MicroRNA 301A Promotes Intestinal Inflammation and Colitis-Associated Cancer Development by Inhibiting BTG1

Gastroenterology 2017;152:1434–1448

MicroRNA 301A Promotes Intestinal Inflammation and Colitis-Associated Cancer Development by Inhibiting BTG1 Chong He,1,* Tianming Yu,1,* Yan Shi,1 Caiyun Ma,1 Wenjing Yang,1 Leilei Fang,1 Mingming Sun,1 Wei Wu,1 Fei Xiao,2 Feifan Guo,2 Minhu Chen,3 Hong Yang,4 Jiaming Qian,4 Yingzi Cong,5 and Zhanju Liu1 1 Department of Gastroenterology, The Shanghai Tenth People’s Hospital, Tongji University, Shanghai, China; 2Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institute for Biological Sciences, Chinese Academy of Sciences, the Graduate School of the Chinese Academy of Sciences, Shanghai, China; 3Department of Gastroenterology, The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China; 4Department of Gastroenterology, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, China; 5Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, TX

BASIC AND TRANSLATIONAL AT

BACKGROUND & AIMS: Intestinal tissues from patients with inflammatory bowel disease (IBD) and colorectal cancer have increased expression of microRNA-301a (MIR301A) compared with tissues from patients without IBD. We studied the mechanisms of MIR301A in the progression of IBD in human tissues and mice. METHODS: We isolated intestinal epithelial cells (IECs) from biopsy samples of the colon from 153 patients with different stages of IBD activity, 6 patients with colitisassociated cancer (CAC), and 35 healthy individuals (controls), enrolled in the study in Shanghai, China. We measured expression of MIR301A and BTG anti-proliferation factor 1 (BTG1) by IECs using quantitative reverse-transcription polymerase chain reaction. Human colon cancer cell lines (HCT-116 and SW480) were transfected with a lentivirus that expresses MIR301A; expression of cytokines and tight junction proteins were measured by quantitative reverse transcription polymerase chain reaction, flow cytometry, and immunofluorescence staining. We generated mice with disruption of the microRNA-301A gene (MIR301A-knockout mice), and also studied mice that express a transgene-encoding BTG1. Colitis was induced in knockout, transgenic, and control (C57BL/B6) mice by administration of dextran sulfate sodium (DSS), and mice were given azoxymethane to induce colorectal carcinogenesis. Colons were collected and analyzed histologically and by immunohistochemistry; tumor nodules were counted and tumor size was measured. SW480 cells expressing the MIR301A transgene were grown as xenograft tumors in nude mice. RESULTS: Expression of MIR301A increased in IECs from patients with IBD and CAC compared with controls. MIR301Aknockout mice were resistant to the development of colitis following administration of DSS; their colon tissues expressed lower levels of interleukin 1b (IL1b), IL6, IL8, and tumor necrosis factor than colons of control mice. Colon tissues from MIR301A-knockout mice had increased epithelial barrier integrity and formed fewer tumors following administration of azoxymethane than control mice. Human IECs expressing transgenic MIR301A down-regulated expression of cadherin 1 (also called E-cadherin or CDH1). We identified BTG1 mRNA as a target of MIR301A; levels of BTG1 mRNA were reduced in inflamed mucosa from patients with active IBD compared with controls. There was an inverse correlation between levels of BTG1 mRNA and levels of MIR301A in inflamed mucosal tissues from patients with active IBD. Human colon cancer cell lines that expressed a MIR301A transgene increased proliferation;

they had increased permeability and decreased expression of CDH1 compared with cells transfected with a control vector, indicating reduced intestinal barrier function. BTG1 transgenic mice developed less severe colitis than control mice following administration of DSS. SW480 cells expressing anti-MIR301A formed fewer xenograft tumors in nude mice than cells expressing a control vector. CONCLUSIONS: Levels of MIR301A are increased in IECs from patients with active IBD. MIR301A reduces expression of BTG1 to reduce epithelial integrity and promote inflammation in mouse colon and promotes tumorigenesis. Strategies to decrease levels of MIR301A in colon tissues might be developed to treat patients with IBD and CAC.

Keywords: Mouse model; Crohn’s disease; Ulcerative colitis; Immune regulation.

I

nflammatory bowel diseases (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), are chronic inflammatory disease of the intestines that appear to be disorders of the host immune response to gut microbiota, which involves a state of local immune hyper-reactivity. Accumulating evidences from a variety of mouse models that develop chronic intestinal inflammation resembling human IBD have provided strong support for the hypothesis that IBD is caused by a dysregulated mucosal T-cell response to gut microbiota antigens initiated by an abnormal innate response in a genetically susceptible host.1 However, the mechanisms remain largely unclear. *Authors share the co-first authorship. Abbreviations used in this paper: AOM, azoxymethane; BTG1, BTG anti-proliferation factor 1; CAC, colitis-associated cancer; CD, Crohn’s disease; CRC, colorectal cancer; DSS, dextran sulfate sodium; UC, ulcerative colitis; IBD, inflammatory bowel disease; IEC, intestinal epithelial cells; IL, interleukin; PCNA, proliferating cell nuclear antigen; PCR, polymerase chain reaction; Tg, transgenic; Th, T-helper; TNF, tumor necrosis factor; MIR, microRNA. Most current article © 2017 by the AGA Institute 0016-5085/$36.00 http://dx.doi.org/10.1053/j.gastro.2017.01.049

May 2017

BACKGROUND AND CONTEXT microRNA-301a (MIR301A) is involved in the pathogenesis of a variety of autoimmune diseases and cancers. However, its roles in maintaining intestinal epithelial cell (IEC) homeostasis and inducing colitisassociated cancer (CAC) are still elusive. NEW FINDINGS IEC-expressing MIR301A down-regulates BTG1 expression, which contributes to compromised epithelial barrier integrity and the induction of intestinal mucosal inflammation and CAC. LIMITATIONS This study did not determine the role of MIR301A in the pathogenesis of colorectal cancer in non-IBD tumor models. IMPACT Strategies to decrease MIR301A expression in intestinal epithelial tissues may be warranted to treat patients with IBD and CAC.

The intestines contain a dynamic community of massive and diverse microbiota that is comprised of more than 1000 species.2 With such an enormous microbiota challenge, it is crucial for the intestinal immune system to respond properly to live in harmony with the microbiota.3–7 Among the multiple mechanisms involved in maintaining intestinal homeostasis, intestinal epithelial cells (IEC) serve as the first line of defense against microbiota by establishing the barrier function through formation of complex proteinprotein networks, including adherens junctions and tight junctions, which mechanically link adjacent cells and seal the intercellular space.8 Intestinal barrier dysfunction has been considered to be critical in the predisposition to and exacerbation of IBD by providing opportunity for invading gut microbiota to stimulate immune cell production of proinflammatory cytokines, which play a critical role in the pathogenesis of IBD.9 In addition to antigen-presenting cells and T cells, which are 2 major sources of cytokine production, IEC have also been identified as important producers of a variety of cytokines to modulate intestinal mucosal immunity and barrier function.10 It has been shown that several tight junction proteins and adherens junction proteins are down-regulated, which results in increased intestinal epithelial permeability in both colitis and colorectal cancer.11,12 However, the mechanisms by which the IEC barrier function is regulated, thus maintaining the intestinal homeostasis, is still largely unknown. MicroRNAs (MIRs), a group of short, non-coding and single-stranded RNAs, have been implicated in regulating a variety of immune responses and diseases through binding to the 3’-UTR of target genes as endogenous inhibitors of translational processes.13,14 MIR301A, which is increased more in colon cancer tissues and cell lines than in normal intestinal mucosal tissues,15 has been shown to enhance NF-kB activation through targeting the NF-kB-repressing

factor (NKRF) gene,16 and promote autoimmune demyelination through regulating myelin-reactive T-helper (Th)17 cell differentiation in experimental autoimmune encephalomyelitis.17 Our previous studies further demonstrated that MIR301A is highly expressed in the peripheral blood mononuclear cells and inflamed mucosa from patients with active IBD, and possibly contributes to the pathogenesis of IBD through enhancing Th17 cell differentiation and production of tumor necrosis factor (TNF) by targeting SNIP1.18 In the current study, we demonstrated that MIR301A expression was increased in IEC from patients with IBD and colitis-associated cancer (CAC) compared with that from healthy donors. MIR301A–/– mice were resistant to the development of colitis induced by dextran sulfate sodium (DSS), as well as of azoxymethane (AOM)/DSS-induced CAC. Mechanistically, MIR301A promoted NF-kB activation in IEC and induced intestinal barrier dysfunction by targeting BTG anti-proliferation factor 1 (BTG1), whose expression was decreased in the IEC of IBD patients compared with that from healthy donors. Our data thereby demonstrated that expression of MIR301A in IEC regulated intestinal inflammation and CAC through impairing epithelial barrier function.

Materials and Methods More detailed information is provided in the online Supplementary Materials and Methods section.

Subjects Colonoscopic biopsies were obtained from inflamed and unaffected sites of the colons from 45 patients with active CD (A-CD), 36 CD patients in remission (R-CD), 38 patients with active UC (A-UC), and 34 UC patients in remission (R-UC), as well as from normal colonic mucosa of 35 healthy individuals. EDTA anti-coagulated blood samples (20 mL) were also collected from the same patterns of patients and healthy individuals after overnight fasting as described above. All patients enrolled in this study were from the Department of Gastroenterology at the Shanghai Tenth People’s Hospital of Tongji University (Shanghai, China). The baseline characteristics are described in online Supplementary Table 1. The diagnosis of CD or UC was based on clinical, radiological, and endoscopic examination and histologic findings. The disease severity was assessed according to international standard criteria such as the CD activity index for the diagnosis of CD patients and Mayo scores for UC patients. Moreover, intestinal mucosal biopsies were also obtained from IBD patients with CAC for further analysis, including 1 CD (male, 48 years old) and 1 UC (male, 56 years old) patients from the Shanghai Tenth People’s Hospital, 2 UC patients (male, 68 years old; female, 73 years old) from the First Affiliated Hospital of Anhui Medical University (Hefei, China), and 2 UC patients (both males; 58 and 68 years old, respectively) from the Peking Union Medical College Hospital (Beijing, China). Diagnosis was made on the basis of hematoxylin and eosin staining on both endoscopic biopsies and surgical specimens obtained from colonoscopy and surgical resection, respectively. The study was approved by the Institutional Review Board for Clinical Research of the

BASIC AND TRANSLATIONAL AT

EDITOR’S NOTES

MIR301A in IBD and CAC 1435

1436 He et al Shanghai Tenth People’s Hospital of Tongji University. Written informed consent was also obtained from all subjects before the study protocol.

