/2019/10/12/jimmun ol.1900467 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2019/10/12/jimmunol.190046

Material 7.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 30, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2019 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published October 14, 2019, doi:10.4049/jimmunol.1900467 The Journal of Immunology

Uhrf1-Mediated Tnf-a Gene Methylation Controls Proinflammatory Macrophages in Experimental Colitis Resembling Inflammatory Bowel Disease

Shanshan Qi,*,1 Yongkui Li,*,1 Zheng Dai,* Mengxi Xiang,* Guobin Wang,*,† Lin Wang,*,‡ and Zheng Wang*,†

Macrophages drive the pathological process of inflammatory bowel diseases (IBD) mostly by secreting proinflammatory cytokines, such as Tnf-a. Recent studies have indicated the association between epigenetic modifications and macrophage functions. How- ever, epigenetic mechanisms regulating macrophages’ functional involvement in IBD remain unknown. In this study, we inves- tigated whether the epigenetic regulator Uhrf1 plays a role in innate immunity by functionally regulating macrophages in fl/fl intestines. We employed two transgenic strains of mice (one with Uhrf1 deficiency in macrophages [Uhrf1 Lyz2-Cre mice] Downloaded from and the other with the two mutations at Uhrf1’s DNA methylation regulatory site [Uhrf1YP187/188AA mice]) to assess their susceptibility to dextran sodium sulfate–induced colitis. We examined the cytokines derived from Uhrf1fl/flLyz2-Cre and Uhrf1YP187/188AA macrophages in response to LPS stimulation. We also analyzed the effects of proinflammatory cytokines on Uhrf1 expression in macrophages. The data demonstrated that Uhrf1 deficiency and Uhrf1YP187/188AA mutation resulted in severe colitis in the dextran sodium sulfate–treated mice. In vitro analysis revealed the hypomethylation of Tnf-a promoter and the fl/fl YP187/188AA increased Tnf-a expression in Uhrf1 Lyz2-Cre and Uhrf1 macrophages in response to LPS stimulation, and anti–Tnf- http://www.jimmunol.org/ a therapy implied the key role of Tnf-a to the aggravated colitis in Uhrf1-deficient mice. Exogenous Tnf-a destabilized Uhrf1 through ubiquitination-mediated protein degradation, triggering macrophage activation. In conclusion, we identified that Uhrf1-mediated DNA methylation controls Tnf-a expression of macrophages in the experimental colitis resembling IBD. The epigenetic mechanisms that activate macrophages may provide new therapeutic targets for IBD treatment. The Journal of Immunology, 2019, 203: 000–000.

nflammatory bowel disease (IBD) is a gastrointestinal in- to dextran sodium sulfate (DSS) treatment (11). These observa- flammation disorder characterized by abdominal pain, tions reveal a promoting role of macrophages in IBD. I diarrhea, and bloody diarrhea (1, 2). Although the exact Macrophages acquire and maintain different phenotypes by guest on September 30, 2021 mechanisms of IBD are not fully understood, breakdown of in- depending on transcriptional and posttranscriptional regulations flammatory homeostasis is reportedly critical for IBD pathogen- (12, 13). Epigenetic modifications are thought to be involved in esis (3–5). Macrophages, a critical member of innate immunity, macrophage activation and differentiation. For instance, DNA are abundant in normal intestines (6, 7). In active IBD, the in- methyltransferase Dnmt1, an epigenetic regulator responsible creased number of macrophages infiltrate into lamia propria (8) for epigenetically silencing target via methylating their and produce cytokines, such as Tnf-a (9). Macrophage depletion promoter DNA (14, 15), can hypermethylate PPAR-g pro- reportedly represses the development of spontaneous colitis in moter, consequently inducing the production of proinflammatory Il102/2 mice (10). Furthermore, M-CSF–deficient mice that lack cytokines (Tnf-a and Il-1b) (16). Administration with DNA mature macrophages develop attenuated acute colitis in response methylation inhibitor 5-Aza-2-deoxycytidine or Dnmt1 deletion

*Research Center for Tissue Engineering and Regenerative Medicine, Union and the Academic Medical Doctor Supporting Program (2018) of Tongji Medical Hospital, Tongji Medical College, Huazhong University of Science and Technol- College, Huazhong University of Science and Technology. ogy, Wuhan 430022, China; †Department of Gastrointestinal Surgery, Union Hos- L.W., Z.W., and G.W. designed the experiments and wrote the manuscript. S.Q. and pital, Tongji Medical College, Huazhong University of Science and Technology, Y.L. performed the experiments and analyzed the data. Z.D. and M.X. generated the Wuhan 430022, China; and ‡Department of Clinical Laboratory, Union Hospital, Uhrf1fl/flLyz2-Cre mice. Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430022, China The sequences presented in this article have been submitted to the Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE136311) under 1S.Q. and Y.L. contributed equally to this work. accession number GSE136311. ORCID: 0000-0002-9330-0728 (Z.W.). Address correspondence and reprint requests to Dr. Guobin Wang, Dr. Lin Wang, and Received for publication April 25, 2019. Accepted for publication September 19, Dr. Zheng Wang, Research Center for Tissue Engineering and Regenerative Medicine, 2019. Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, 1277 Jiefang Avenue, Wuhan 430022, China. E-mail addresses: [email protected] (G.W.), This work was supported by the Major State Basic Research Development Program [email protected] (L.W.), and [email protected] (Z.W.) of China (973 Program, 2015CB554007), National Natural Science Foundation of China programs (81272559, 81572866, 81502572, 81773263, and 81773104), the The online version of this article contains supplemental material. International Science and Technology Cooperation Program of Chinese Ministry Abbreviations used in this article: BMDM, bone marrow–derived macrophage; ChIP, of Science and Technology (2014DFA32920), the Science and Technology Program chromatin immunoprecipitation; CKO, conditional knockout; DSS, dextran sodium of Chinese Ministry of Education (113044A), the Frontier Exploration Program of sulfate; IBD, inflammatory bowel disease; IEC, intestinal epithelial cell; qPCR, quan- Huazhong University of Science and Technology (2015TS153), the Natural Science titative real-time PCR; RIPA, radioimmunoprecipitation assay; Uhrf1, -like Foundation Program of Hubei Province (2015CFA049), the Research Fund of Public with PHD and RING finger domains 1; WT, wild-type. Welfare in Health Industry (201402015) from the Health and Family Plan Committee of China, the Integrated Innovative Team for Major Human Diseases Program (2017), Copyright Ó 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1900467 2 Uhrf1 AND INFLAMMATORY BOWEL DISEASES promotes M2 macrophage phenotype characteristic of anti- (extensive damage and extension into deeper structure of the bowel wall). inflammation (17, 18). These observations support the idea that For Tnf-neutralizing therapy, the mice were injected i.p. with Tnf-a Ab epigenetic regulators play a key role in regulating the functions (no. 506332; 20 mg/mouse/d; BioLegend) on days 3–7 of the DSS-induced colitis experiment. of macrophages. However, epigenetic regulations on inflamma- tory macrophages during IBD are poorly understood. Bone marrow–derived macrophage Genome-wide association studies implicated the association of The protocol of bone marrow–derived macrophage (BMDM) production IBD with ubiquitin-like with PHD and RING finger domains 1 was according to the work (33). Briefly, the 6–8-wk-old mice were sac- (Uhrf1) (19, 20). Uhrf1 is a key epigenetic regulator (also known rificed. Femurs were cut off and flushed with 10 ml 10% FBS (Life as Np95 in mice and ICBP90 in humans), acting as a primary Technologies) DMEM medium (GE Healthcare Life Sciences). The mass was removed by 40-mm filters (no. REF352350; Corning), and RBCs were mediator of DNA methylation (21–23). Uhrf1 binds to H3K9me2/ lysed with ammonium/chloride/potassium buffer (no. C3720; Beyotime 3 (di- and trimethyl K9 histone H3) of the newly assembled nu- Biotechnology). Suspended cells were collected and cultured with 10% cleosomes, which is a prerequisite for Uhrf1 to recruit Dnmt1 onto FBS (Life Technologies) DMEM medium (GE Healthcare Life Sciences) replication forks where Dnmt1 methylates newly replicated DNA containing 25% L929 cell (no. CCL-1; American Type Culture Collection) (24–26). Thus, Uhrf1 and Dnmt1 cooperate to maintain methyl- supernatants to differentiate BMDMs for 7 d. ation status of target genes. Mononuclear cells isolation and flow cytometry A factor central to IBD pathogenesis is Tnf-a (27, 28). Ther- Isolation of mononuclear cells from colonic lamina propria was carried out apeutic strategies against Tnf-a were effective for IBD to some as described (34). Briefly, colons were opened, cleared by PBS, and in- extent in clinical trials (29, 30). Although Tnf-a is reportedly cubated in PBS containing 5 mM EDTA and 1 mM DTT at 37˚C for regulated by several signaling pathways in macrophages (31, 32), 20 min to remove intestinal epithelial cells (IECs). Then, residual colon Downloaded from epigenetic regulation of Tnf-a in IBD was unclear. In this study, tissues were cut into pieces and dissected by shaking at 37˚C for 30 min with digestion solution containing 0.5 mg/ml collagenase D (no. we revealed a negative role for Uhrf1 in regulating Tnf-a ex- 11088858001; Roche), 0.5 mg/ml DNase I (no. AMPD1-1KT; Sigma- pression of inflammatory macrophages in intestinal inflammation. Aldrich), and 5% FBS (Life Technologies) in DMEM medium (GE We demonstrated that macrophages lacking Uhrf1 contributed to Healthcare Life Sciences). Forty-micrometer filters were used to remove the the development of severe DSS-induced colitis resembling IBD mass. Cell suspensions were subjected to Ficoll separation (no. LDS1090; TBDscience) and then mononuclear cells were collected. and anti–Tnf-a treatment could prevent the excessive colitis. http://www.jimmunol.org/ For intracellular Tnf-a staining, cells were incubated with PMA (1 ml/ml; Uhrf1 mediated Dnmt1-dependent methylation of Tnf-a promoter, no. P8139; Sigma-Aldrich) and GolgiStop (0.67 ml/ml; no. 554724; BD contributing to suppression of macrophage activation. Notably, Biosciences). Four hours later, the cells were stained with surface Abs exogenous addition of Tnf-a, however, destabilized Uhrf1 and (F4/80-APC; no. 123115; BioLegend; CD11b-FITC; no. 101206; Bio- induced endogenous Tnf-a production, thus forming a proin- Legend) for 30 min. Then, the cells were fixed and permeabilized with Fixation/Permeabilization Buffer (no. 00-5521; Invitrogen) and stained flammatory feedback loop. Our findings might help to understand with the intracellular Ab (Tnf-a Pacific Blue; no. 506318; BioLegend) for the mechanisms underlying pathological processes of IBD. 30 min. Flow cytometer (BD Biosciences) and FlowJo software (Tree Star, Ashland, OR) were used to analyze the data.

