Loss of MicroRNA-21 Influences the Gut Microbiota, Causing Reduced Susceptibility in a Murine Model of Colitis

Loss of MicroRNA-21 Influences the Gut Microbiota, Causing Reduced Susceptibility in a Murine... Abstract Background and Aims microRNAs regulate gene expression and influence the pathogenesis of human diseases. The present study investigated the role of microRNA-21 [miR-21] in the pathogenesis of intestinal inflammation, because miR-21 is highly expressed in inflammatory bowel disease. Inflammatory bowel disease is associated with intestinal barrier dysfunction and an altered gut microbiota. Recent studies have demonstrated that host microRNAs can shape the microbiota. Thus, we determined the influence of miR-21 on the gut microbiota and observed the subsequent impact in a dextran sodium sulphate [DSS]-induced colitis model. Methods The influence of miR-21 on the gut microbiota and inflammation was assessed in wild-type [WT] and miR-21–/– mice, in co-housed mice, following antibiotic depletion of the microbiota, or by colonization of germ-free [GF] mice with fecal homogenate, prior to DSS administration. We carried out 16S rRNA sequencing on WT and miR-21–/– mice to dissect potential differences in the gut microbiota. Results miR-21–/– mice have reduced susceptibility to DSS-induced colitis compared with WT mice. Co-housing conferred some protection to WT mice, while GF mice colonized with fecal homogenate from miR-21–/– were protected from DSS colitis compared with those colonized with WT homogenate. Further supporting a role for the microbiota in the observed phenotype, the protection afforded by miR-21 depletion was lost when mice were pre-treated with antibiotics. The 16S rRNA sequencing revealed significant differences in the composition of WT and miR-21–/– intestinal microbiota. Conclusions These findings suggest that miR-21 influences the pathogenesis of intestinal inflammation by causing propagation of a disrupted gut microbiota. MicroRNA, microbiota, inflammation 1. Introduction Inflammatory bowel disease [IBD], which includes Crohn’s disease [CD] and Ulcerative colitis [UC], is a lifelong chronic inflammation of the gastrointestinal tract. Inflammatory bowel disease is a complex and multifactorial disease of unknown etiology; however, it is generally believed to arise from a complex interaction between genetic factors, environmental factors and the host immune system. It is thought that abnormal interaction of the gut microbiota with the underlying mucosal immune system leads to an aberrant immune activation and chronic inflammation.1 Indeed, IBD has been shown to have an association with an altered gut microbiota [dysbiosis], and patients with IBD have an impaired intestinal barrier, which either contributes to the disease or arises as a result of inflammation, subsequently exacerbating the disease. Whether microbial dysbiosis causes impaired barrier function and development of inflammation or whether dysbiosis is a consequence of an altered barrier remains to be determined. Recently, there has been increasing evidence that this inflammation is associated with the altered expression of microRNA-21 [miR-21]. miRNAs are short ~22-nucleotide non-coding RNAs that act as post-transcriptional regulators of mRNA function by targeting specific RNAs for destruction or by repressing their translation.2–4 miRNAs regulate many cellular processes in response to a wide variety of stimuli, including processes associated with the progression of IBD, such as barrier function, mucin secretion, apoptosis and immune cell activation and function.5 However, although altered miRNA expression profiles have been identified in serum and tissue from IBD patients, the mechanistic basis underlying some of these associations is not yet understood.4 miR-21 is overexpressed in both IBD patients and experimental models of IBD.6–9 It has several reported functions that may impact on the disease, including roles in apoptosis and cytoskeletal rearrangement in the context of the barrier.8,9 It also has functions in both innate and adaptive immunity. In particular, miR-21 has been shown to modulate the responses of macrophages to bacterially derived Toll-like receptor [TLR] agonists, including lipopolysaccharides [LPSs].10 Recently, the role of miRNAs in interactions between the host and its microbiota has begun to emerge, with the miRNA profile being dependent on the presence of an endogenous