Mice Female C57BL/B6 mice were purchased from the Nanjing Biomedical Research Institution of Nanjing University (Nanjing, China), bred and maintained in the animal facility of the Shanghai Tenth People’s Hospital, Tongji University. MIR301A–/– mice in a C57BL/B6 background were generated at the Shanghai Key Laboratory of Regulatory Biology (Shanghai, China) using a CRISPR/Cas9 system as reported previously,19,20 and maintained in the animal facility of the Shanghai Tenth People’s Hospital. All mice were maintained under specific pathogen-free conditions with a 12-hour light cycle and given a regular chow diet and water ad libitum. Mice were used at 8–10 weeks of age. The BTG1 transgenic mice were generated and described previously,21 and maintained in the animal facility of the Shanghai Tenth People’s Hospital. Animal experiments in this study were approved by the Institutional Animal Care and Use Committee at Tongji University.

Generation of MIR301A–/– Mice

BASIC AND TRANSLATIONAL AT

MIR301A sgRNAs were designed by CRISPR Design Tool (http://crispr.mit.edu) according to MIR301A gene sequence (NCBI gene ID: 723834). Two sequences were chosen to design MIR301A sgRNAs: 5’-TTCCTGCTCACTCCTGCTAACGG-3’ and 5’-CCGCGAGCAGGTTGCACACCTTT-3’. Two sgRNAs were transcribed by in vitro Transcription T7 Kit (Cat. No. 6140; Takara; Shiga, Japan). The Cas9 open reading frame was amplified from pX33043 (Addgene #42230; Cambridge, MA) and recombined into pET28 vector (Cat. No. VT0129; Novagen; Milwaukee, WI). Cas9 mRNA was obtained by mMESSAGE mMACHINE SP6 Transcription Kit (Cat. No. AM1340; Life Technologies; Carlsbad, CA) according to the manufacturer’s instructions. Cas9 mRNA and MIR301A sgRNA were microinjected into the zygotes from female C57BL/B6 mice after ovulation induction. To identify the genotyping of founder mice and their progeny, the mouse genomic DNA containing mutated sequence was amplified from mouse tail by polymerase chain reaction (PCR) (forward 5’-TATGGTAGGTAACCCAGACGC-3’; reverse 5’-TTCCCAGAACCTATTCAAGTC-3’). The PCR products (700 bp) were sent for sequencing after annealing and digestion with T7 endonuclease I (Cat. No. M0302S; New England Biolabs, Beverly, MA). The offspring with mutated sequence were verified by alignment with WT MIR301A sequence. The mutant alleles were crossed to C57BL/B6 mice for further 4 generations backcrossed. The homozygous MIR301A–/– mice were used in this study. The founder 1 heterozygotes were shipped to the animal facility at Shanghai Tenth People’s Hospital, which were then crossed with each other to obtain the homozygotes for MIR301A for further study.

DSS-Induced Colitis in Mice The DSS-induced colitis model in mice was established as reported previously.18,22 Briefly, mice were treated with 2% DSS (molecular weight, 36,000–50,000 Da; MP Biomedicals, LLC, Solon, OH) in drinking water for 7 days, followed by DSS-free water for an additional 3 days. The severity of colitis was scored daily by recording standard parameters including

Gastroenterology Vol. 152, No. 6 body weight, diarrhea, and bloody stools. Colonic tissues were removed, fixed in 10% formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. The histologic grading of colonic inflammation was graded from 0 to 3 as described previously.23 The disease activity index (DAI) was measured as reported previously.24

AOM/DSS-Induced CAC in Mice CAC was induced in mice as reported elsewhere.25,26 Briefly, 8- to 10-week-old female MIR301A–/– and WT C57BL/B6 mice were first injected intraperitoneally with AOM (A5486; Sigma-Aldrich; St. Louis, MO; 12 mg/kg). Seven days after post-AOM intervention, these mice were administrated with 2% DSS (MP Biomedicals) in drinking water for 7 consecutive days, followed by 14 days of recovery. This cycle was repeated twice followed by regular drinking water until day 81, when all mice were sacrificed. During the study, body weight, diarrhea, and haematochezia were monitored and described as DAI. All mice were sacrificed, and the colon was then cut open longitudinally. The presence of gross tumors was measured, and the tumor nodules were quantified. Tissue samples were also frozen in liquid nitrogen for RNA analysis or fixed in 10% formalin for histopathologic analysis.

Statistical Analysis Data were expressed as mean ± SEM, and analyzed using Prism 5.0 software (Graphpad Software, San Diego, CA). Statistical comparisons were performed using unpaired Student t test, paired Student t test (Figures 1B and 2B), and one-way analysis of variance (ANOVA). Statistical significance was set at P < .05. Pearson’s correlation was performed to analyze the correlation of interleukin 1b (IL1b) levels in sera and IEC and MIR301A expression in IEC from IBD patients.

Results MIR301A Expression Is Up-Regulated in IEC of IBD Patients We recently demonstrated that MIR301A expression was significantly increased in the peripheral blood mononuclear cell and inflamed mucosa of patients with IBD compared with that in healthy donors, and that IEC was a major source of MIR301A in addition to CD4þ T cells.18 However, the expression pattern of MIR301A in IEC of IBD patients and its role in the pathogenesis of IBD remain unknown. To examine MIR301A expression in the IEC of IBD patients, we isolated primary IEC from colonic biopsies of IBD patients and healthy volunteers during endoscopic examination. As shown in Figure 1A, expression of MIR301A was markedly up-regulated in IEC from both active CD and UC patients compared with that in healthy controls. Moreover, MIR301A expression was also increased in the IEC of inflamed mucosa compared with that in normal mucosa adjacent to inflamed area from the same IBD patients, while the level of MIR301A expression in the IEC from unaffected mucosa of IBD patients was comparable to that in healthy volunteers (Figure 1B). Because MIR301A has been reported to regulate NF-kB activation,16 we determined NF-kB

MIR301A in IBD and CAC 1437

BASIC AND TRANSLATIONAL AT

May 2017

Figure 1. MIR301A expression is significantly increased in IEC of IBD patients. (A) IEC (1  106) were isolated from colonic biopsies of healthy controls (HC, n ¼ 35), inflamed mucosa from active CD (A-CD, n ¼ 45) and active UC (A-UC, n ¼ 38) patients, and colonic mucosa from CD patients with remission (R-CD, n ¼ 36) and UC patients with remission (R-UC, n ¼ 34). Expression of MIR301A in IEC was analyzed by qRT-PCR and normalized to U6 expression. ***P < .001. (B) Relative expression of MIR301A in IEC from unaffected and inflamed intestinal mucosa from the same A-CD patients (n ¼ 45) and A-UC patients (n ¼ 38). *** P < .001. (C) The mRNA levels of NF-kB p65 in IEC described in Figure 1A were analyzed by qRT-PCR. ***P < .001. (D) A correlation analysis was performed between the relative levels of NF-kB p65 mRNA and MIR301A expression in IEC of inflamed mucosa from 45 A-CD patients and 38 active A-UC patients (Spearman’s rank correlation coefficient, CD, R ¼ .6231, P ¼ .0003; UC, R ¼ .6490, P < .0001). (E) A correlation analysis was done between the relative levels of IL1b mRNA and MIR301A expression in IEC of inflamed mucosa from 45 A-CD patients and 38 A-UC patients (Spearman’s rank correlation coefficient, CD, R ¼ .7490, P < .001; UC, R ¼ .7438, P < .001). (F) A correlation analysis was performed between the serum levels of IL1b and MIR301A expression in IEC of colonic mucosa from 45 A-CD patients and 38 A-UC patients (Spearman’s rank correlation coefficient, CD, R ¼ .7343, P < .001; UC, R ¼ .6239, P < .001).

1438 He et al

BASIC AND TRANSLATIONAL AT

p65 expression in IEC of IBD patients. As shown in Figure 1C, expression of NF-kB p65 was increased in IEC of IBD patients compared with that in healthy controls. Interestingly, expression of MIR301A in IEC of IBD patients was positively correlated with the levels of NF-kB p65 mRNA (Spearman’s rank correlation coefficient in CD, R ¼ .6231, P ¼ .0003; in UC, R ¼ .6490, P < .0001) (Figure 1D), indicating that expression of MIR301A in IEC could positively regulate the NF-kB pathway. Because proinflammatory cytokines (eg, TNF, IL1b, and IL6) are highly expressed in inflamed mucosa of IBD patients, and certain MIRs have been found to be regulated by different cytokines,3,13,27 we then investigated whether proinflammatory cytokines could regulate MIR301A expression in IEC of IBD patients. For this purpose, primary IEC were isolated from normal colonic mucosa of patients with colorectal cancer who underwent a colectomy and stimulated in vitro with different cytokines as indicated for 6 hours. The levels of MIR301A expression were analyzed by quantitative reverse transcription-PCR (qRT-PCR). We found that IL1b markedly induced, while IL6 modestly induced, MIR301A expression (Supplementary Figure 1). Interestingly, the levels of IL1b in sera were positively correlated with MIR301A expression in IEC from CD and UC patients, respectively (Spearman’s rank correlation coefficient in CD, R ¼ .7343, P < .01; in UC, R ¼ .6239, P < .01) (Figure 1E). The same correlation was also observed between the levels of IL1b mRNA and MIR301A expression in IEC from CD and UC patients, respectively (Spearman’s rank correlation coefficient in CD, R ¼ .7490, P < .01; in UC, R ¼ .7438, P < .01) (Figure 1F). These results indicate that IL1b promotes expression MIR301A in IEC of IBD patients.

IL1b Induces MIR301A Expression in IEC in a c-Jun–Dependent Manner We then investigated the underlying mechanisms whereby IL1b regulates MIR301A expression in IEC. Because human primary IECs are short-lived after isolation ex vivo, 5 human colon cancer cell lines (ie, HCT-116, SW480, SW620, HT29, and LoVo) were utilized to perform the mechanistic studies. These cell lines expressed significantly higher levels of MIR301A compared with freshly isolated primary IEC from healthy individuals (Supplementary Figure 2). We then stimulated those colon cancer cell lines in vitro with a series of doses of IL1b (0, 1, 5, and 10 ng/mL) for 72 hours, and found that IL1b strongly stimulated expression of MIR301A in a dose-dependent manner (Supplementary Figure 3A). Because IL1b signaling has been shown to activate NF-kB and the JNK MAPK pathways,28 we utilized the MEME-ChIP datasets for the motif scanning from large nucleotide and LASAGNA-Search 2.0 database, and searched transcription factor binding sites of the MIR301A motif. As shown in Supplementary Figure 3B, c-Jun was predicted to potentially bind to MIR301A promoter. We then investigated whether MIR301A expression is promoted by IL1b-induced c-Jun binding of the MIR301A promoter. IL1b strongly induced

Gastroenterology Vol. 152, No. 6

c-Jun phosphorylation in a dose-dependent manner in all 5 cell lines (Supplementary Figure 3C). Moreover, ChIP and luciferase assays further verified the direct binding of c-Jun to the MIR301A promoter (Supplementary Figure 3D and E). Interestingly, in vitro culture of 5 cell lines with JNK inhibitor (JNK-IN-7) significantly abrogated IL1b-induced MIR301A expression (Supplementary Figure 3F).