Materials and Methods Quantitative real-time PCR and RNA sequencing by guest on September 30, 2021 Mice Total RNA of colon tissues and BMDMs were extracted with TRIzol (no. All the experiments with mice were performed according to the protocol 108-95-2; Takara Bio). One microgram RNA was used for reverse tran- approved by the Ethics Committee of Union Hospital, Tongji Medical scription with the Reverse Transcription Reagent Kit (no. R021-01; Vazyme College, Huazhong University of Science and Technology on Care and Biotech) following the manufacturer’s instructions. Quantitative real-time Use of Laboratory Animals. Uhrf1fl/fl and Uhrf1YP187/188AA mice were PCR (qPCR) was performed to quantify the gene expression using kindly provided by Dr. J. Wong (Shanghai Key Laboratory of Regu- SYBR Green qPCR Master Mix (no. TE221CB; Vazyme Biotech) on the latory Biology, Shanghai, China). To generate the mice with StepOnePlus Software System (Applied Biosystems). Relative gene ex- 2ΔΔ macrophage-specific Uhrf1 deficiency, Uhrf1fl/fl mice were crossed pression levels were normalized to Hprt using the 2 CT quantifica- with Lyz2-Cre mice (purchased from the Model Animal Research tion method. The primers for qPCR were listed in the Supplemental Center of Nanjing University, Nanjing, China). All the mice were of Table I. RNA from DSS-challenged colon tissues was used for RNA se- C57BL/6 background and housed under specific pathogen–free con- quencing with NEBNext Ultra RNA Library Prep Kit for Illumina system dition in the animal facility of Tongji Medical College Huazhong (New England Biolabs). The RNA sequencing data generated the University of Science and Technology. The primers for identification dataset GSE136311 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? of genotype were as follows: Lyz2-Cre, 59-CCCAGAAATGCCA- acc=GSE136311). Differentially expressed genes with the 2-fold GATTACG-39 (forward) and 59-CTTGGGCTGCCAGAATTTCTC-39 change in expression and p values ,0.05 between wild-type (WT; (reverse); Uhrf1fl/fl,59-ACCACATCACTCTTGATCTGCC-39 (forward) and n =4)andUhrf1fl/flLyz2-Cre (n = 3) were subject to Kyoto Encyclo- 59-GGATGTTAGGTGTGAGCCACCATG-39 (reverse); and Uhrf1YP187/188AA, pedia of Genes and Genomes (KEGG) analysis. 59-TCTGTGGGTACTGATAGTGCTCG-39 (forward) and 59-GTTGCCAG- CCATCTGTTGTT-39 (reverse). Immunohistochemistry Murine colitis model The terminal colon tissues were fixed in 4% formaldehyde for 24 h, em- bedded in paraffin laterally, and sectioned for H&E staining. For immu- The sex- and age-matched mice (6–8 wk old) were challenged with nohistochemistry staining, colonic sections were deparaffinized and boiled drinking water containing 2.5% (wt/vol) DSS (molecular mass 36–50 kDa; in EDTA solution (pH 9) for Ag retrieval. A total of 3% H2O2 was used to no. 60316ES60; Yeasen Biotechnology) for 5 d and then provided with quench the activity of endogenous peroxidase, and the background signals normal water for 2 d. Mice were sacrificed on day 7. Body weight, stool were blocked with 5% BSA (no. D-64683; BioFroxx). The sections were consistency, and rectal bleeding were monitored daily. The clinical score incubated with CD68 primary Ab (no. 28058-1-AP; Proteintech) at 4˚C and histological score were assessed as follows: weight loss, 0 (,5% loss), overnight, followed by alkaline phosphatase-linked secondary Ab (no. AP- 1 (5–10% loss), 2 (10–20% loss), and 3 (.20% loss); stool consistency, 9500; Vector Laboratories) for 1 h, and the color was developed with 0 (normal stool), 1 (pasty stool), and 2 (watery stool); and bleeding, 0 (no diaminobenzidine solution (no. SK-4100; Vector Laboratories). The dou- blood), 2 (hemoccult positive), 4 (visible bleeding). Histological scoring ble immunohistochemistry staining was performed as follows. Briefly, was assessed based on two independent parameters: inflammatory cell after incubation with CD68 Ab (no. 28058-1-AP; Proteintech) and alkaline infiltration and crypt damage. Inflammatory cell infiltration was assessed phosphatase-linked secondary Ab (no. AP-9500; Vector Laboratories), the as follows: 0 (rare inflammatory cells), 1 (increased inflammatory cells), 2 sections were stained with second primary Ab against Uhrf1 (no. SC- (abundant inflammatory cells infiltration into the submucosa), and 3 373750; Santa Cruz Biotechnology) overnight. Then, the sections were (transmural infiltration of inflammatory cells). Crypt damage was assessed incubated with HRP-conjugated secondary Ab (no. MP-7401; Vector as follows: 0 (no damage), 1 (discrete lesions), 2 (focal ulceration), and 3 Laboratories). The color was developed with diaminobenzidine solution The Journal of Immunology 3