microbiota; furthermore, host miRNAs have been shown to influence the activity of the gut microbiota.11–13 Given these observations and the ever-increasing evidence implicating the intestinal microbiota in the pathogenesis of IBD, we questioned whether miR-21 expression may influence the composition of the microbiota, and whether this was contributing to the pathogenesis of IBD. Here we demonstrated that miR-21 plays a pathological role in the development of intestinal inflammation. We showed that mice deficient in miR-21 [miR-21–/–] are less susceptible to DSS-induced colitis compared with wild-type [WT] mice. We demonstrated a protective microbiota in the miR-21–/– mice that contributed to the observed reduced susceptibility in the DSS-induced colitis model. Furthermore, using 16s rRNA sequencing analysis we confirmed alterations in the miR-21–/– microbial composition that were indicative of protection, including a reduced abundance of the phylum Bacteroidetes and an increased abundance of Firmicutes. Interestingly, analysis also revealed an increased abundance of the protective Clostridia classes upon loss of miR-21. Taken together, this work identified a novel role for miR-21 in shaping the gut microbiota and in the subsequent development of intestinal inflammation such as IBD. 2. Materials and Methods 2.1. Animals miR-21–/– mice were developed by Taconic Artemis using a Cre/lox approach. Briefly, miR-21 was modified by the insertion of two loxP sites that enable excision of the floxed miR-21 segment through Cre-mediated recombination. Chimeric offspring were backcrossed onto the C57BL/6J background for a total of eight generations. Homozygous deletion of miR-21 was confirmed by PCR genotyping. Homozygous miR-21–/– mice and WT littermates were used for animal studies. Animals were maintained in ventilated cages at 21 ± 1°C, humidity 50 ± 10% and with a 12 h-light/12 h-dark light cycle under specific pathogen-free conditions, in line with Irish and European Union regulations. Food and water were available ad libitum throughout all of the experiments. All experiments were subject to ethical approval by the Trinity College Dublin’s Animal Research Ethics Committee [AREC] and were carried out in accordance with the Irish Health Products Regulatory Authority. Germ-free [GF] animal experiments were carried out at the Weizmann Institute, Revolt, Israel, according to The Institutional Animal Care and Use Committee [IACUC]. 2.2. DSS-induced experimental colitis Colitis was induced in age- and sex-matched WT and miR-21–/– mice. Mice were weighed at the start of the trial. To induce acute experimental colitis, mice were administered 2.5% [w//v] dextran sodium sulphate [DSS, 36–50 kDa; MP Biomedicals, UK] in their drinking water ad libitum for 5 days [unless otherwise stated in the figure legend]. To assess experimental colitis and repair, a recovery model was implemented; specifically, mice were administered 2.5% [w/v] DSS in their drinking water ad libitum for 5 days followed by 5 days normal water. Control mice were given normal drinking water throughout. In all colitis models, mice were checked daily for morbidity and body weight was recorded. Each mouse was scored daily for pathological features, including stool consistency [diarrhea score], presence of fecal occult [blood score], and weight loss. Individual scores were combined to generate the Disease Activity Index [DAI], which was calculated daily for each mouse. The maximum DAI score was 12, based on assigning a 1–4 scoring system for each parameter: Score 0, no weight loss, normal stool and no blood; Score 1, 1–3% weight loss; Score 2, 3–6% weight loss, loose stool [a loose stool was defined as the formation of a stool that readily becomes paste upon handling] and blood visible in stool; Score 3, 6–9% weight loss; and Score 4, >9% weight loss, diarrhea and gross bleeding. At the end of the trial, mice were euthanized by CO2 asphyxiation. The colons were removed and measured as an indication of colonic inflammation. Subsequently, colonic sections were taken and stored in 10% formalin for histological analysis, or snap-frozen in liquid nitrogen and stored at –80°C. 2.3. Co-housing experiments Prior to induction of DSS experimental colitis, sex-matched WT and miR-21–/– mice aged 4 weeks were co-housed for a minimum of 4 weeks, to allow natural transfer of the fecal microbiota. 