MIR301A Enhances NF-kB Signaling in IEC After confirming there was a positive correlation between the levels of NF-kB p65 mRNA and MIR301A expression in IEC of IBD patients (Figure 1D), we next investigated whether MIR301A could promote NF-kB signaling in IEC. To this end, HCT-116 and SW480 cells were transduced with a lentivirus encoding pre-MIR301A (LV-MIR301A) or control MIR (LV-MIRctrl). Overexpression of MIR301A was verified in cells transduced with LV-MIR301A (Supplementary Figure 4A). We found that NF-kB activation was markedly increased in these cells overexpressing MIR301A (Supplementary Figure 5A). Moreover, we also used a lentivirus that encoded a reverse complementary sequence of MIR301A (LV-anti-MIR301A), which is able to inhibit its expression, and examined the effect of LV-anti-MIR301A on NF-kB activation in HCT-116 and SW480 cells after transduction (Supplementary Figure 4B). Consistently, in vitro silencing of MIR301A significantly suppressed the activation of NF-kB in HCT-116 and SW480 cells compared with transduction with vehicle lentivirus (Supplementary Figure 5A). Knockdown of MIR301A also inhibited TNF-, IL1b- or LPS-induced NF-kB activation (Supplementary Figure 5B). Moreover, ectopic expression of MIR301A promoted, whereas knockdown of MIR301A inhibited, expression of IL1b, IL6, IL8, and TNF (Supplementary Figure 5C and E). The levels of various NF-kB-associated proteins were also up-regulated by MIR301A (Supplementary Figure 5D and F). Taken together, these data indicate that MIR301A promotes NF-kB signaling in IEC.

MIR301A Targets BTG1 In an attempt to define how MIR301A regulates intestinal inflammation, we sought to determine the downstream targets of MIR301A that modulate intestinal inflammation. To this end, we used 3 bioinformatics prediction tools (Miranda, miRDB, TargetScan) to predict potential targets of MIR301A based on their sequence complementarity. BTG1, a member of BTG/Tob family, which has been shown to regulate cell cycle progression in a variety of cells,29–31 was predicted as a potential target (Supplementary Figure 6A). To determine whether the BTG1 gene is indeed the target of MIR301A, we constructed dual-luciferase reporter vectors containing the predicted seed sequence in the 3’-UTR of BTG1, as well as the corresponding mutant vectors in which 3 random nucleotide mutants were introduced into the seed sequences. As illustrated in Supplementary Figure 6B, co-transduction of LV-MIR301A markedly suppressed the activity of Renilla luciferase containing the BTG1 3’-UTR reporter sequence, but not the mutant BTG1 (BTG1 mut) 3’-UTR reporter

sequence, indicating that MIR301A specifically downregulates BTG1 expression. To further confirm MIR301A targeting BTG1 in human IEC, 5 colon cancer cell lines (ie, HCT-116, SW480, SW620, HT29, and LoVo) were transduced with LV-MIR301A or LV-MIRctrl, and BTG1 mRNA was analyzed by qRT-PCR. Ectopic expression of MIR301A significantly down-regulated expression of BTG1 mRNA compared with controls (Supplementary Figure 6C). Importantly, expression of BTG1 in intestinal mucosa was significantly up-regulated in MIR301A–/– mice compared with that in WT mice (Supplementary Figure 6D). Taken together, these data indicate that MIR301A indeed targets BTG1 both in human and mice. Because BTG1 has been shown to inhibit NF-kB activation,32 we then investigated whether MIR301A promotes NF-kB signaling through inhibiting BTG1. HCT116 or SW480 cells were stimulated with IL1b in the presence or absence of BTG1 plasmid, and activation of NF-kB was determined by BioPlex enzyme-linked immunosorbent assay (Bio-Rad Laboratories, Shanghai, China). Overexpression of BTG1 significantly down-regulated activation of NF-kB in HCT-116 and SW480 cells induced by IL1b (Supplementary Figure 7A). IL1b-inducible c-Jun phosphorylation was also inhibited by BTG1 transfection (Supplementary Figure 7B). These data thereby suggest that MIR301A promotes NF-kB activation, possibly through targeting BTG1, leading to inhibiting c-Jun phosphorylation.

BTG1 Expression Is Decreased in Inflamed Mucosa of IBD Patients We then assessed BTG1 expression in intestinal mucosa from IBD patients and healthy controls. In contrast to increased expression of MIR301A, BTG1 expression was decreased in inflamed mucosa of patients with active CD or UC compared with that in healthy controls (Figure 2A). BTG1 expression was also decreased in inflamed mucosa compared with that in unaffected mucosa from the same IBD patients (Figure 2B). Immunohistochemical staining further confirmed that the number of BTG1-positive cells was significantly decreased in the epithelia and lamina propria of inflamed mucosa from active CD and UC patients compared with that in healthy controls (Figure 2C). Importantly, the levels of IEC-derived BTG1 were found to inversely correlate with MIR301A expression in IBD patients (Figure 2D).

MIR301A–/– Mice Are Resistant to DSS-Induced Colitis To further determine whether MIR301A promoted intestinal inflammation, we generated MIR301A–/– mice (Supplementary Figure 8). WT (n ¼ 10) and MIR301A–/– mice (n ¼ 10) were treated with 2% DSS in drinking water for 7 days, followed by 3 days of regular water alone. The clinical signs of colitis, including body weight, stool consistency, and rectal bleeding, were monitored daily. WT mice developed severe colitis (Figure 3 A–C). Consistent with IBD patient data, expression of MIR301A was significantly increased, whereas expression of BTG1 was markedly

MIR301A in IBD and CAC 1439

decreased in IEC from DSS-treated WT mice compared with that in untreated WT mice (Supplementary Figure 9). However, MIR301A–/– mice developed only mild colitis with a slight decrease of body weight, colon length, and pathologic scores (Figure 3 A–C). The number of neutrophils, DC, and CD4þ T cells was markedly decreased in the colons of MIR301A–/– mice compared with that in WT mice (Figure 3D).

Loss of MIR301A Enhances Intestinal Barrier Function To further determine the role of IEC-derived MIR301A in regulating intestinal mucosal inflammation, we generated the BM chimeric mice by reconstituting lethally irradiated WT or MIR301A–/– mice with WT BM cells. Colitis was then induced with DSS as described above. As shown in Figure 4 A–C, MIR301A–/– chimeras reconstituted with WT BM cells (ie, MIR301A expression was deficient in nonhemopoietic cells, including IEC) developed a milder form of colitis with less body weight loss, lower DAI, and lower pathologic scores compared with WT recipients. Interestingly, expression of cadherin 1 (also called E-cadherin or CDH1), an important molecule in regulating epithelial barrier function, was increased in the epithelia from MIR301A–/– chimeras than that in WT chimeras (Figure 4D). Both epithelial permeability and IEC production of IL1b, IL6, IL8, and TNF were increased to a lesser extent in MIR301A–/– chimeras after DSS treatment compared with those in WT chimeras (Figure 4 E and F). Moreover, the WT chimeras reconstituted with MIR301A–/– BM cells (ie, MIR301A expression was deficient in hemopoietic cells, including CD4þ T cells) developed mild colitis (Supplementary Figure 10 A and B), characterized by decreased colonic IL17A transcript expression compared with WT mice (Supplementary Figure 10C). We next determined the role of MIR301A in modulating tight junction and intestinal permeability. HCT-116 and SW480 cells were transduced with LV-MIR301A and LV-MIRctrl, respectively, and then cultured in a monolayer on Transwell supports (Corning, MA). Overexpression of MIR301A inhibited transepithelial electrical resistance (Supplementary Figure 11A) but promoted fluorescein isothiocyanate permeability (Supplementary Figure 11B). Moreover, ectopic expression of MIR301A also enhanced ethylhydroxyethyl cellulose translocation, an indicator of the dysfunction of intestinal barrier33,34 (Supplementary Figure 11C), but inhibited expression of CDH1, indicating a dramatic disruption of tight junctions (Supplementary Figure 11 D and E). Collectively, these data demonstrate that expression of MIR301A in IEC could promote mucosal permeability, leading to epithelial barrier dysfunction.

BTG1 Protects Mice from DSS-Induced Colitis Because we have identified BTG1 as a target of MIR301A and confirmed its expression was down-regulated in inflamed mucosa of IBD patients, we then investigated whether BTG1 plays a protective role in intestinal inflammation by treating WT and BTG1 transgenic (Tg) mice21

BASIC AND TRANSLATIONAL AT

May 2017

1440 He et al

Gastroenterology Vol. 152, No. 6

BASIC AND TRANSLATIONAL AT

Figure 2. BTG1 expression is decreased in inflamed mucosa of IBD patients. (A and B) IEC (1  106) were isolated from colonic biopsies as described in Figure 1, and BTG1 expression was analyzed by qRT-PCR. ***P < .001. (C) Representative intestinal sections were prepared from colonic mucosa of a healthy control (HC) (left panel), and inflamed mucosa of an A-CD patient (middle panel) and an A-UC patient (right panel), and stained for BTG1 by immunohistochemistry. Magnification: 200, upper panels; 400, lower panels. The red arrows indicate BTG1þ cells. (D) A correlation analysis was performed between the relative levels of BTG1 mRNA and MIR301A expression in IECs of colonic mucosa of 45 A-CD patients and 38 A-UC patients (Spearman’s rank correlation coefficient, CD, R ¼.8664, P < .001; UC, R ¼.8543, P < .001).

with DSS to induce colitis. BTG1 Tg mice exhibited significantly less body weight loss, lower DAI, and less colon shortening than the WT counterparts after DSS exposure (Figure 5 A–D). Histologic analysis and colonic pathologic scores also revealed less intestinal inflammation in BTG1 Tg mice than that in WT littermates (Figure 5 E and F). Taken together, these data indicate that BTG1 might play a protective role in mice upon DSS insult, thus, further suggesting that miR-301 regulates colitis probably through inhibiting BTG1.