(no. SK-4100; Vector Laboratories), and hematoxylin (no. BL702A; Bio- Technology; anti-rabbit; no. 7074; Cell Signaling Technology) for 1 h at Sharp) was used to counterstain. All sections were photographed under a room temperature. Then, the membranes were incubated in ECL-detection microscope (Olympus Dp73). reagents (MilliporeSigma), and the bands were detected with Azure Bio- systems c600 imaging system (P/N 97-0400-02). Immunofluorescence Methylated DNA bisulfite sequencing Immunofluorescence staining of cells planted on the coverslips were per- formed as follows. Cells were fixed by 4% formaldehyde and permeabilized Bisulfite CT conversion of genomic DNA was performed using the EZ by 1% Triton X-100 (no. 0694; BioSharp) diluted in PBS. After washing, DNA Methylation-Gold Kit (no. D5005; Zymo Research) according cells were blocked with 5% BSA (no. D-64683; BioFroxx) and incubated to the instruction books. The amplified PCR products were cloned to with the primary Ab against Uhrf1 (no. SC-373750; Santa Cruz Biotech- the T-Vector pMD19 (no. 3271; Takara Bio), and 10 clones for each nology) overnight at 4˚C. Then, cells were incubated with secondary Ab sample were sequenced. The primers were listed as follows: 59- (no. SA00009-1; Proteintech), and the nucleus was counterstained with TTTTGATGTTTGGGTGTTTTTAATT-39 (forward) and 59-TTCTC- DAPI (no. AR1176; Boster Biological Technology). For immunofluores- CCTCCTAACTAATCCCTTAC-39 (reverse). cence staining of colonic sections, tissues were stained with the Ab against Zo1 (no. SC-33725; Santa Cruz Biotechnology). All sections were pho- Chromatin immunoprecipitation tographed under a microscope (Olympus Dp73). Chromatin immunoprecipitation (ChIP) analysis was carried out as ELISA previously described (35). Briefly, chromatin and protein complexes were incubated with specific Abs (Uhrf1; no. SC-373750; Santa Cruz BMDMs (1 3 106 cells/ml) were planted in 12-well plates and incubated Biotechnology; Dnmt1; no. SC-271729; Santa Cruz Biotechnology; with LPS (100 ng/ml; no. L2630; Sigma-Aldrich) for 4 h. The supernatants mouse lgG; no. SC-2025; Santa Cruz Biotechnology) and pulled were collected, and the concentrations of Tnf-a were quantified by ELISA down by Protein A/G PLUS Agarose (no. SC-2003; Santa Cruz

Kit (no. EMC102a; QuantiCyto) based on the instruction books. For Biotechnology). Precipitated immune complexes were washed, and Downloaded from the colonic samples, the segments of terminal colon tissues were ho- DNA was extracted. Semiquantitative PCR was used to visualize the mogenized with radioimmunoprecipitation assay (RIPA) buffer (no. presence of DNA precipitated by specific Abs. The primers were as P0013C; Beyotime Biotechnology). Protein levels were measured with follows: 59-CCTTGATGCCTGGGTGTCCCCAACT-39 (forward) and the bicinchoninic acid assay kit (no. BCA02; DingGuo Biotech). Tnf-a 59-TTCTCCCTCCTGGCTAGTCCCTTGC-39 (reverse). levels in protein extracts were quantified by ELISA kit (Tnf-a; no. EMC102a; QuantiCyto). In vitro ubiquitination assay

Western-blotting BMDMs planted in 10-cm cell culture dishes were incubated with 100 nM http://www.jimmunol.org/ rTnf-a (no. C600052; Sangon Biotech) or not for 6 h. Then, the cells were BMDMs were harvested and lysed in RIPA buffer (no. P0013C; Beyotime treated with 20 mM MG132 (no. HY-13259; MedChemExpress) for an- Biotechnology) containing mixture (no. HY-K0010; MedChemExpress). other 6 h. After that, cells were harvested and lysed in RIPA buffer (no. An equal amount of protein was loaded in 12% SDS-PAGE to be separated P0013D; Beyotime Biotechnology) containing 2% SDS and 5 mM DTT. and then transferred to the nitrocellulose membranes. After blocking with The lysates were denatured and diluted 10 times with PBS. Then, the ly- 5% BSA, the membranes were incubated with primary Abs (Uhrf1, sates were incubated with primary Ab against Uhrf1 (no. SC-373750; no. SC-373750, Santa Cruz Biotechnology; Gapdh, no. 6004-1-lg, Pro- Santa Cruz Biotechnology), and immune complexes were pulled down teintech; Ub, no. SC-47721, Santa Cruz Biotechnology) at 4˚C overnight by Protein A/G PLUS Agarose (no. SC-2003; Santa Cruz Biotechnology). and following with secondary Abs (anti-mouse; no. 7074; Cell Signaling After washing, precipitated immune complexes were boiled in SDS-PAGE by guest on September 30, 2021

FIGURE 1. Uhrf1fl/flLyz2-Cre mice are more susceptible to DSS-induced acute colitis. (A–D) WT and Uhrf1fl/flLyz2-Cre mice were adminis- tered with 2.5% DSS in drinking water to in- duce acute colitis. Body weight changes (A), clinical scores (B), and colon length (C) were recorded and analyzed. Representative images of H&E staining and histological scoring in terminal colon tissues were shown (D)(n =5 per group). (E) The relative mRNA levels of Zo1, tricellulin, Claudin5, and Claudin10 in colon tissues from WT and Uhrf1fl/flLyz2-Cre mice under steady-state or with DSS-induced colitis were analyzed by qPCR (n = 5 per group). The data represent three independent experiments. Scale bar, 100 mm. Error bars represent mean 6 SEM. *p , 0.05 (unpaired t test). N.S., not significant. 4 Uhrf1 AND INFLAMMATORY BOWEL DISEASES Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 2. Uhrf1 deletion upregulates Tnf-a expression. (A) KEGG analysis of RNA sequencing data showed the top 20 gene set that activated in DSS- treated Uhrf1fl/flLyz2-Cre mice versus WT mice. (B) Heatmap displayed the mRNA expression of cytokines and chemokines in colon tissues from DSS- treated WT (n = 4) and Uhrf1fl/flLyz2-Cre mice (n = 3). (C and D) BMDMs derived from WT and Uhrf1fl/flLyz2-Cre mice were treated with LPS (100 ng/ml) for 4 h. The mRNA expression of inflammation-associated genes were analyzed by qPCR (C). Tnf-a protein levels in culture supernatants were analyzed by ELISA (D). (E and F) Tnf-a expression by CD11b+ F4/80+ macrophages in colonic lamina propria of WT and Uhrf1fl/flLyz2-Cre mice received 2.5% DSS challenge were assessed by flow cytometry (E), and the Tnf-a levels in the colon tissue homogenates were measured by (Figure legend continues) The Journal of Immunology 5

FIGURE 3. Loss of H3K9me2/3-binding activity of Uhrf1 results in increased Tnf-a expression. (A) Bisulfite sequencing for the CpG methylation status in the Tnf-a promoter region (2205 to +50 bp) was analyzed. Ten respective replicates from WT and Uhrf1fl/flLyz2-Cre samples were subjected to the analysis. (B) The occu- pation of Uhrf1 on the Tnf-a promoter in BMDMs were detected by ChIP assays with the Ab against Uhrf1. Isotype lgG served as a control. (C) Bisulfite sequencing for the CpG methylation status in the Tnf-a promoter region (2205 to +50 bp) was analyzed. Ten respective YP187/188AA replicates from WT and Uhrf1 samples were Downloaded from subjected to the analysis. (D and E) BMDMs derived from WT and Uhrf1YP187/188AA mice were stimulated by LPS (100 ng/ml) for 4 h. The relative mRNA levels of Tnf-a were analyzed by qPCR (D), and the Tnf-a pro- tein levels in culture supernatants were measured by ELISA (E). (F and G) Tnf-a expression by CD11b+ + F4/80 macrophages in colonic lamina propria of WT http://www.jimmunol.org/ and Uhrf1YP187/188AA mice received DSS challenge were assessed by flow cytometry (F), and the Tnf-a levels in colon tissue homogenates were analyzed by ELISA (G) (n = 5–6 per group). The data represent three indepen- dent experiments. Error bars represent mean 6 SEM. *p , 0.05 (unpaired t test). by guest on September 30, 2021