2.4. Colonization of germ-free mice with fecal microbiota A homogenate of frozen feces from WT or miR-21–/– mice was prepared in sterile PBS under anaerobic conditions and subsequently filtered through a sterile 70 μm strainer to remove soil particles. For colonization of GF mice, each mouse was administered 100 μL fecal suspension containing 20 mg feces by oral gavage. Mice were then allowed to rest for 9 days to ensure colonization. The same dose was used to mimic fecal microbial transplant in conventional mice. 2.5. Colonoscopy Colonoscopy was performed on experimental mice using a high‐resolution mouse video endoscopic system [Carl Storz, Tuttlingen, Germany]. The severity of colitis was scored in a blinded manner using MEICS [Murine Endoscopic Index of Colitis Severity], which is based on five parameters: granularity of mucosal surface, vascular pattern, translucency of the colon mucosa, visible fibrin, and stool consistency. 2.6. Antibiotic treatment of mice Mice were put on a 4-way antibiotic treatment course ad libitum for 2 weeks prior to DSS treatment. The course of treatment consisted of vancomycin [0.5 g/L], ampicillin [1 g/L], kanamycin [1 g/L] and metronidazole [1 g/L] in their drinking water. 2.7. Histology Sections from the distal colon of each mouse were fixed in 10% buffered formalin and analysed using haematoxylin and eosin [H&E] staining. H&E-stained sections of the distal colon were blind scored to ascertain the extent of colitis, using a previously established protocol.14 A combined score of inflammatory cell infiltration and tissue damage was determined as follows: (i) cell infiltration [Score 0, occasional inflammatory cells in the lamina propria [LP]; Score 1, increased infiltrate in the LP, predominantly at the base of crypts; Score 2, confluence of inflammatory infiltrate extending into the mucosa; Score 3, transmural extension of infiltrate]; (ii) tissue damage (Score 0, no mucosal damage; Score 1, partial [up to 50%] loss of crypts in large areas; Score 2, partial to total [50–100%] loss of crypts in large areas, epithelium intact; Score 3, total loss of crypts in large areas and epithelium lost). 2.9. qRT-PCR Total RNA was isolated from colonic sections using the RNeasy PlusMini kit [QIAGEN] and quantified using a Nanodrop 2000 UV-visible spectrophotometer. cDNA was prepared using 20–100 ng/mL total RNA by RT-PCR using a high capacity cDNA reverse transcription kit [Applied Biosystems], according to the manufacturer’s instructions. Real-time qPCR for miR-21 and IL-1β expression was performed on cDNA using Taqman probes. qPCR was performed on a 7900 HT Fast Real-Time PCR System [Applied Biosystems] using KAPA FAST Master Mix High ROX [Kapa Biosystems]. Fold changes in expression were calculated by the Delta Delta Ct method, using mouse Rps18 or RNU6B as an endogenous control for mRNA expression. 2.10. 16S sequencing and analysis To determine the gut microbiota composition of WT and miR-21–/– mice, 16S rRNA sequence analysis was performed on fecal samples from each cohort. Fecal samples were snap-frozen in liquid nitrogen, and DNA extraction was performed using a MOBIO PowerLyser DNA extraction kit according to the manufacturers’ instructions. The 16S rRNAgene [V3–V4 region] was PCR amplified, following the Illumina 16S Sample Preparation Guide, and the resulting amplicons were sequenced on the Illumina MiSeq platform using v3 sequencing chemistry with 2 × 250 bp paired-end reads. Sequences were further filtered on the basis of quality [removal of low-quality nucleotides at the 3' end, and removal of windows of 20 nucleotides with a low average quality] and length [removal of sequences with <200 nucleotides] with prinseq-lite, and joined using fastq-join [https://expressionanalysis.github.io/ea-utils/].15 The sequences were clustered with 97% identity level [calculated in operational taxonomic units; OTUs], using the closed-reference usearch v7.0 algorithm,16 and the cluster sequences assigned using the RDP’s Classifier.17 Alpha and Beta-diversity was determined using QIIME.18 Data were statistically analyzed by Adonis for beta-diversity analysis. Statistical differences between multiple samples were estimated using the Kruskal–Wallis test, and the False discovery rate [FDR, qvalue] control based on the Benjamini–Hochberg procedure was used to correct for multiple testing with the R statistical package [https://www.r-project.org/]. 