MIR301A–/– Mice Are Resistant to AOM/DSS-induced CAC We then analyzed expression of MIR301A in colonic tumor tissues from patients with CAC and colorectal cancer

and found that it was increased in colonic tumor tissues from patients with CAC compared with healthy controls (Figure 6A), consistent with a previous report.15 To determine whether MIR301A regulated the development of intestinal tumorigenesis, we treated WT and MIR301A–/– mice with AOM/DSS to induce CAC as described previously.35 Over the duration of CAC induction, MIR301A–/– mice displayed less weight loss, slight colon length shortening, and attenuated rectal prolapse compared with WT mice (Figure 6B, C and D, upper panel). Histologic analysis revealed that AOM/DSS-treated MIR301A–/– mice developed better-differentiated adenocarcinomas compared with AOM/DSS-treated WT mice (Figure 6E). When sacrificed at day 81, the numbers of macroscopically visible tumors were significantly decreased in MIR301A–/– mice compared with those in WT mice (Figure 6D, lower panel). Moreover, the

May 2017

MIR301A in IBD and CAC 1441

size of tumors from MIR301A–/– mice was much smaller compared with that from WT mice (Figure 6F).

MIR301A Functions as a Novel Oncogene To investigate how MIR301A regulated CAC development, immunostaining for Ki-67 and proliferating cell nuclear antigen (PCNA) expression in colonic tissues and tumor tissues was performed to determine whether MIR301A regulates tumor cell proliferation. As shown in Figure 7A, expression of Ki-67 and PCNA was up-regulated in colonic tissues of WT mice after AOM/DSS treatment. Interestingly, their expression was decreased in MIR301A–/– mice compared with controls. Moreover, transduction of LV-anti-MIR301A also significantly suppressed the cell

BASIC AND TRANSLATIONAL AT

Figure 3. MIR301A–/– mice are resistant to DSSinduced colitis. The induction of colitis was performed as described in the Material and Methods. (A) Changes of body weight over a period of observation are shown as percent of the initial weight at the start of the experiments. *P < .05 versus DSS-treated WT mice at the same time points. (B) Gross morphology and length of the large bowel in different groups. **P < .01. (C) Histologic sections and colonic pathologic scores of mice in each group. ***P < .001. (D) Percentages of Ly6GþCD11bþ neutrophils, MHC IIþCD11cþ DC, and CD4þ T cells infiltrated in whole colon from each group were analyzed by flow cytometry. *P < .05, **P < .01.

growth in SW480 and HCT-116 cells in vitro (Figures 7B and C), but inhibition of MIR301A did not influence the apoptosis of SW480 cells (Supplementary Figure 12). Next, we sought to establish a xenograft model in nude mice to further determine how MIR301A regulates tumor growth. SW480 cells were transduced with LV-anti-MIR301A and LV-anti-MIRctrl, respectively, and then transplanted into nude mice. As expected, LV-anti-MIR301A–transduced cell growth was markedly inhibited compared with controls (Figure 7D and E). Importantly, MIR301A expression was also significantly down-regulated in tumor tissues derived from LV-anti-MIR301A–transduced SW480 cells compared with controls (Figure 7F).

1442 He et al

Gastroenterology Vol. 152, No. 6

BASIC AND TRANSLATIONAL AT

Figure 4. MIR301A exacerbates intestinal barrier function. Lethally irradiated WT (n ¼ 4) or MIR301A–/– (n ¼ 4) mice received WT BM cells (1  107) to generate chimeric mice. DSS-induced colitis in these chimeric mice was established using the methods as described above. (A) Changes of body weight over a period of observation are shown as percentage of the initial weight at the start of the experiments. *P < .05 vs DSS-treated WT chimeric mice at the same time points. (B) Changes of the DAI. *P < .05 compared with DSS-treated WT chimeric mice at the same time points. (C) Histologic sections and colonic pathologic scores of mice in each group. *P < .05 compared with DSS-treated WT chimeric mice. (D) Representative immunofluorescent images for detection of CDH1 (red) and DAPI (blue). The white arrows indicate representative CDH1þ cells (magnification: 200). (E) Epithelial permeability was determined by fluorescein isothiocyanate-dextran assay, and fold changes of permeability were shown as bar chart. *P < .05 compared with DSS-treated WT chimeras. (F) Expression of IL1b, IL6, IL8, and TNF in colonic epithelia was examined by qRT-PCR. *P < .05 compared with DSS-treated WT chimeras.

Discussion Despite much attention drawn to an essential role of adaptive immunity in intestinal mucosal inflammation, IEC are considered to be one of the most important components of the innate immunity and play a crucial role in maintaining intestinal barrier function.36–39 A compromised epithelial

barrier and accompanying excessive immune responses to gut microbiota have been implicated in the pathogenesis of IBD.9,40 In the current study, we demonstrated that IEC expression of MIR301A was significantly increased in patients with IBD and CAC. By generating MIR301A–/– mice to establish experimental models of colitis and CAC, we also

May 2017

MIR301A in IBD and CAC 1443

found that MIR301A did not only promote intestinal inflammation, possibly through facilitating NF-kB signaling in IEC and compromised integrity of intestinal barrier, but also induced colon tumorigenesis by enhancing tumor cell proliferation. Furthermore, we identified BTG1 as a target gene of MIR301A in inhibiting the development of colitis and CAC. Several MIRs have been implicated in regulating intestinal homeostasis.41 MIR29 is highly expressed in IEC, and further increased in intestinal mucosa of patients with irritable bowel syndrome with diarrhea.42 Interestingly, increased expression of MIR29 is inversely correlated

BASIC AND TRANSLATIONAL AT

Figure 5. BTG1 protects mice from DSS-induced colitis. BTG1 Tg mice were generated and treated with 2% DSS in drinking water to establish experimental colitis model. (A) Changes of body weight over a period of observation were shown as percentage of the initial weight at the start of the experiments. *P < .05 compared with DSStreated WT mice at the same time points. (B) Changes of the DAI. *P < .05 compared with DSS-treated WT mice at the same time points. (C) Gross morphology of colon of each group. (D) The length of colon from each group, **P < .01. (E) Histologic sections of colonic tissues in each group. Magnification: 200. (F) Pathologic scores of colonic tissues from all groups, **P < .01.

with expression of Claudin1 in IECs, leading to disruption of the epithelial barrier integrity in irritable bowel syndrome with diarrhea.42 TNF promotes MIR122A expression, which down-regulates occludin expression and induces the compromised tight junction.43 Moreover, epithelial vitamin D receptor signaling could alleviate colitis via suppressing IEC apoptosis, and MIR346 is able to inhibit vitamin D receptor expression in IEC from patients with IBD, thus, facilitating the development of intestinal inflammation.44 In this study, we demonstrated that IEC expression of MIR301A was increased in IBD patients. Furthermore, the mice deficient in MIR301A

1444 He et al

Gastroenterology Vol. 152, No. 6

BASIC AND TRANSLATIONAL AT

Figure 6. MIR301A–/– mice are resistant to AOM/DSS-induced CAC. (A) The levels of MIR301A in normal colonic biopsies from healthy donors (n ¼ 36), inflamed mucosa from A-CD (n ¼ 40) and A-UC patients (n ¼ 42), colonic tumor tissues from patients with CAC (n ¼ 6), and paired normal (NC) and colonic tumor tissues from patients with CRC (n ¼ 43) were analyzed by qRT-PCR and normalized to U6 expression, ***P < .001. (B) Changes of body weight were recorded during modeling in all mice induced by AOM/DSS. (C) An attenuated rectal prolapse in MIR301A–/– mice compared with WT mice. (D) Gross morphology of the colon on day 81 of AOM/DSS administration (upper panel), **P < .01. Colons were opened longitudinally, and the number of tumors was measured (lower panel), ***P < .001. (E) Representative images of colonic tumor tissues of WT and MIR301A–/– mice on day 81 of the experimental procedure. Magnification: 100. (F) Tumor sizes (left panel) and the tumor size distribution according to tumor numbers (right panel) were recorded. ***P < .001.

were resistant to colitis induced by DSS insults, which indicated a pathogenic role of MIR301A in development of colitis. Consistent with increased expression of MIR301A in patients with colorectal cancer (CRC), MIR301A–/– mice were resistant to the development of AOM/DSS-induced CAC, indicating that MIR301A also possibly regulates CRC development. Although our data suggested that MIR301A could potentially function as an oncogene by promoting proliferation of colon cancer cells, it is still unclear currently that MIR301A regulation of

tumorigenesis in AOM-DSS model is caused by a direct oncogenic effect of MIR301A or because of a consequence of reduced inflammation in MIR301A–/– mice. The epithelial tight junctions are essential for maintaining intestinal homeostasis, cell-cell adhesion, and apicalbasal polarity. Several tight junction proteins, including CDH1, an adherens junction protein, are down-regulated in both colitis and colorectal cancer.11,12 Our current data showed that MIR301A regulated IEC barrier function and permeability, as well as proinflammatory cytokine

MIR301A in IBD and CAC 1445

BASIC AND TRANSLATIONAL AT

May 2017

Figure 7. MIR301A is identified as a novel oncogene. Colonic mucosa and tumor tissues were obtained as described in Figure 6B. (A) Immunohistochemical staining for Ki67 (upper panels) and PCNA (lower panels) from colonic mucosa of WT and MIR301A–/– mice before AOM injection (day 0) and from tumor tissues of these mice on day 81 after AOM/DSS treatment. Magnification: 200. The percentage of Ki67þ and PCNAþ IEC per view by immunohistochemistry staining was calculated and presented as bar charts. *P < .05, **P < .01. (B and C) SW480 or HCT-116 cells were transduced with LV-anti-MIR301A or LV-anti-MIRctrl, and then cultured in DMEM supplemented with 10% FBS. (B) MTT assays were performed to determine cell growth (SW480, upper panel; HCT-116, lower panel). *P < .05 compared with cells transduced with LV-anti-MIRctrl. (C) Representative images of colony formation are recorded, and the numbers of colonies containing > 50 cells were scored and illustrated as bar charts. *P < .05 compared with cells transduced with LV-anti-MIRctrl. (D–F) SW480 cells transduced with LV-anti-MIR301A or LV-anti-MIRctrl were injected into the hind limbs of nude mice (n ¼ 6). (D) Sizes of tumors were measured at the indicated time points. *P < .05, **P < .01 compared with mice receiving cells transduced with LV-anti-MIRctrl. (E) Representative images of tumors collected 24 days post-implantation are shown. (F) MIR301A expression was examined in resected tumor tissues from mice transplanted with LV-anti-MIR301A- or LV-anti-MIRctrl-transduced SW480 cells, respectively. *P < .05 compared with mice receiving cells transduced with LV-anti-MIRctrl.

production, through promoting activation of NF-kB in IBD patients. MIR301A deficiency facilitated intestinal barrier function during experimental colitis, characterized by increased expression of CDH1 compared with WT mice, indicating that MIR301A regulates the pathogenesis of IBD

and CRC, possibly through moderating epithelial barrier function. Our studies identified BTG1 as a target gene of MIR301A to regulate IEC function. Currently, 6 members of the BTG family have been found including BTG1, BTG2, BTG3, BTG4,