Sample Loading Buffer (no. P0015A; Beyotime Biotechnology) for im- clinical scores (Fig. 1B), and shortened colons (Fig. 1C). Con- munoblot analysis. sistently, DSS-treated Uhrf1fl/flLyz2-Cre mice also had much Statistical analysis worsened pathohistological alterations: pronounced epithelial disruption, increased crypt distortion, and massive inflammatory 6 Quantitative data were presented as mean SEM. The statistical signi- cell infiltration of lamina propria in colon tissues in comparison ficance between groups was assessed using one-way ANOVA or two-tailed unpaired Student t test. All the data were analyzed by GraphPad Prism with WT mice (Fig. 1D). As tightly joined IECs are essential for version 5 (GraphPad Software, San Diego, CA). All p values , 0.05 were gut to act as an effective barrier (39), we assessed the expression accepted as statistically significant. of tight junction–associated genes in colon tissues. Zo1, tricellu- lin, Claudin5, and Claudin10 in colon tissues from Uhrf1fl/flLyz2- Results Cre mice were transcriptionally decreased compared with WT colon Uhrf1-deficient macrophages aggravate DSS-induced colitis tissues (Fig. 1E). We next determined whether Uhrf1fl/flLyz2-Cre To study the role of the epigenetic regulator Uhrf1 in macrophages, mice spontaneously developed colitis. Compared with WT mice, we generated the Uhrf1fl/flLyz2-Cre mice where macrophages were the colon tissues from Uhrf1fl/flLyz2-Cre mice had no differences Uhrf1 deficient (Supplemental Fig. 1A). Uhrf1fl/flLyz2-Cre mice in the expression of tight junction–associated genes under steady- and WT mice were challenged with DSS to induce colitis that state (Fig. 1E) and showed no obvious inflammation 6 mo after pathologically resembles IBD, as previously reported (36–38). birth (Supplemental Fig. 1B). The immunofluorescence staining Uhrf1fl/flLyz2-Cre mice developed more severe IBD symptoms revealed that the expression of the tight junction–associated gene than WT mice, such as drastic weight loss (Fig. 1A), elevated Zo1 in the colon tissues of Uhrf1fl/flLyz2-Cre mice was similar to

ELISA (F)(n = 5–6 per group). (G–J) Tnf-a–neutralization therapy of DSS-treated WT and Uhrf1fl/flLyz2-Cre mice. Body weight changes (G) and clinical scores (H) were recorded and analyzed. Representative images of H&E staining and histological scoring in terminal colon tissues were shown (I). The mRNA expression of Zo1, tricellulin, Claudin5, and Claudin10 in colon tissues was analyzed by qPCR (J)(n = 5 per group). The data represent three independent experiments. Scale bar, 100 mm. Error bars represent mean 6 SEM. *p , 0.05 (unpaired t test). N.S., not significant. 6 Uhrf1 AND INFLAMMATORY BOWEL DISEASES Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 4. Uhrf1YP187/188AA mutant mice are sensitive to DSS-induced colitis. (A–D) WT and Uhrf1YP187/188AA mice were administered with 2.5% DSS in drinking water to induce acute colitis. Body weight changes (A), colon length (B), and clinical scores (C) were recorded and analyzed. Representative images of H&E staining and histological scoring in terminal colon tissues were shown (D)(n = 5–6 per group). (E) The relative mRNA levels of Zo1, tricellulin, Claudin5, and Claudin10 in colon tissues from WT and Uhrf1YP187/188AA mice under steady-state or with DSS-induced colitis were analyzed by qPCR (n = 5 per group). The data represent three independent experiments. Scale bar, 100 mm. Error bars represent mean 6 SEM. *p , 0.05 (unpaired t test). N.S., not significant. that in WT mice (Supplemental Fig. 1C). These data suggest that levels of several key inflammation-associated genes, with Tnf-a Uhrf1 conditional knockout (CKO) in macrophages aggravates, being upregulated the most (Fig. 2C). The enhanced Tnf-a but not spontaneously induces, colitis. secretion resulting from Uhrf1 deficiency was also observed (Fig. 2D). Although the abundance of Tnf-a–expressing mac- a Uhrf1 represses Tnf- expression in macrophages rophages in the colonic lamina propria was similar between We next explored whether Uhrf1 deficiency impacted differenti- Uhrf1-deficient mice and WT mice (Supplemental Fig. 1F), ation of macrophages in vitro and in vivo. We isolated the bone Uhrf1 deficiency drastically upregulated Tnf-a level in colonic marrow progenitor cells from Uhrf1fl/flLyz2-Cre and WT mice and macrophages of DSS-treated mice (Fig. 2E). Moreover, DSS- generated matured BMDMs. Uhrf1 deletion did not affect induced production of Tnf-a was increased in colon tissues of BMDMs differentiation in vitro (Supplemental Fig. 1D). The Uhrf1fl/flLyz2-Cre mice in comparison with WT mice (Fig. 2F). number of macrophages in the colonic lamina propria of Uhrf1- These results indicate that Uhrf1 negatively regulates Tnf-a deficient mice were similar to that of WT mice (Supplemental expression. Fig. 1E). To assess the role of Uhrf1 in macrophage functional Given that Uhrf1 repressed Tnf-a expression in macrophages, regulation, we performed RNA sequencing and analyzed the we hypothesized that Tnf-a might play a key role in the exacer- transcriptomes of colon tissues from DSS-treated mice. KEGG bated colitis of DSS-treated Uhrf1fl/flLyz2-Cre mice. To test the analysis and heatmap revealed that colon tissues from Uhrf1 CKO idea, we treated the DSS-challenged mice with Tnf-a Ab via i.p. mice expressed high levels of several proinflammatory cytokines injection to neutralize endogenous Tnf-a. In contrast with DSS- and chemokines relative to WT mice (Fig. 2A, 2B). We also treated mice, the presence of Tnf-a Ab removed the differences in stimulated BMDMs with LPS that is capable of activating mac- body weight loss (Fig. 2G) and clinical scores (Fig. 2H) between rophages and initiating innate immune responses (33, 40, 41) and WT mice and Uhrf1fl/flLyz2-Cre mice. It has been highlighted that examined the expression of proinflammatory cytokines from Tnf-a overproduction could lead to IBD through inducing apo- macrophages. Intriguingly, Uhrf1 deficiency increased mRNA ptosis of IECs (27, 28, 42). We found that the severe IECs damage The Journal of Immunology 7 in Uhrf1fl/flLyz2-Cre mice was relieved by Tnf-a Ab therapy treatment had elevated abundance of macrophages expressing (Fig. 2I). Also, no differences in the mRNA expression of tight Tnf-a and the increased amount of Tnf-a in colon tissues (Fig. 3F, junction–associated genes were found between colon tissues from 3G). These results demonstrate that Uhrf1 epigenetically regulates WT and Uhrf1fl/flLyz2-Cre mice (Fig. 2J). These data suggest that methylation of Tnf-a gene to dampen its expression, which would Tnf-a overexpression and Tnf-a–induced damage of IECs induce restrain macrophages’ proinflammatory functions. the colitis in Uhrf1fl/flLyz2-Cre mice. Loss of H3K9me2/3-binding activity of Uhrf1 enhances Uhrf1 epigenetically regulates methylation status of DSS-induced colitis a Tnf- promoter Given our observation that loss of H3K9me2/3-binding ability of Because Uhrf1 often represses transcription of target genes by Uhrf1 increased Tnf-a expression, we wondered whether Uhrf1’s regulating DNA methylation of target genes’ promoters (25, 26), functional role in colitis phenotypes were dependent on the we hypothesized that Uhrf1 deficiency altered DNA methylation H3K9me2/3-binding activity. Similar to Uhrf1fl/flLyz2-Cre mice, status of Tnf-a promoter. Supporting this, Uhrf1 CKO in macro- Uhrf1 YP187/188AA mice were fertile with no colitis under normal phages drastically reduced methylation of CG dinucleotides at the condition (Supplemental Fig. 2E, 2F). However, after being ad- region from 221 to +18 bp in Tnf-a promoter (Fig. 3A). The ChIP ministered with DSS, Uhrf1 YP187/188AA mice developed severe analysis indicated Uhrf1’s binding to the promoter of Tnf-a colitis, with the symptoms including pronounced body weight (Fig. 3B). Given Uhrf1’s H3K9me2/3-binding activity is required loss (Fig. 4A), shortened colon (Fig. 4B), and higher clinical for Uhrf1 to recruit Dnmt1 to methylate target genes’ promoters, scores (Fig. 4C). Consistently, in comparison with WT mice, YP187/188AA YP187/188AA we employed the Uhrf1 transgenic mice where the two Uhrf1 mice had higher colitis histological scores Downloaded from mutations, respectively, affected tyrosine 187 and proline 188 of (Fig. 4D) and decreased expression of epithelial tight junction Uhrf1 protein (Supplemental Fig. 2A). These two-point muta- genes (Fig. 4E). Thus, loss of H3K9me2/3-binding of Uhrf1 tions abolish Uhrf1’s H3K9me2/3-binding activity (43). Notably, enhances DSS-induced colitis, which is highly likely because of Uhrf1YP187/188AA BMDMs exhibited reduced methylation in Tnf-a Uhrf1’s incapability of maintaining DNA methylation. promoter (221 to +18 bp) (Fig. 3C). Although Uhrf1 mutant did Tnf-a downregulates Uhrf1 protein level in macrophages