2.11. Fecal sIgA detection Fecal pellets were taken, weighed and homogenized in protease inhibition buffer. Homogenate was centrifuged at 16 000g for 5 min at 4°C before IgA was measured using an IgA ELISA kit [BD Pharmingen] in accordance with the manufacturer’s instructions. 2.12. Myeloperoxidase activity assay Colon sections were snap-frozen in liquid nitrogen before homogenization, and myeloperoxidase [MPO] activity was assessed using a slightly modified version of a previously described assay.19 The modifications are as follows: 3,3',5,5'-tetramethylbenzidine [TMB] was used a substrate for the assay, and MPO activity was expressed in arbitrary units [AUs]. 2.13. Statistical analysis Numerical results are given as arithmetic means ± standard error of the means. Statistical differences were analysed by GraphPad Prism 5.0 statistical software [GraphPad Software Inc., San Diego, USA] to perform Student’s t-tests or the Chi squared test for the Kaplan–Meier survival curve. A p-value of <0.05 [p ≤ 0.05] was considered statistically significant. 3. Results 3.1. miR-21 deletion is protective in DSS-induced colitis We initially compared WT and miR-21–/– mice in an acute model of DSS-induced colitis. It was evident that miR-21–deficient mice were significantly protected from DSS-induced colitis, as evidenced by significantly reduced weight loss [Figure 1A], strikingly reduced levels of occult blood in the stool [Figure 1B] and a reduction in diarrhea score [Figure 1C]. Indeed, miR-21–/– mice had a significantly reduced DAI score [Figure 1D], indicative of reduced susceptibility to disease. The colons of miR-21–/– mice were also significantly longer [Figure 1E]. miR-21 expression was elevated in DSS-treated WT mice when compared with control animals [Figure 1F]. In a DSS-recovery model, in which mice received normal drinking water following 5 days treatment with 2.5% DSS, miR-21–/– mice again demonstrated protection from the DSS-induced colitis, recovering faster and showing reduced disease symptoms for each parameter [Figure 2A– D]. While there was no significant difference in colon length between WT and miR-21–/– mice treated with DSS, there was a significant difference between WT control mice and WT mice treated with DSS, which was absent between the miR-21–/– groups [Figure 2E]. Histological analysis was performed to assess the inflammation in the colon of WT and miR-21–/– mice post-DSS treatment. Control mice displayed no inflammation, whereas WT mice treated with DSS had significantly more inflamed colons than miR-21–/– mice, as assessed by a cumulative histology score [Figure 2F] generated from crypt damage score and inflammation score [Figure 2G and 2H]. Representative images show this difference in inflammation [Figure 2I]. Figure 1. View largeDownload slide Loss of miR-21 reduces susceptibility to DSS-induced colitis. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index score [DAI] [D]. On Day 5, the mice were sacrificed and their colons measured [E]. miR-21 expression in WT mice post-DSS treatment was compared with that in water controls [F]. [A–E] n = 5 mice per group, mean values ± SEM are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001. [F] miR-21 expression was assessed relative to RNU6B in mice treated with water [n = 2] or 2.5% DSS for 5 days [n = 4]. Mean values ± SD are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 1. View largeDownload slide Loss of miR-21 reduces susceptibility to DSS-induced colitis. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index score [DAI] [D]. On Day 5, the mice were sacrificed and their colons measured [E]. miR-21 expression in WT mice post-DSS treatment was compared with that in water controls [F]. [A–E] n = 5 mice per group, mean values ± SEM are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001. [F] miR-21 expression was assessed relative to RNU6B in mice treated with water [n = 2] or 2.5% DSS for 5 days [n = 4]. Mean values ± SD are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 2. View largeDownload slide View largeDownload slide miR-21–deficient mice were found to be protected compared with WT mice in a DSS-induced colitis recovery model. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days, followed by normal drinking water for a further 5 days. Control mice were given normal drinking water throughout. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index [DAI] score [D]. On Day 10, the mice were sacrificed and their colons measured [E]. Colitis severity of H&E-stained colonic tissues was assessed by the combined histological score [F] of tissue disruption [crypt damage score; 0–3, according to the severity of mucosal and crypts damages] [G] and colon cellular infiltration [inflammation score; 0–3, according to the extent of inflammation throughout the intestinal wall] [H]. Representative microscopic pictures of H&E-stained colon sections [×100 magnification] from water and 2.5% DSS–treated WT and miR-21–/– mice are shown [I]. Data represent mean values ± SEM of n = 3 mice per group [n = 2 in the miR-21–/– water control group]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 2. View largeDownload slide View largeDownload slide miR-21–deficient mice were found to be protected compared with WT mice in a DSS-induced colitis recovery model. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days, followed by normal drinking water for a further 5 days. Control mice were given normal drinking water throughout. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index [DAI] score [D]. On Day 10, the mice were sacrificed and their colons measured [E]. Colitis severity of H&E-stained colonic tissues was assessed by the combined histological score [F] of tissue disruption [crypt damage score; 0–3, according to the severity of mucosal and crypts damages] [G] and colon cellular infiltration [inflammation score; 0–3, according to the extent of inflammation throughout the intestinal wall] [H]. Representative microscopic pictures of H&E-stained colon sections [×100 magnification] from water and 2.5% DSS–treated WT and miR-21–/– mice are shown [I]. Data represent mean values ± SEM of n = 3 mice per group [n = 2 in the miR-21–/– water control group]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. 3.2. Co-housing confers protection to WT mice against DSS-induced colitis To determine any influence of miR-21 expression on the composition and activities of the gut microbiota and subsequent development of DSS-induced colitis, WT and miR-21–/– mice were co-housed for 4 weeks prior to DSS treatment. Control mice co-housed with mice of the same genotype displayed a similar disease phenotype to that observed in previous experiments [Figure 3A, B]. However, there was a slight reduction in disease severity in WT mice that had been co-housed with miR-21–/– [CH-WT], as indicated by a reduced DAI score compared with WT housed with other WT mice [Figure 3A, C]. Although there was no reduction in weight loss [Figure 3D], CH-WT had reduced diarrhea and blood scores compared with WT controls [Figure 3E, F]. Furthermore, the colon lengths of the CH-WT mice were longer than those of the WT controls [Figure 3G]. Although subtle, these results suggested that natural transfer of the fecal microbiota from miR-21–/– mice offers some protection against the severity of DSS-induced colitis. As such, the miR-21–/– fecal microbiota appears to display a less colitogenic phenotype compared with WT fecal microbiota, and can protect against development of colitis in recipient WT mice. We next attempted to determine whether the miR-21–/– fecal microbiota could also influence the course of established colitis [Supplementary figure S1]. To do this, WT mice were given DSS for 5 days prior to inoculation with either PBS [control] or a homogenate of miR-21–/– feces [prepared in the same way as for the GF colonization model]. On the day of inoculation, DSS was removed and mice received normal drinking water ad libitum for a further 5 days. Mice were scored daily throughout the model, as before, for disease parameters and DAI generated. No difference in the rates of recovery between control and miR-21–/– fecal homogenate– treated mice was observed, other than reduced shortening of colons in miR-21–/– fecal homogenate–treated mice [Supplementary figure S1E]. Figure 3. View largeDownload slide WT mice co-housed with miR-21–deficient mice were found to be protected from DSS-induced colitis. WT and miR-21–/– mice aged 4–6 weeks were co-housed for 4 weeks with either mice of the opposite genotype [e.g. WT mouse co-housed with miR-21–/– = CH-WT] or other mice of the same genotype [e.g. WT mouse co-housed with another WT = WT] before being weighed on Day 0 and given 3.5% DSS [w/v] ad libitum for 7 days. Weight, diarrhea score and blood score was determined daily and combined to generate a disease activity index [DAI] score [A]. Individual plots of selected groups’ DAI, weight