1446 He et al

BASIC AND TRANSLATIONAL AT

Transducer of ErbB-2, and TOB2, which suppress the proliferation and cycle progression of tumor cells and inducing cell differentiation.31,45,46 BTGs are able to shuttle in nucleocytoplasmic counterparts because of their nuclear localization and export signals.47 BTG1 is highly expressed in the G0/G1 phases of the cell cycle, but markedly reduced throughout G1 phase. Moreover, BTG1 is also increased in apoptotic cells and enhances antisense Bcl-2-induced cytotoxic activities.48,49 Recently, an anti-tumor role of BTG1 has been reported that inhibits gastric cell proliferation, migration, and invasion,29,50 and positively correlates with increased expression levels of Cyclin D1 and Bax, also known as anti-tumor proteins.51 The levels of BTG1 are also decreased in kidney cancer and correlate with poor disease prognosis.52 In the present study, we found, for the first time, that BTG1 was significantly decreased in intestinal mucosa and IEC of patients with IBD and colorectal cancer. BTG1 inhibited NF-kB activation in IEC induced by IL1b. In addition, colitis was markedly mitigated in BTG1 transgenic mice after DSS insults compared with that in WT mice. These data thereby indicate that BTG1 might protect the gut from inflammation. In summary, we demonstrated that MIR301A expression was significantly increased in the IEC of patients with active IBD and CRC. IL1b promoted MIR301A expression through a c-Jun binding to MIR301A promoter pathway. Moreover, MIR301A induced production of several proinflammatory cytokines (eg, IL1b, IL6, IL8, and TNF) by IEC and compromised intestinal barrier through inhibiting BTG1, which was significantly decreased in colonic mucosa of patients with active CD and UC (Supplementary Figure 13). Because MIR301A supports both intestinal inflammation and neoplastic growth, possibly through inhibiting BTG1, both could potentially serve as targets for therapeutic treatment of patients with IBD and CRC. As great efforts have been made on development of MIR therapeutics, and some of which have already been in preclinical development stage, it will be interesting to see if blockade of MIR301A can prevent/ treat colitis in chronic models, as well as in CRC models, by delivering anti-MIR301A or related decoy transcripts. It will also be interesting to investigate if administration of BTG1 can prevent/treat colitis and CRC. If successful, both strategies will provide therapeutics for patients with IBD or CRC.

Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at www.gastrojournal.org, and at http://dx.doi.org/10.1053/ j.gastro.2017.01.049.

References 1. Cader MZ, Kaser A. Recent advances in inflammatory bowel disease: mucosal immune cells in intestinal inflammation. Gut 2013;62:1653–1664.

Gastroenterology Vol. 152, No. 6 2. Hecht GA. Intestinal microbes in health and disease. Introduction. Gastroenterology 2009;136:1849–1850. 3. Maloy KJ, Powrie F. Intestinal homeostasis and its breakdown in inflammatory bowel disease. Nature 2011; 474:298–306. 4. Ventham NT, Kennedy NA, Nimmo ER, et al. Beyond gene discovery in inflammatory bowel disease: the emerging role of epigenetics. Gastroenterology 2013; 145:293–308. 5. Smits LP, Bouter KE, de Vos WM, et al. Therapeutic potential of fecal microbiota transplantation. Gastroenterology 2013;145:946–953. 6. O’Toole PW, Jeffery IB. Gut microbiota and aging. Science 2015;350:1214–1215. 7. Maynard CL, Elson CO, Hatton RD, et al. Reciprocal interactions of the intestinal microbiota and immune system. Nature 2012;489:231–241. 8. Hogan SP, Seidu L, Blanchard C, et al. Resistin-like molecule beta regulates innate colonic function: barrier integrity and inflammation susceptibility. J Allergy Clin Immunol 2006;118:257–268. 9. Luissint AC, Parkos CA, Nusrat A. Inflammation and the intestinal barrier: leukocyte-epithelial cell interactions, cell junction remodeling, and mucosal repair. Gastroenterology 2016;151:616–632. 10. Neurath MF. Cytokines in inflammatory bowel disease. Nat Rev Immunol 2014;14:329–342. 11. Liu W, Chen Y, Golan MA, et al. Intestinal epithelial vitamin D receptor signaling inhibits experimental colitis. J Clin Invest 2013;123:3983–3996. 12. Rokavec M, Oner MG, Li H, et al. IL-6R/STAT3/miR-34a feedback loop promotes EMT-mediated colorectal cancer invasion and metastasis. J Clin Invest 2014; 124:1853–1867. 13. Kalla R, Ventham NT, Kennedy NA, et al. MicroRNAs: new players in IBD. Gut 2015;64:504–517. 14. O’Connell RM, Rao DS, Chaudhuri AA, et al. Physiological and pathological roles for microRNAs in the immune system. Nat Rev Immunol 2010;10:111–122. 15. Liu L, Nie J, Chen L, et al. The oncogenic role of microRNA-130a/301a/454 in human colorectal cancer via targeting Smad4 expression. PLoS One 2013; 8:e55532. 16. Lu Z, Li Y, Takwi A, et al. microRNA-301a as an NFkappaB activator in pancreatic cancer cells. EMBO J 2011;30:57–67. 17. Mycko MP, Cichalewska M, Machlanska A, et al. MicroRNA-301a regulation of a T-helper 17 immune response controls autoimmune demyelination. Proc Natl Acad Sci U S A 2012;109:E1248–E1257. 18. He C, Shi Y, Wu R, et al. miR-301a promotes intestinal mucosal inflammation through induction of IL-17A and TNF-alpha in IBD. Gut 2016;65:1938–1950. 19. Li D, Qiu Z, Shao Y, et al. Heritable gene targeting in the mouse and rat using a CRISPR-Cas system. Nat Biotechnol 2013;31:681–683. 20. Shao Y, Guan Y, Wang L, et al. CRISPR/Cas-mediated genome editing in the rat via direct injection of one-cell embryos. Nat Protoc 2014;9:2493–2512.

21. Xiao F, Deng J, Yu J, et al. A novel function of B-cell translocation gene 1 (BTG1) in the regulation of hepatic insulin sensitivity in mice via c-Jun. FASEB J 2016; 30:348–359. 22. Cao AT, Yao S, Gong B, et al. Th17 cells upregulate polymeric Ig receptor and intestinal IgA and contribute to intestinal homeostasis. J Immunol 2012;189: 4666–4673. 23. Wu W, He C, Liu C, et al. miR-10a inhibits dendritic cell activation and Th1/Th17 cell immune responses in IBD. Gut 2015;64:1755–1764. 24. Scheibe K, Backert I, Wirtz S, et al. IL-36R signalling activates intestinal epithelial cells and fibroblasts and promotes mucosal healing in vivo. Gut 2016 [Epub ahead of print]. http://dx.doi.org/10.1136/gutjnl-2015310374. 25. Deng Z, Mu J, Tseng M, et al. Enterobacteria-secreted particles induce production of exosome-like S1P-containing particles by intestinal epithelium to drive Th17-mediated tumorigenesis. Nat Commun 2015; 6:6956. 26. Chang B, Tessneer KL, McManus J, et al. Epsin is required for disheveled stability and Wnt signalling activation in colon cancer development. Nat Commun 2015; 6:6380. 27. Xavier RJ, Podolsky DK. Unravelling the pathogenesis of inflammatory bowel disease. Nature 2007;448:427–434. 28. Weber A, Wasiliew P, Kracht M. Interleukin-1 (IL-1) pathway. Sci Signal 2010;3:cm1. 29. Lee AS, Kranzusch PJ, Cate JH. eIF3 targets cell-proliferation messenger RNAs for translational activation or repression. Nature 2015;522:111–114. 30. Rouault JP, Rimokh R, Tessa C, et al. BTG1, a member of a new family of antiproliferative genes. EMBO J 1992; 11:1663–1670. 31. Waanders E, Scheijen B, van der Meer LT, et al. The origin and nature of tightly clustered BTG1 deletions in precursor B-cell acute lymphoblastic leukemia support a model of multiclonal evolution. PLoS Genet 2012; 8:e1002533. 32. Cho IJ, Lee AK, Lee SJ, et al. Repression by oxidative stress of iNOS and cytokine gene induction in macrophages results from AP-1 and NF-kappaB inhibition mediated by B cell translocation gene-1 activation. Free Radic Biol Med 2005;39:1523–1536. 33. Johansson ME, Gustafsson JK, Holmen-Larsson J, et al. Bacteria penetrate the normally impenetrable inner colon mucus layer in both murine colitis models and patients with ulcerative colitis. Gut 2014;63: 281–291. 34. Wang Z, Li R, Tan J, et al. Syndecan-1 acts in synergy with tight junction through Stat3 signaling to maintain intestinal mucosal barrier and prevent bacterial translocation. Inflamm Bowel Dis 2015; 21:1894–1907. 35. Barrett CW, Reddy VK, Short SP, et al. Selenoprotein P influences colitis-induced tumorigenesis by mediating stemness and oxidative damage. J Clin Invest 2015; 125:2646–2660.