not affect the BMDM differentiation (Supplemental Fig. 2B) or http://www.jimmunol.org/ numbers of macrophages present in colon tissues (Supplemental Next, we investigated the regulation of Uhrf1 expression in Fig. 2C, 2D), this mutant Uhrf1 permitted a drastic upregulation of macrophages. Intestine-infiltrating CD68+ macrophages expressed Tnf-a expression in BMDMs (Fig. 3D, 3E) in response to LPS Uhrf1 at a moderate level, whereas its expression in macro- stimulation. Moreover, Uhrf1YP187/188AA mice receiving DSS phages was sharply dropped in DSS-induced colitis (Fig. 5A). by guest on September 30, 2021

FIGURE 5. The stability of Uhrf1 protein is reduced by Tnf-a in macrophages. (A) Immunohistochemical staining of Uhrf1 (brown) and CD68 (red) in colon tissues from WT mice received DSS treatment (n = 6 per group). Scale bar, 20 mm. (B) Immunoblot analysis of Uhrf1 protein level in BMDMs incubated with Il6 (100 nM), Il1-b (100 nM), or rTnf-a (100 nM) for 12 h. (C) DSS-challenged WT mice were treated with Tnf-a Ab. Immunohisto- chemical staining of Uhrf1 (brown) and CD68 (red) in colon tissues was shown (n = 5 per group). (D) Decay assay of Uhrf1 protein in cycloheximide (CHX; 10 mM)–treated BMDMs incubated with or without rTnf-a (100 nM). (E) BMDMs incubated with rTnf-a (100 nM) for 12 h were subjected to anti- Uhrf1 immunoprecipitation, followed by immunoblot analysis with Ab against Ub. (F) The occupation of Uhrf1 and Dnmt1 on the Tnf-a promoter in BMDMs incubated with rTnf-a (100 nM) for 12 h were detected by ChIP assays. Isotype lgG served as a control. (G)Tnf-a mRNA levels in BMDMs incubated with rTnf-a (100nM)wereanalyzedbyqPCR.Thedatarepresent three independent experiments. Scale bar, 20 mm. Error bars represent mean 6 SEM. *p , 0.05 (unpaired t test). 8 Uhrf1 AND INFLAMMATORY BOWEL DISEASES

(from 221 to +18 bp) of Tnf-a promoter was hypomethylated in Uhrf1-deficient macrophages. Uhrf1’s epigenetic regulation on Tnf-a expression via Dnmt1- dependent DNA methylation was further demonstrated using the Uhrf1YP187/188AA-transgenic mice. The proper recruitment of Dnmt1 relies on Uhrf1 first binding to H3K9me2/3 of nucleo- somes of target genes (24–26), where Uhrf1 recruits Dnmt1 via Set and Ring-associated (SRA) domain (52). YP187/188AA mu- tations in Tandem Tudor (TTD) domain abolishes Uhrf1’s H3K9me2/ 3-binding activity, which would reduce Uhrf1’s capability of recruiting Dnmt1 for methylating DNA in target genes’ promoters. Indeed, Uhrf1YP187/188AA BMDMs exhibited a reduction in DNA methylation in Tnf-a promoter, which was followed by enhanced Tnf-a expression. Phenotypically, Uhrf1YP187/188AA mice also showed FIGURE 6. The regulatory role of Uhrf1 in intestinal macrophages. In severe inflammatory responses to DSS treatment. Our observations normal intestine, Uhrf1 suppresses the proinflammatory functions of collectively indicate that Uhrf1 epigenetically regulates Tnf-a ex- macrophages via maintaining DNA methylation of inflammatory cytokine pression. Given that constant intestinal bacteria exposure imposes such as Tnf-a. Uhrf1 protein can be destabilized in feedback by Tnf-a persistent pressure on macrophages to produce Tnf-a (8, 53), Uhrf1 abnormally stimulated by intestinal pathological factors, which further might contribute to dynamically regulate Tnf-a expression toward Downloaded from enhances Tnf-a expression and promotes inflammation progression. maintaining intestinal immune homeostasis. In inflammatory intestines, both immune cells, such as macro- Accompanied by this downregulation, Tnf-a was upregulated in phages (8, 54), and IECs (49) can produce Tnf-a. Although IECs’ the colitis tissue (Supplemental Fig. 3A). Notably, exogenous responses to Tnf-a are important to intestinal disease pathogenesis addition of rTnf-a, but not Il1-b and Il6, downregulated Uhrf1 (28, 55), the impact of Tnf-a on macrophages was poorly un-

protein in BMDMs (Fig. 5B, Supplemental Fig. 3B, 3C). Con- derstood. Our data showed that (1) rTnf-a destabilized Uhrf1 by http://www.jimmunol.org/ sistent with the in vitro results, Tnf-a Ab treatment showed that promoting its ubiquitination in macrophages (2); rTnf-a enhanced neutralization of endogenous Tnf-a in acute colitis mice could Tnf-a expression by reducing Uhrf1 and Dnmt1’s binding to the recover the Uhrf1 expression of macrophages (Fig. 5C). The promoter of Tnf-a in macrophages. Given these observations, it protein decay assay showed that Tnf-a destabilized Uhrf1 protein was possible that once excessive amount of Tnf-a was produced by increased ubiquitination (Fig. 5D, 5E). These observations by macrophages, it could downregulate Uhrf1 in macrophages, suggest a positive feedback loop between Tnf-a–induced Uhrf1 resulting in reduced methylation on Tnf-a gene, permitting degradation and Uhrf1-mediated Tnf-a methylation in macro- overproduction of Tnf-a, and thus forming a positive feedback phages during inflammation. Supporting this notion, exogenous loop that might contribute to IBD pathogenesis. Further study Tnf-a incubation with BMDMs decreased the association of is needed to explore the therapeutic value of blocking this by guest on September 30, 2021 Uhrf1 and Dnmt1 with Tnf-a promoter (Fig. 5F) and induced Uhrf1/Tnf-a epigenetic loop in IBD. endogenous Tnf-a expression (Figs. 5G, 6). Acknowledgments Discussion We thank Dr. Jiemin Wong (Shanghai Key Laboratory of Regulatory Biol- and proliferation promoted by Uhrf1, a key epigenetic ogy, Shanghai, China) for kindly providing Uhrf1fl/fl and Uhrf1 YP187/188AA regulator, via recruiting Dnmt1 to methylate DNA, thus achieving mice. methylation regulation of target genes, were extensively explored (23, 44, 45). However, research efforts directed to dissect Uhrf1’s Disclosures effects on development and functions of macrophages during in- The authors have no financial conflicts of interest. flammatory diseases are limited. In this study, we uncovered a functional role for Uhrf1 in controlling Tnf-a expression from macrophages during IBD pathogenesis (Fig. 6). Although Uhrf1 References 1. Podolsky, D. K. 2002. Inflammatory bowel disease. N. Engl. J. Med. 347: 417– had no effect on macrophage differentiation, Uhrf1-deficient mac- 429. rophages did overexpress Tnf-a, leading to the damage of IECs and 2. Ko, J. K., and K. K. Auyeung. 2014. Inflammatory bowel disease: etiology, promoting the development of severe experimental colitis. pathogenesis and current therapy. Curr. Pharm. Des. 20: 1082–1096. 3. Xavier, R. J., and D. K. Podolsky. 2007. Unravelling the pathogenesis of in- Tnf-a production by macrophages is known to be critical to flammatory bowel disease. Nature 448: 427–434. IBD onset, as its inhibition not only reduces inflammatory 4. Kamada, N., S. U. Seo, G. Y. Chen, and G. Nu´n˜ez. 2013. Role of the gut markers, but also relieves structural damages to mucosa (46–48). microbiota in immunity and inflammatory disease. Nat. Rev. Immunol. 13: 321– 335. The regulatory mechanism of Tnf-a expression in macrophages is, 5. Kaistha, A., and J. Levine. 2014. Inflammatory bowel disease: the classic gas- however, not yet fully understood. In this study, we identified Tnf- trointestinal autoimmune disease. Curr. Probl. Pediatr. Adolesc. Health Care 44: 328–334. a as a target gene of Uhrf1. In zebrafish, loss of DNA methylation 6. Lee, S. H., P. M. Starkey, and S. Gordon. 1985. Quantitative analysis of total function of Uhrf1 led to aberrant production of Tnf-a in IECs, macrophage content in adult mouse tissues. Immunochemical studies with followed by development of severe symptoms of IBD (49). monoclonal antibody F4/80. J. Exp. Med. 161: 475–489. 7. Brandtzaeg, P. 2009. Mucosal immunity: induction, dissemination, and effector In our study, we identified that, in macrophages, Uhrf1 nega- functions. Scand. J. Immunol. 70: 505–515. tively regulated Tnf-a transcription. Mammalian Tnf-a contains a 8. Bain, C. C., C. L. Scott, H. Uronen-Hansson, S. Gudjonsson, O. Jansson, CpG-poor promoter and has no CpG cluster, which is in sharp O. Grip, M. Guilliams, B. Malissen, W. W. Agace, and A. M. Mowat. 2013. Resident and pro-inflammatory macrophages in the colon represent alternative contrast to the CpG-rich promoter in zebrafish. Consistent context-dependent fates of the same Ly6Chi monocyte precursors. Mucosal with the notion that methylation of individual CG sites around Immunol. 6: 498–510. 9. Allison, M. C., and L. W. Poulter. 1991. Changes in phenotypically distinct the transcription start site of Tnf-a may determine the tran- mucosal macrophage populations may be a prerequisite for the development of scriptional status (50, 51), we found that the small region inflammatory bowel disease. Clin. Exp. Immunol. 85: 504–509. The Journal of Immunology 9

10. Watanabe, N., K. Ikuta, K. Okazaki, H. Nakase, Y. Tabata, M. Matsuura, production and epithelial cell apoptosis in acute murine colitis. Immunology H. Tamaki, C. Kawanami, T. Honjo, and T. Chiba. 2003. Elimination of local 118: 112–121. macrophages in intestine prevents chronic colitis in interleukin-10-deficient 32. Munitz, A., E. T. Cole, A. Beichler, K. Groschwitz, R. Ahrens, K. Steinbrecher, mice. Dig. Dis. Sci. 48: 408–414. T. Willson, X. Han, L. Denson, M. E. Rothenberg, and S. P. Hogan. 2010. Paired 11. Ghia, J. E., F. Galeazzi, D. C. Ford, C. M. Hogaboam, B. A. Vallance, and immunoglobulin-like receptor B (PIR-B) negatively regulates macrophage ac- S. Collins. 2008. Role of M-CSF-dependent macrophages in colitis is driven by tivation in experimental colitis. Gastroenterology 139: 530–541. the nature of the inflammatory stimulus. Am. J. Physiol. Gastrointest. Liver 33. Ying, W., P. S. Cheruku, F. W. Bazer, S. H. Safe, and B. Zhou. 2013. Investi- Physiol. 294: G770–G777. gation of macrophage polarization using bone marrow derived macrophages. 12. Dai, M., L. Wu, Z. He, S. Zhang, C. Chen, X. Xu, P. Wang, A. Gruzdev, J. Vis. Exp. 76: e50323. D. C. Zeldin, and D. W. Wang. 2015. Epoxyeicosatrienoic acids regulate mac- 34. Weigmann, B., I. Tubbe, D. Seidel, A. Nicolaev, C. Becker, and M. F. Neurath. rophage polarization and prevent LPS-induced cardiac dysfunction. J. Cell. 2007. Isolation and subsequent analysis of murine lamina propria mononuclear Physiol. 230: 2108–2119. cells from colonic tissue. Nat. Protoc. 2: 2307–2311. 13. Takeuch, O., and S. Akira. 2011. Epigenetic control of macrophage polarization. 35. Cui, T. X., I. Kryczek, L. Zhao, E. Zhao, R. Kuick, M. H. Roh, L. Vatan, Eur. J. Immunol. 41: 2490–2493. W. Szeliga, Y. Mao, D. G. Thomas, et al. 2013. Myeloid-derived suppressor cells 14. Kittan, N. A., R. M. Allen, A. Dhaliwal, K. A. Cavassani, M. Schaller, enhance stemness of cancer cells by inducing microRNA101 and suppressing the K. A. Gallagher, W. F. Carson, IV, S. Mukherjee, J. Grembecka, T. Cierpicki, corepressor CtBP2. Immunity 39: 611–621. et al. 2013. Cytokine induced phenotypic and epigenetic signatures are key to 36. Cooper, H. S., S. N. Murthy, R. S. Shah, and D. J. Sedergran. 1993. Clinico- establishing specific macrophage phenotypes. PLoS One 8: e78045. pathologic study of dextran sulfate sodium experimental murine colitis. Lab. 15. Cheng, C., C. Huang, T. T. Ma, E. B. Bian, Y. He, L. Zhang, and J. Li. Invest. 69: 238–249. 2014. SOCS1 hypermethylation mediated by DNMT1 is associated with 37. Okayasu, I., S. Hatakeyama, M. Yamada, T. Ohkusa, Y. Inagaki, and R. Nakaya. lipopolysaccharide-induced inflammatory cytokines in macrophages. Toxicol. Lett. 1990. A novel method in the induction of reliable experimental acute and 225: 488–497. chronic ulcerative colitis in mice. Gastroenterology 98: 694–702. 16. Yu, J., Y. Qiu, J. Yang, S. Bian, G. Chen, M. Deng, H. Kang, and L. Huang. 2016. 38. Kiesler, P., I. J. Fuss, and W. Strober. 2015. Experimental models of inflam- DNMT1-PPARg pathway in macrophages regulates chronic inflammation and matory bowel diseases. Cell. Mol. Gastroenterol. Hepatol. 1: 154–170. atherosclerosis development in mice. Sci. Rep. 6: 30053. 39. Turner, J. R. 2009. Intestinal mucosal barrier function in health and disease. Nat. 17. Wang, X., Q. Cao, L. Yu, H. Shi, B. Xue, and H. Shi. 2016. Epigenetic regulation Rev. Immunol. 9: 799–809. Downloaded from of macrophage polarization and inflammation by DNA methylation in obesity. 40. Doyle, S. L., and L. A. O’Neill. 2006. Toll-like receptors: from the discovery of JCI Insight 1: e87748. NFkappaB to new insights into transcriptional regulations in innate immunity. 18. Thangavel, J., S. Samanta, S. Rajasingh, B. Barani, Y. T. Xuan, B. Dawn, and Biochem. Pharmacol. 72: 1102–1113. J. Rajasingh. 2015. Epigenetic modifiers reduce inflammation and modulate 41. Zhang, G., and S. Ghosh. 2001. Toll-like receptor-mediated NF-kappaB ac- macrophage phenotype during endotoxemia-induced acute lung injury. J. Cell tivation: a phylogenetically conserved paradigm in innate immunity. J. Clin. Sci. 128: 3094–3105. Invest. 107: 13–19. 19. Franke, A., D. P. McGovern, J. C. Barrett, K. Wang, G. L. Radford-Smith, 42. Zeissig, S., C. Bojarski, N. Buergel, J. Mankertz, M. Zeitz, M. Fromm, and