MIR301A in IBD and CAC 1447 36. Nalle SC, Turner JR. Intestinal barrier loss as a critical pathogenic link between inflammatory bowel disease and graft-versus-host disease. Mucosal Immunol 2015; 8:720–730. 37. Hooper LV. Epithelial cell contributions to intestinal immunity. Adv Immunol 2015;126:129–172. 38. Turner JR. Intestinal mucosal barrier function in health and disease. Nat Rev Immunol 2009;9:799–809. 39. Su L, Nalle SC, Shen L, et al. TNFR2 activates MLCK-dependent tight junction dysregulation to cause apoptosis-mediated barrier loss and experimental colitis. Gastroenterology 2013;145:407–415. 40. Ordas I, Eckmann L, Talamini M, et al. Ulcerative colitis. Lancet 2012;380:1606–1619. 41. O’Neill LA, Sheedy FJ, McCoy CE. MicroRNAs: the fine-tuners of Toll-like receptor signalling. Nat Rev Immunol 2011;11:163–175. 42. Zhou Q, Costinean S, Croce CM, et al. MicroRNA 29 targets nuclear factor-kappaB-repressing factor and Claudin 1 to increase intestinal permeability. Gastroenterology 2015;148:158–169 e8. 43. Ye D, Guo S, Al-Sadi R, et al. MicroRNA regulation of intestinal epithelial tight junction permeability. Gastroenterology 2011;141:1323–1333. 44. Chen Y, Du J, Zhang Z, et al. MicroRNA-346 mediates tumor necrosis factor alpha-induced downregulation of gut epithelial vitamin D receptor in inflammatory bowel diseases. Inflamm Bowel Dis 2014;20: 1910–1918. 45. Ho KJ, Do NL, Otu HH, et al. Tob1 is a constitutively expressed repressor of liver regeneration. J Exp Med 2010;207:1197–1208. 46. Tijchon E, Havinga J, van Leeuwen FN, et al. B-lineage transcription factors and cooperating gene lesions required for leukemia development. Leukemia 2013; 27:541–552. 47. Kawamura-Tsuzuku J, Suzuki T, Yoshida Y, et al. Nuclear localization of Tob is important for regulation of its antiproliferative activity. Oncogene 2004;23: 6630–6638. 48. Corjay MH, Kearney MA, Munzer DA, et al. Antiproliferative gene BTG1 is highly expressed in apoptotic cells in macrophage-rich areas of advanced lesions in Watanabe heritable hyperlipidemic rabbit and human. Lab Invest 1998;78:847–858. 49. Nahta R, Yuan LX, Fiterman DJ, et al. B cell translocation gene 1 contributes to antisense Bcl-2-mediated apoptosis in breast cancer cells. Mol Cancer Ther 2006;5:1593–1601. 50. Li Y, Choi PS, Casey SC, et al. MYC through miR-17-92 suppresses specific target genes to maintain survival, autonomous proliferation, and a neoplastic state. Cancer Cell 2014;26:262–272. 51. Zheng HC, Li J, Shen DF, et al. BTG1 expression correlates with pathogenesis, aggressive behaviors and prognosis of gastric cancer: a potential target for gene therapy. Oncotarget 2015;6:19685–19705. 52. Sun G, Liu Q, Cheng Y, et al. B cell translocation gene 1 reduces the biological outcome of kidney cancer through

BASIC AND TRANSLATIONAL AT

May 2017

1448 He et al induction of cell proliferation, cell cycle arrest, cell apoptosis and cell metastasis. Int J Mol Med 2015; 35:777–783.

Author names in bold designate shared co-first authorship.

Gastroenterology Vol. 152, No. 6 Immunology, University of Texas Medical Branch, Galveston. e-mail: [email protected]. Acknowledgments The authors thank Drs Naizhong Hu (The Affiliated Hospital of Anhui Medical University, Hefei, China) and Bin Wu (Peking Union Medical College Hospital, Beijing, China) for collecting biopsies for CAC from IBD patients. The authors thank Dr Qing Wei (Department of Pathology, The Shanghai Tenth People’s Hospital) for expert histologic diagnosis.

Received September 21, 2016. Accepted January 24, 2017. Reprint requests Address requests for reprints to: Zhanju Liu, MD, PhD, Department of Gastroenterology, The Shanghai Tenth People’s Hospital, Tongji University, Shanghai 200072, China. e-mail: [email protected]; fax: þ86-21-66303983 or Yingzi Cong, PhD, Department of Microbiology and

Conflicts of interest The authors disclose no conflicts. Funding This work was funded by grants from the National Natural Science Foundation of China (nos. 81470822 and 81630017).

BASIC AND TRANSLATIONAL AT

May 2017

Supplementary Materials and Methods Antibodies and Reagents Dulbecco’s modified Eagle medium (DMEM), fetal bovine serum (FBS), penicillin, streptomycin, and L-gentamycin were all purchased from HyClone (Logan, UT). Recombinant IL1b, IL6, IL12, IL17A, IL23, IFNg, and TNF were purchased from BioLegend (San Diego, CA). All primary antibodies for immunoblotting were purchased from Abcam (Cambridge, UK), and horseradish peroxidase-conjugated anti-IgG was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal antibody to b-actin was purchased from Abcam. The inhibitor of JNK (JNK-IN-7) was purchased from Selleckchem (Shanghai, China).

Isolation of IEC IEC were isolated as described previously.1,2 Briefly, the colon was removed from the sacrificed mice, cut into 0.5-cm pieces, and placed in cold phosphate-buffered saline to remove debris. Colonic biopsies from IBD patients were obtained during endoscopic examination, and directly used as indicated. After incubating at 37 C for 202 minutes in phosphate-buffered saline with 2 mmol/L DTT and 1 mmol/L EDTA under gently shaking to isolate primary IEC. These cells were then collected and further purified via density gradient centrifugation with 20% and 40% percoll-RPMI solution.

Analysis of mRNA and MIR301A Expression Total RNA from primary IEC and epithelial cell lines was extracted using Trizol reagent (Invitrogen, Carlsbad, CA). RNA quantity and quality were assessed on a NanoVue spectrophotometer (GE Healthcare, Uppsala, Sweden), with a 260/280 ratio of > 1.8 and 28S/18S ratio of >1.4 for the majority of the samples. mRNA and MIR reverse transcription were performed using a 5All-In-One RT MasterMix kit (Applied Biological Materials Inc., Richmond, BC, Canada) and RT-PCR miRcute miRNA First-Strand cDNA Synthesis Kit (Tiangen Biotech, Beijing, China), respectively. For mRNA expression analysis, qRT-PCR was performed using a SYBR Green PCR kit (TaKaRa, Dalian, China), and the mRNA levels were normalized to the expression of GAPDH. MIR301A expression was detected by qRT-PCR using a miRcute miRNA qPCR Detection kit (Tiangen Biotech), and the relative expression of MIR301A was normalized to U6 expression. Each PCR amplification was performed in triplicate wells.

Generation of BM Chimeras BM cells were flushed and collected from the femurs of WT C57BL/B6 mice and transferred via intravenous injection into lethally irradiated (1100 Rad, split dose) MIR301A/ recipients (a total of 1107 cells/mouse). All chimeric mice were allowed to reconstitute for 4 weeks before DSS insult.

Luciferase Reporter Assay for MIR301A Target Dual-luciferase reporter vectors were constructed containing the predicted seed sequence in the 3’-UTR of

MIR301A in IBD and CAC 1448.e1

BTG1, as well as the corresponding mutant vectors in which 3 random nucleotide mutants were introduced into the seed sequences. The Renilla luciferase activity was determined as described previously.3,4

Cell Culture and Transduction All colon cancer cell lines used in this study were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) from the Central Laboratory for Medical Research at our hospital. These cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/ streptomycin in a CO2 incubator at 37 C. Transduction of colon cancer cell lines was performed using lentivirusencoding MIR301A (LV-MIR301A) or the empty vector (LV-MIRctrl) (GeneChem, Shanghai, China) in 6-well plates when cells grew at about 70% confluence according to the manufacturer’s instructions.

Immunohistochemical Staining To localize BTG1 expression, colonic sections were stained with rabbit polyclonal Ab against BTG1 (Abcam, Cambridge, UK) according to the manufacturer’s instructions. Briefly, colonic biopsies were fixed, embedded in paraffin, and then cut into slices. Slices were dewaxed and hydrated. After treatment of antigen retrieval, slices were incubated with Envision Flex Peroxidase-Blocking Reagent (Dako, Glostrup, Denmark) to block endogenous peroxidese. They were then incubated with goat serum before staining with anti-BTG1 monoclonal antibody (dilution 1:100) at 4 C overnight. After washing, slices were incubated with horseradish peroxidase-conjugated anti-IgG (dilution 1:400) for 1 hour at room temperature. Subsequently, slices were colored with diaminobenzidine in dark for 10 minutes. Slices were counterstained with hematoxylin. BTG1 expression was analyzed by optical microscopy.

BioPlex Enzyme-Linked Immunosorbent Assay A sandwich enzyme-linked immunosorbent assay (ELISA) was performed to analyze the phosphorylation status of serine 63 of c-Jun protein in IEC stimulated with IL1b. The data were analyzed with a BioPlex FlexMap3D analyzer (Bio-Rad Laboratories, Hercules, CA) using the BioPlex Manager software.

Apoptosis Assay For analyzing apoptosis of SW480 cells after transduction with LV-anti-MIR301A or LV-anti-MIRctrl, an Annexin V-fluorescein isothiocyanate (FITC) Apoptosis Detection Kit II (BD Biosciences, San Diego, CA) was used according to the manufacturer’s instructions. Briefly, SW480 cells were trypsinized, washed, and counted, followed by resuspended in Binding Buffer at a concentration of 1  106 cells/mL. Subsequently, cells were stained with FITC-conjugated Annexin V and PI for 15 minutes at room temperature. After adding binding buffer, flow cytometry data were acquired on a BD FACSCanto II and analyzed using FlowJo software (Tree Star, Inc, Ashland, OR).

1448.e2 He et al

References 1. Nguyen HT, Dalmasso G, Muller S, et al. Crohn’s disease-associated adherent invasive Escherichia coli modulate levels of microRNAs in intestinal epithelial cells to reduce autophagy. Gastroenterology 2014;146: 508–519. 2. Yan F, Cao H, Cover TL, et al. Colon-specific delivery of a probiotic-derived soluble protein ameliorates intestinal inflammation in mice through an

Gastroenterology Vol. 152, No. 6 EGFR-dependent mechanism. J Clin Invest 2011; 121:2242–2253. 3. Liu L, Nie J, Chen L, et al. The oncogenic role of microRNA-130a/301a/454 in human colorectal cancer via targeting Smad4 expression. PLoS One 2013; 8:e55532. 4. Xiao F, Deng J, Yu J, et al. A novel function of B-cell translocation gene 1 (BTG1) in the regulation of hepatic insulin sensitivity in mice via c-Jun. FASEB J 2016; 30:348–359.

MIR301A in IBD and CAC 1448.e3

20

∗∗∗

15 10

∗ ∗

5

6 IL

A 17

F

IL

TN

Nγ IF

23 IL

ed

IL

iu

1β

m

0

M

Relative MIR301A expression

May 2017

8

∗∗∗ ∗∗∗

6

∗∗ ∗∗ 4

∗

SW 0 62 0 H T2 9 Lo Vo C D - IE U C C - IE H C C - IE C

48

SW

11

6

0

C H

∗∗ ∗∗

2

T-

Relative MIR301A expression

Supplementary Figure 1. IL1b strongly induces MIR301A expression in human primary IEC. Primary IEC (1  106) were obtained from colonic biopsy samples of healthy donors (n ¼ 10), stimulated in vitro with IL1b (10 ng/mL), IL23 (10 ng/mL), IFNg (10 ng/mL), TNF (10 ng/mL), IL17A (10 ng/mL), and IL6 (20 ng/mL), respectively, for 6 hours, and expression of MIR301A was determined by qRT-PCR and normalized to U6 expression. *P < .05, ***P < .001 vs cells cultured in medium alone.