T. Ahmad, C. W. Lees, T. Balschun, J. Lee, R. Roberts, et al. 2010. Genome- J. D. Schulzke. 2004. Downregulation of epithelial apoptosis and barrier repair in http://www.jimmunol.org/ wide meta-analysis increases to 71 the number of confirmed Crohn’s disease active Crohn’s disease by tumour necrosis factor alpha antibody treatment. susceptibility loci. Nat. Genet. 42: 1118–1125. [Published erratum appears in 2004 Gut 53: 1722.] Gut 53: 1295–1302. 20. Jostins, L., S. Ripke, R. K. Weersma, R. H. Duerr, D. P. McGovern, K. Y. Hui, 43. Zhao, Q., J. Zhang, R. Chen, L. Wang, B. Li, H. Cheng, X. Duan, H. Zhu, W. Wei, J. C. Lee, L. P. Schumm, Y. Sharma, C. A. Anderson, et al; International IBD J. Li, et al. 2016. Dissecting the precise role of H3K9 methylation in crosstalk with Genetics Consortium (IIBDGC). 2012. Host-microbe interactions have shaped DNA maintenance methylation in mammals. Nat. Commun. 7: 12464. the genetic architecture of inflammatory bowel disease. Nature 491: 119–124. 44. Obata, Y., Y. Furusawa, T. A. Endo, J. Sharif, D. Takahashi, K. Atarashi, 21. Bostick, M., J. K. Kim, P. O. Este`ve, A. Clark, S. Pradhan, and S. E. Jacobsen. M. Nakayama, S. Onawa, Y. Fujimura, M. Takahashi, et al. 2014. The epigenetic 2007. UHRF1 plays a role in maintaining DNA methylation in mammalian cells. regulator Uhrf1 facilitates the proliferation and maturation of colonic regulatory Science 317: 1760–1764. T cells. Nat. Immunol. 15: 571–579. 22. Sharif, J., M. Muto, S. Takebayashi, I. Suetake, A. Iwamatsu, T. A. Endo, 45. Tien, A. L., S. Senbanerjee, A. Kulkarni, R. Mudbhary, B. Goudreau, J. Shinga, Y. Mizutani-Koseki, T. Toyoda, K. Okamura, et al. 2007. The SRA S. Ganesan, K. C. Sadler, and C. Ukomadu. 2011. UHRF1 depletion causes a

protein Np95 mediates epigenetic inheritance by recruiting Dnmt1 to methylated G2/M arrest, activation of DNA damage response and apoptosis. Biochem. by guest on September 30, 2021 DNA. Nature 450: 908–912. J. 435: 175–185. 23. Nishiyama, A., L. Yamaguchi, J. Sharif, Y. Johmura, T. Kawamura, K. Nakanishi, 46. Targan, S. R., S. B. Hanauer, S. J. van Deventer, L. Mayer, D. H. Present, S. Shimamura, K. Arita, T. Kodama, F. Ishikawa, et al. 2013. Uhrf1-dependent T. Braakman, K. L. DeWoody, T. F. Schaible, and P. J. Rutgeerts. 1997. A short- H3K23 ubiquitylation couples maintenance DNA methylation and replication. term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha Nature 502: 249–253. for Crohn’s disease. Crohn’s disease cA2 study group. N. Engl. J. Med. 337: 24. Rothbart, S. B., K. Krajewski, N. Nady, W. Tempel, S. Xue, A. I. Badeaux, 1029–1035. D. Barsyte-Lovejoy, J. Y. Martinez, M. T. Bedford, S. M. Fuchs, et al. 2012. 47. Kriegel, C., and M. Amiji. 2011. Oral TNF-a gene silencing using a polymeric Association of UHRF1 with methylated H3K9 directs the maintenance of DNA microsphere-based delivery system for the treatment of inflammatory bowel methylation. Nat. Struct. Mol. Biol. 19: 1155–1160. disease. J. Control. Release 150: 77–86. 25. Rothbart, S. B., B. M. Dickson, M. S. Ong, K. Krajewski, S. Houliston, 48. Danese, S., L. Vuitton, and L. Peyrin-Biroulet. 2015. Biologic agents for IBD: D. B. Kireev, C. H. Arrowsmith, and B. D. Strahl. 2013. Multivalent histone practical insights. Nat. Rev. Gastroenterol. Hepatol. 12: 537–545. engagement by the linked tandem Tudor and PHD domains of UHRF1 is re- 49. Marjoram, L., A. Alvers, M. E. Deerhake, J. Bagwell, J. Mankiewicz, quired for the epigenetic inheritance of DNA methylation. Genes Dev. 27: 1288– J. L. Cocchiaro, R. W. Beerman, J. Willer, K. D. Sumigray, N. Katsanis, et al. 1298. 2015. Epigenetic control of intestinal barrier function and inflammation in 26. Liu, X., Q. Gao, P. Li, Q. Zhao, J. Zhang, J. Li, H. Koseki, and J. Wong. 2013. zebrafish. Proc. Natl. Acad. Sci. USA 112: 2770–2775. UHRF1 targets DNMT1 for DNA methylation through cooperative binding of 50. Bird, A. 2002. DNA methylation patterns and epigenetic memory. Genes Dev. hemi-methylated DNA and methylated H3K9. Nat. Commun. 4: 1563. 16: 6–21. 27. Papadakis, K. A., and S. R. Targan. 2000. Role of cytokines in the pathogenesis 51. Jones, P. A., and S. B. Baylin. 2002. The fundamental role of epigenetic events in of inflammatory bowel disease. Annu. Rev. Med. 51: 289–298. cancer. Nat. Rev. Genet. 3: 415–428. 28. Ma, T. Y., G. K. Iwamoto, N. T. Hoa, V. Akotia, A. Pedram, M. A. Boivin, and 52. Achour, M., X. Jacq, P. Ronde´, M. Alhosin, C. Charlot, T. Chataigneau, H. M. Said. 2004. TNF-alpha-induced increase in intestinal epithelial tight M. Jeanblanc, M. Macaluso, A. Giordano, A. D. Hughes, et al. 2008. The in- junction permeability requires NF-kappa B activation. Am. J. Physiol. Gastro- teraction of the SRA domain of ICBP90 with a novel domain of DNMT1 is intest. Liver Physiol. 286: G367–G376. involved in the regulation of VEGF gene expression. 27: 2187–2197. 29. Rutgeerts, P., G. D’Haens, S. Targan, E. Vasiliauskas, S. B. Hanauer, 53. Nguyen-Chi, M., B. Laplace-Builhe, J. Travnickova, P. Luz-Crawford, D. H. Present, L. Mayer, R. A. Van Hogezand, T. Braakman, K. L. DeWoody, G. Tejedor, Q. T. Phan, I. Duroux-Richard, J. P. Levraud, K. Kissa, G. Lutfalla, et al. 1999. Efficacy and safety of retreatment with anti-tumor necrosis factor et al. 2015. Identification of polarized macrophage subsets in zebrafish. eLife 4: antibody (infliximab) to maintain remission in Crohn’s disease. Gastroenterol- e07288. ogy 117: 761–769. 54. Rugtveit, J., E. M. Nilsen, A. Bakka, H. Carlsen, P. Brandtzaeg, and H. Scott. 1997. 30. Hanauer, S. B., B. G. Feagan, G. R. Lichtenstein, L. F. Mayer, S. Schreiber, Cytokine profiles differ in newly recruited and resident subsets of mucosal mac- J. F. Colombel, D. Rachmilewitz, D. C. Wolf, A. Olson, W. Bao, and rophages from inflammatory bowel disease. Gastroenterology 112: 1493–1505. P. Rutgeerts, ACCENT I Study Group. 2002. Maintenance infliximab for 55. Garcia-Carbonell, R., J. Wong, J. Y. Kim, L. A. Close, B. S. Boland, T. L. Wong, Crohn’s disease: the ACCENT I randomised trial. Lancet 359: 1541–1549. P. A. Harris, S. B. Ho, S. Das, P. B. Ernst, et al. 2018. Elevated A20 promotes 31.Assi,K.,R.Pillai,A.Go´mez-Mun˜oz, D. Owen, and B. Salh. 2006. The TNF-induced and RIPK1-dependent intestinal epithelial cell death. Proc. Natl. specific JNK inhibitor SP600125 targets tumour necrosis factor-alpha Acad. Sci. USA 115: E9192–E9200. 1 Supplementary figures and figure legends