Supplementary Figure 2. MIR301A expression in epithelial cell lines. Expression of MIR301A was determined by qRTPCR in 5 colon cancer cell lines (ie, HCT-116, SW480, SW620, HT29, and LoVo; 1  106), as well as in primary IEC (1  106) from inflamed mucosa of patients with active CD (CD-IEC, n ¼ 10) or active UC (UC-IEC, n ¼ 10), and normal colonic mucosa of healthy controls (HC-IEC, n ¼ 10). *P < .05; **P < .01; ***P < .001 vs HC-IEC.

1448.e4 He et al

Gastroenterology Vol. 152, No. 6

15

15 HCT116 ∗∗∗

10

0

∗∗∗

10

∗∗

5

∗∗

5

0 1 5 10

0

IL1 β (ng/ml)

18

SW480

0

IL1 β (ng/ml)

20

∗∗∗

LoVo ∗∗∗

15 10

∗∗

∗∗

5 0

0 1 5 10 IL1 β (ng/ml)

D

c-Jun binding site of MIR301A motif

Bits

B

HT29

10

12 6

0 1 5 10

15

SW620 ∗∗∗

∗∗

5 0

0 1 5 10 IL1 β (ng/ml)

MIR301A/U6 relative expression

MIR301A/U6 relative expression

A

0 1 5 10 IL1 β (ng/ml)

15

∗∗∗

10

∗∗

5 0

0

5

10

IL1β (ng/ml)

p-cJun levels (fold change)

C

18

18

HCT116

12

∗∗

∗∗

6

0

0

0 1 5 10 IL1β (ng/ml)

MIR301A promoter activity (fold change)

E

MIR301A/U6 relative expression

SW620

0

24

HT29

25

∗∗

16

18 9

∗∗

∗∗

0 1 5 10

0 1 5 10

IL1 β (ng/ml)

IL1β (ng/ml)

0

∗

15

∗

10

8

∗∗

LoVo

20

∗∗

5

0 1 5 10

0

IL1β (ng/ml)

0 1 5 10 IL1β (ng/ml)

MIR301A promoter mutant MIR301A promoter

15

HCT116

15

∗∗

10

SW 480

∗

0

0

5

10

0

18 12 6

0 IL1β (10 ng/ml) JNK inhibitor

HCT116 ∗∗

12

#

− + + − − − + +

6 0

∗∗

5

10

0

SW480 ∗∗∗

− + + − − − + +

5

10

0

IL1β (ng/ml) 15 10

##

0

5 0

SW620 ∗∗∗

0

5

10

HT29 ∗∗

12 ##

− + + − − − + +

6 0

∗

3 0

IL1β (ng/ml) 18

∗∗∗

9 6

∗

5

LoVo

12 ∗∗

10

3

0

HT29

15 ∗∗

9

IL1β (ng/ml) 18

SW 620

6

∗

5

5

12 ∗∗

10

IL1β (ng/ml)

F

∗∗

12

6

27

SW480

##

− + + − − − + +

0

5

10

IL1β (ng/ml) 15

LoVo

10

∗∗∗

5 0

#

− + + − − − + +

May 2017

=

MIR301A in IBD and CAC 1448.e5

Supplementary Figure 3. IL1b promotes MIR301A expression through c-Jun activation. HCT-116, SW480, SW620, HT29, or LoVo cells were cultured in vitro under stimulation with IL1b at a series of doses (0, 1, 5, and 10 ng/mL) for 72 hours. (A) MIR301A expression was analyzed by qRT-PCR and normalized to U6 expression. **P < .01; ***P < .001 vs cells cultured in medium alone. (B) The LASAGNA-Search 2.0 database reveals the presence of c-Jun binding sites on the promoter of MIR301A gene. (C) Phosphorylated c-Jun (p-c-Jun) was detected by ELISA as described in the Methods. *P < .05; **P < .01 vs cells cultured in medium alone. (D) HCT-116 cells were treated with IL1b at different concentrations (0, 5, and 10 ng/mL) for 72 hours. Chromatin Immunoprecipitation (ChIP) analysis was performed according to the manufacturer’s instructions, and expression of MIR301A promoter was assessed by qRT-PCR. **P < .01; ***P < .001 vs cells cultured in medium alone. (E) HCT-116, SW480, SW620, HT29, and LoVo cells were treated with IL1b at different concentrations (0, 5, and 10 ng/mL) for 72 hours, and luciferase activity of MIR301A and mutant MIR301A gene promoter were then analyzed. *P < .01; **P < .01; ***P < .001 vs cells transfected with mutant MIR301A promoter in the same group. Data are shown as mean ± SEM. (F) HCT-116, SW480, SW620, HT29, and LoVo cells were cultured and stimulated with IL1b (10 ng/mL) or JNK inhibitor (JNK-IN-7, 1 mm) or a combination of both IL1b and JNK inhibitor for 72 hours, and MIR301A expression was analyzed by qRT-PCR. **P < .01; ***P < .001 vs cells cultured in medium alone. #P < .05; ##P < .01 vs cells cultured with IL1b alone.

Gastroenterology Vol. 152, No. 6

10

4

2

2

0

0 1A 30

LV

LV

-M

M

ed

iu

ct rl

1A

-M IR

LV

-M

IR

30

iu ed M

ct rl

4

m

6

-M IR

8

6

LV

SW480 ∗∗∗

IR

10 HCT-116 8 ∗∗∗

m

A

MIR301A/U6 relative expression

1448.e6 He et al

1.5

1.0

1.0

SW480

∗∗∗ 0.5

0.0

0.0

LV

-a

IR

30

iu nt i-M

ed -a

M LV

M

-a LV

∗∗∗

m

0.5

1 nt i-M A IR ct rl

1.5 HCT-116

ed nt i-M iu m LV IR 30 -a 1 nt i-M A IR ct rl

MIR301A/U6 relative expression

B

Supplementary Figure 4. The efficiency of lentivirus-mediated transduction. HCT-116 or SW480 cells were transduced with LV-MIR301A and LV-MIRctrl, respectively (A), or with LV-anti-MIR301A and LV-anti-MIRctrl, respectively (B), as described in the Material and Methods. Total RNA was isolated from these cells, and MIR301A expression was determined by qRT-PCR and normalized to the U6 expression. ***P < .001 vs cells cultured in medium alone.

IL

1β

IL

8

1A IR 30

1A

rl

m

Bcl-2

c-Myc

Bcl-xL

COX-2

β-actin

#

Bcl-2

c-Myc

Bcl-xL

COX-2

β-actin

S

#

LP

∗ 1β

HCT-116

IL

m

S

F

iu TN

ed

LP

1β

F

∗

LV -M

ct

IR

IL

TN

m

∗∗

30

0.0

IR

∗

rl

1.0

LV -M

F

i-M

LV-anti-MIR301A LV-anti-MIRctrl Medium

ct

E

IR

0

∗

nt

6

m

5

5

-a

IL

∗

i-M

F

∗ iu

D

15

LV

TN

∗ ed

LV-MIR301A LV-MIRctrl Medium

∗ 0

nt

M

C iu

0

m

∗ #

ed

2

p-NFκB levels (fold change)

3

iu

1.5

##

m

∗∗

B

-a

∗

8

∗ SW480

LV

0.5

∗∗

IL

1

1β

15

IL

4

6

F

10

HCT-116

ed

N

M LV ed -M iu IR m LV L 30 -a V-M 1A nt IR iLV MIR ctrl -a nti 301 -M A IR M ctrl LV ed -M iu IR m LV L 30 -a V-M 1A nt I LV i-M Rct -a IR3 rl nt i-M 01A IR ct rl

p-NFκB levels (fold change) 5

IL

T

Relative mRNA expressio n

A

M

Relative mRNA expressio n

May 2017 MIR301A in IBD and CAC 1448.e7

LV-anti-MIR301A LV-anti-MIRctrl

SW480

10 #

#

1448.e8 He et al

=

Gastroenterology Vol. 152, No. 6

Supplementary Figure 5. MIR301A enhances NF-kВ activation in IEC. (A) HCT-116 or SW480 cells were transduced with LV-MIR301A, LV-MIR-ctrl, LV-anti-MIR301A, and LV-anti-MIRctrl, respectively. Levels of p-NF-kВ were determined by ELISA as described in the Material and Methods. **P < .001 vs HCT-116 cells transduced with LV-MIRctrl. #P < .05; ##P < .01 vs SW480 cells transduced with LV-anti-MIRctrl. (B) HCT-116 or SW480 cells were transduced with LV-anti-MIR301A or LV-anti-MIRctrl, respectively, and then stimulated in vitro with TNF (10 ng/mL), IL1b (10 ng/mL), and LPS (1 mg/mL), respectively, for 72 hours. The levels of p-NF-kВ were determined by ELISA. *P < .05 vs HCT-116 cells transduced with LV-anti-MIRctrl. #P < .05 vs SW480 cells transduced with LV-anti-MIRctrl. (C and D) HCT-116 cells were transduced with LV-MIR301A and LV-MIRctrl, respectively, or cultured in medium alone. The mRNA levels of TNF, IL6, IL1b, and IL8 in HCT-116 cells were analyzed by qRT-PCR (C). *P < .05 vs cells transduced with LV-miR-ctrl. Protein levels of Bcl-2, c-Myc, Bcl-xl, and COX-2 were analyzed by Western blotting (D). (E and F) HCT-116 cells were transduced with LV-anti-MIR301A or LV-anti-MIRctrl, or cultured in medium alone. (E) The mRNA levels of TNF, IL6, IL1b, and IL8 in HCT116 cells were analyzed by qRT-PCR. *P < .05 vs cells transduced with LV-anti-MIRctrl. Protein levels of Bcl-2, c-Myc, Bcl-xl, and COX-2 were analyzed by Western blotting (F).

MIR301A in IBD and CAC 1448.e9

∗∗

0.5 0.0

1 G BT

E

m

pt y

1A -M IRc t rl

m

t rl

-M IR c

1A 30 IR

di u

m

0.0

LV

LV

∗

0.6

LV

∗∗

0.5

-M

1A -M IRc t rl

m

30 IR

di u

-M

LV

IR

30 1A -M IRc t rl

di u Me

LV

1.2

0.0

0.0 m

30 1A -M IR ctr l LV

15

∗∗∗ 10 5

T

−/

−

0

M

IR

30

1A

W

-M

Relative BTG1 expression

LV

∗∗

0.5

0.0

IR

di u

m

0.0

D

∗

0.5

1.0

1.0

Me

∗

0.5

-M

1.0

1.8 LoVo

30

1.5

1.0

ut m

1.5 HT29

LV

1.5

2.0 SW620

Me

1.5 SW480

LV

2.0 HCT-116

Me

Relative BTG1 expression

C

1 G BT

IR

BTG1 3’UTR

1.0

di u

MIR301A BTG1

.. ..

-M

..

Medium LV-MIRctrl LV-MIR301A

1.5

Me

B

LV

A

Relative Luciferase activity

May 2017

Supplementary Figure 6. MIR301A targets BTG1. (A) The predicted binding sites of MIR301A in the BTG1 3’-UTR. (B) The luciferase activity of the BTG1 or mutant BTG1 (BTG1 mut) 3’-UTR reporter in 293T cells transduced with LV-MIR301A or LV-MIRctrl (both at 100 nmol/L). **P < .01 vs cells transduced with LV-MIRctrl or cultured in medium alone in the same group. (C) HCT-116, SW480, SW620, HT29, and LoVo cells were transduced with LV-MIR301A and LV-MIRctrl, respectively, or cultured in medium alone. Expression of BTG1 was determined by qRT-PCR. *P < .05; **P < .01 vs cells transduced with LV-MIRctrl or cultured in medium alone. (D) The mRNA expression of BTG1 was analyzed in colonic tissues from both WT and MIR301A/ mice by qRT-PCR. ***P < .001 vs WT mice.

1448.e10 He et al

Gastroenterology Vol. 152, No. 6

HCT-116

A

∗∗ ∗∗∗ ∗∗

8

p-NFκB levels (fold change)

SW480

∗∗ ∗∗ ∗

6

∗∗

∗

4 2

p-c-Jun levels (fold change)

15

G1 BT

I β + L 1β BT G1

m di u

G1

Me

BT

G1

β

BT

IL 1

IL 1

B

IL 1

β+

Me

di u

m

0

HCT-116

SW480

∗∗ ∗∗∗ ∗∗

∗∗ ∗∗ ∗

10

∗∗

∗∗

5

G1 BT

IL 1 β BT G1 β+ IL 1

m di u

Me

G1 BT

IL 1 I β+ L1β BT G1

Me

di u

m

0

Supplementary Figure 7. MIR301A enhances NF-kB activation through inhibiting BTG1. HCT-116 or SW480 cells were cultured and stimulated with IL1b (10 ng/mL) or BTG1 plasmid (5 mg/100 mL Optimen) or a combination of both IL1b and BTG1 plasmid for 72 hours, and the levels of phosphorylated NF-kB (on S536) (A) and phosphorylated c-Jun (on S63) (B) were analyzed as described in the Material and Methods. *P < .05; **P < .01; ***P < .001.

Supplementary Figure 8. Genotyping of the MIR301A/ mice. Genotyping of the MIR301A mutant alleles was performed by PCR from murine genomic DNA. The PCR incubation sequence was set at 94 C for 4 minutes, followed by 30 cycles at 94 C for 40 s, 60 C for 40 s, 72 C for 60 s, and then a final extension at 72 C for 10 minutes; 10 mL of resulting products was digested with 0.1 mL T7 endonuclease I (T7 EI) for 20 minutes and detected using 1.5% agarose gel electrophoresis. The mutant alleles had lower molecular weight bands after digestion with T7 EI just as lane 1–3. Lane 4 is the WT mice, and lane M represents the DNA ladder (Cat. No. SM0331; Thermo Scientific, Waltham, MA). The PCR products of mutant alleles were sent for sequencing.

May 2017

MIR301A in IBD and CAC 1448.e11

B

4

BTG1

∗ T

MI

R3

01

W

A −/−

T W

A −/− 01 R3 MI

H2 O

DSS

R3

R3

MI

MI

DSS

∗

15 10

W

A −/− 01

T MI

MI

R3

R3

W

A −/−

0

T

∗

5

01

Relative BTG1 expression

H2 O

C

T

WT

A −/−

0

W

1

β-actin

01

2

A −/−

No detection

No detection

3

01

MIR301A/U6 relative expression

A

H2 O

DSS

Supplementary Figure 9. Expression of MIR301A and BTG1 in IEC from DSS-induced colitis in mice. Colitis was induced in WT and MIR301A/ mice as indicated in Figure 3A, and IEC were collected on day 10 for further analysis. (A) Expression of MIR301A was examined in colonic epithelial cells by qRT-PCR and normalized to U6 expression. *P < .05 vs WT mice treated with regular water alone. (B) Protein levels of BTG1 in IEC were assessed by Western blotting, and (C) expression of BTG1 mRNA was analyzed by qRT-PCR. *P < .05.

1448.e12 He et al

Gastroenterology Vol. 152, No. 6

H2 O

H2O

C

DSS

∗

4 3 2 0

H2O

WT BM ɠWT

1 WT BM ɠWT

WT BM ɠWT

∗

MIR301A−/− BM ɠWT

no inflammation

0

WT BM ɠWT

1

no inflammation

2

MIR301A−/−BM ɠWT

Pathological scores

3

WT BM ɠWT

ɠWT

MIR301A−/− BM ɠWT

WT BM ɠWT

ɠWT

B

DSS MIR301A−/− BM

IL17A mRNA relative expression

MIR301A−/− BM

MIR301A−/− BM ɠWT

A

DSS

Supplementary Figure 10. MIR301A deficiency in hemopoietic cells alleviates mucosal damage and IL17A expression during DSS-induced colitis. BM chimeric mice were generated in lethally irradiated WT mice reconstituted with WT or MIR301A/ BM cells, and experimental colitis was established in these chimeric mice with oral DSS treatment as indicated in Figure 4. (A) Histologic sections and (B) pathologic scores of colonic tissues in each group. (C) Expression of IL17A in colonic mucosa was determined by qRT-PCR. *P < .05.

LV

∗∗∗

SW480

CDH1 CDH1

β-actin β-actin

iu

1.0

0.5

m

- M 1A IR ct rl

30

ed

IR

M

-M LV

LV

200

- M 1A IR ct rl

2.0

30

m

0.0

300

IR

∗

-M

1.5

iu

HCT-116

ed

0

Percent of relative intensity

100

M

m - M 1A IR ct rl

30

iu

∗

CDH1 mRNA expression

LV

IR

ed

HCT-116

LV

LV

0.5

rl

1.0 0.8 0.6 0.4 0.2 0.0 - M 1A IR ct rl

1.0

ct

LV

-M

M

Percent of relative intensity 200

-M IR

1A

30

2.0

IR

m

LV

m

- M 1A IR ct rl

30

iu

300

30

HCT-116 -M

0 iu

5

ed

∗ CDH1 mRNA expression

SW480

IR

∗∗ LV

IR

ed

0

M

10

LV

m

15

30 - M 1A IR ct rl

IR

-M

M

50

-M

1.0 0.8 0.6 0.4 0.2 0.0

∗∗∗

m

0 100

iu

5

SW480

ed

∗∗

iu

LV

B

LV

10

ed

A

M

HCT-116 150

Ratio (CDH1/β-actin)

15

-M

M

0

TER (Ω*cm2)

50

LV

LV

rl

C

EH EC Translocation (log10 cfu)

ct

m

30 - M 1A IR ct rl

iu

∗∗∗

30

m

ed

IR

M -M

LV

LV

100

IR

iu

m

1A

iu

30

-M IR

IR

ed

HCT-116

1 -M A IR ct rl

-M

ed

LV

-M

M

TER (Ω*cm 2 ) 150

LV

LV

M

E

Ratio (CDH1/β-actin) LV

EH EC Trans location (log10 cfu)

May 2017 MIR301A in IBD and CAC 1448.e13

SW480

∗∗

100 0

D SW480

1.5

∗

0.0

1448.e14 He et al

Gastroenterology Vol. 152, No. 6

A

anti-MIR301A a t 30

0.1

0.1

0.2

PI

0.2

anti-MIRctrl

2.1

1.9

Annexin V 2.5 2.0 1.5 1.0 0.5 0.0 M IR tian

an

ti-

M

IR

30

ct

rl

1A

Percent of apoptotic cells

B

Supplementary Figure 12. MIR301A does not influence apoptotic activities in SW480 cells. SW480 cells were transduced with LV-anti-MIR301A or LV-anti-MIRctrl, and then cultured in DMEM supplemented with 10% FBS. (A) Cells were then harvested and stained with FITC-Annexin V/PI, and apoptotic cells were analyzed by flow cytometry. (B) Percentage of apoptotic cells was shown in a bar chart.

=

Supplementary Figure 11. MIR301A increases intestinal epithelial permeability. HCT-116 or SW480 cells were transduced with LV-MIR301A and LV-MIRctrl, respectively. Effect of exogenous MIR301A on TEER (A), the fluorescence intensity of a flux of FITC-labeled dextran (B), EHEC translocation (C), and CDH1 expression (D) was analyzed by qRT-PCR. (E) Protein levels of CDH1 in these cells were determined by Western blotting. *P < .05; **P < .01; ***P < .001 compared with cells transduced with LV-MIRctrl or cultured in medium alone.

May 2017

MIR301A in IBD and CAC 1448.e15

IL1β

c-Jun MIR301A

BTG1 IL1β IL6

NF-κB

Barrier injury

Cell proliferation

IL8

Inflammation

Tumorigenesis

Supplementary Figure 13. Schematic representation of the role of IEC-derived MIR301A molecular circuit in the pathogenesis of IBD and CRC. In patients with IBD or CRC, up-regulated IL1b activates c-Jun, which is able to bind to the promoter of MIR301A and initiate its transcription. Functionally, MIR301A directly inhibits its target gene, BTG1, through which it plays a crucial role in aggravating intestinal inflammation via impairing barrier function and activating NF-kB leading to increased expression of proinflammatory cytokines (eg, IL1b, IL6, IL8, TNF). Additionally, suppression of BTG1 results in promoting CRC cell proliferation and growth, conferring on MIR301A as a novel oncogene.

Supplementary Table 1.Clinical Characteristics of IBD Patients

No. of patients Age (y) Gender Male Female Disease duration (mo) Current therapy 5-aminosalicylates Immunosuppressants Biologics Nutritional therapy Disease extent (UC)a E1 E2 E3 Disease location (CD)a L1 L2 L3 L4 CRP (mg/L)

Con

CD (A/R)

UC (A/R)

35 36.7 ± 6.2

81 (45/36) 32.3 ± 10.4

72 (38/34) 36.1 ± 12.2

18 17

45 (25/17) 36 (20/16) 40.1 ± 21.5

39 (18/21) 33 (20/13) 43.8 ± 20.8

62 (39/23) 0 0 14 (14/0)

57 (31/26) 0 0 7 (7/0) 22 (10/12) 25 (15/10) 25 (13/12)

19 (8/11) 23 (14/9) 39 (23/16) 0 36.3 ± 16.1

37.5 ± 15.8

Abbreviations: A/R, active/remission; Con, healthy controls; CD, Crohn’s disease; UC, ulcerative colitis a According to the Montreal classification system.