2 3 Supplementary figure 1 4 Uhrf1fl/flLyz2-Cre mice do not spontaneously develop colitis under SPF condition. 5 A. Immunoblot analysis of Uhrf1 in BMDMs derived from WT and Uhrf1fl/flLyz2-Cre 6 mice. 7 B-C. H&E staining (B) and Zo1 immunofluorescent staining (C) of colon tissues from 8 WT and Uhrf1fl/flLyz2-Cre mice (6-month old) (n = 6 per group). Scale bars, 100 µm. 9 D. Bone marrow progenitor cells obtained from WT and Uhrf1fl/flLyz2-Cre mice were 10 cultured with L929 cell supernatants for 7 days. CD11b+ F4/80+ cell subset was 11 analyzed by flow cytometry at day 7. 12 E. CD68 (Red) immunohistochemical staining of colon tissues from normally housed 13 WT and Uhrf1fl/flLyz2-Cre mice. Representative images (Left) and quantification of 14 CD68+ macrophages in colonic lamina propria (Right) were shown (n = 6 per group). 15 Scale bars, 20 µm. 16 F. Tnf-α expression in CD11b+ F4/80+ macrophages in colonic lamina propria were 17 assessed by flow cytometry (n = 6 per group). 18 Data represented three independent experiments. Error bars represented mean ± SEM. 19 N.S., not significant (unpaired t test). 20

1

21

22 23 Supplementary figure 2 24 Uhrf1 mutant does not affect the differentiation of macrophages. 25 A. The coding sequence of Uhrf1 in BMDMs from WT and Uhrf1YP187/188AA mice were 26 verified by sequencing. 27 B. Bone marrow progenitor cells obtained from WT and Uhrf1YP187/188AA mice were 28 cultured with L929 cell supernatants for 7 days. CD11b+ F4/80+ cell subset was analyzed 29 by flow cytometry at day 7. 30 C. CD68 (Red) immunohistochemical staining of colon tissues from normally housed 31 WT mice and Uhrf1YP187/188AA mice. Representative images (Left) and quantification of 32 CD68+ macrophages in colonic lamina propria (Right) were shown (n = 6 per group). 33 Scale bars, 20 µm. 34 D. Tnf-α expression in CD11b+ F4/80+ macrophages from colonic lamina propria were 35 assessed by flow cytometry (n = 6 per group). 36 E-F. H&E staining (E) and Zo1 immunofluorescent staining (F) of colon tissues from 37 WT and Uhrf1YP187/188AA mice (6-month old) (n = 6 per group). Scale bars, 100 µm. 38 Data represented three independent experiments. Error bars represented mean ± SEM. 39 N.S., not significant (unpaired t test). 40 41

2

56

57 Supplementary figure 3

58 Tnf-α reduces the stability of Uhrf1 protein.

59 A. Tnf-α protein levels in colon tissue homogenates from untreated WT mice and DSS-

60 treated WT mice were assessed by enzyme-linked immunosorbent assay (n = 6 per

61 group).

62 B. Uhrf1 mRNA levels in BMDMs incubated with rTnf-α (100 nM) for 12 hrs or not

63 were analyzed by qPCR.

64 C. Immunofluorescence staining of Uhrf1 in BMDMs incubated with rTnf-α (100 nM)

65 for 12 hrs or not. Bright-field images (Left) and fluorescent images (Right) were shown.

66 Scale bars, 10 µm.

67 Data represented three independent experiments. Error bars represented mean ± SEM.

68 *P < .05, N.S., not significant (unpaired t test).

69 5

70 Supplementary Table I

71 Table Ia. The primers for qPCR. Primer Sequence(5’-3’) Tnf-α-forward ATTCGAGTGACAAGCCTGTAGCCCA Tnf-α-reverse CTGGGAGACAAGGTACAACCCA Il-6-forward TGTCTATACCTTCACAAGTCGGAG Il-6-reverse GCACAACTCTTTTCTCATTTCCAC Il-Iβ-forward GCCCATCCTCTGTGACTCA Il-Iβ-reverse AGGCCACAGGTATTTTGTC Il-10-forward GCTCTTACTGACTGGCATGAG Il-10-reverse CGCAGCTCTAGGAGCATGTG iNOS-forward CAGCTGGCTGTACAAACCTT iNOS-reverse CATTGGAAGTGAAGCGTTTCG Arg1-forward GGAATCTGCATGGGCAACCTGTGT Arg1-reverse AGGGTCTACGTCTCGCAAGCCA Arg2-forward TAGGGTAATCCCCTCCCTGC Arg2-reverse AGCAAGCCAGCTTCTCGAAT Ccl2-forward GATGCAGTTAACGCCCCACT Ccl2-reverse ACCCATTCCTTCTTGGGGTC Tgf-β-forward CACCTGCAAGACCATCGACA Tgf-β-reverse CATAGTAGTCCGCTTCGGGC Cd40-forward TTGTTGACAGCGGTCCATCT Cd40-reverse TCTCAAGAGCTGTGCAGTGG Zo1-forward GAGCGGGCTACCTTACTGAAC Zo1-reverse GTCATCTCTTTCCGAGGCATTAG Tricellulin-forward TTCCGAAGCCTATCGTGATGC Tricellulin-reverse GAACACAGCCTTATAGCGTTCT Claudin5-forward TATGAATCTGTGCTGGCGCT Claudin5-reverse GTGCTACCCGTGCCTTAACT Claudin10-forward AATCGTCGCCTTCGTAGTCTC Claudin10-reverse GTTGGCAAAATAAGTGGCTGTG Hprt-forward GTTGGGCTTACCTCACTGCT Hprt-reverse TAATCACGACGCTGGGACTG

72

a The sequence of primers using for qPCR. 6