loss, diarrhea score and blood score are shown [B–F]. On Day 7, the mice were sacrificed and their colons measured [G]. Data represent mean values ± SEM of n = 4 mice per group [except miR-21–/– control group, for which n = 3]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 3. View largeDownload slide WT mice co-housed with miR-21–deficient mice were found to be protected from DSS-induced colitis. WT and miR-21–/– mice aged 4–6 weeks were co-housed for 4 weeks with either mice of the opposite genotype [e.g. WT mouse co-housed with miR-21–/– = CH-WT] or other mice of the same genotype [e.g. WT mouse co-housed with another WT = WT] before being weighed on Day 0 and given 3.5% DSS [w/v] ad libitum for 7 days. Weight, diarrhea score and blood score was determined daily and combined to generate a disease activity index [DAI] score [A]. Individual plots of selected groups’ DAI, weight loss, diarrhea score and blood score are shown [B–F]. On Day 7, the mice were sacrificed and their colons measured [G]. Data represent mean values ± SEM of n = 4 mice per group [except miR-21–/– control group, for which n = 3]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. 3.3. Colonization of GF mice with miR-21 fecal microbiota offers protection against DSS-induced colitis Having observed protection being partially transferred to WT mice via co-housing, we wished to explore whether or not this effect could be seen using a GF setting. GF mice were colonized with fecal homogenates from WT or miR-21–/– animals for 9 days prior to DSS treatment. Upon DSS treatment, mice colonized with WT fecal homogenate exhibited significantly greater weight loss, significantly higher occult blood scores and higher diarrhea scores [Figure 4A–C]. This culminated in significantly higher DAI scores [Figure 4D]. Overall colitis severity was assessed by colonoscopy, where scores were combined from multiple disease parameters [Figure 4E]. Mice colonized with WT homogenate had a higher disease score than their miR-21–/– colonized counterparts [Figure 4E] and this can be seen in representative images [Figure 4F]. Interestingly, miR-21 expression is induced by DSS in the colons of GF mice [Figure 4G]. Figure 4. View largeDownload slide Colonization of germ-free [GF] mice with the fecal microbiota of miR-21–deficient mice protected them against DSS-induced colitis. Germ-free Swiss–Webster mice were colonized with fecal preparations from WT or miR-21–/– mice by oral gavage. Nine days later, mice were given 3% DSS [w/v] ad libitum for 13 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. On Day 13, the mice were anesthetized and underwent a colonoscopy using the ColoView system. Colons were blind scored for a number of clinical parameters that were combined to give a colitis severity score [E]. Representative images are shown [F]. Data are representative of two independent experiments, presented as mean values ± SEM of n = 5 mice per group. Student’s t-test was used to calculate the p values for A–E. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. miR-21 expression relative to RNU6B was assessed by qPCR in colonic tissue from GF mice given 1.5% DSS ad libitum for 10 days; data represent mean values ± SD of n = 3 mice per group. Student’s t-test was used to calculate the p values. A single asterisk indicates p < 0.05 [G]. Figure 4. View largeDownload slide Colonization of germ-free [GF] mice with the fecal microbiota of miR-21–deficient mice protected them against DSS-induced colitis. Germ-free Swiss–Webster mice were colonized with fecal preparations from WT or miR-21–/– mice by oral gavage. Nine days later, mice were given 3% DSS [w/v] ad libitum for 13 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. On Day 13, the mice were anesthetized and underwent a colonoscopy using the ColoView system. Colons were blind scored for a number of clinical parameters that were combined to give a colitis severity score [E]. Representative images are shown [F]. Data are representative of two independent experiments, presented as mean values ± SEM of n = 5 mice per group. Student’s t-test was used to calculate the p values for A–E. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. miR-21 expression relative to RNU6B was assessed by qPCR in colonic tissue from GF mice given 1.5% DSS ad libitum for 10 days; data represent mean values ± SD of n = 3 mice per group. Student’s t-test was used to calculate the p values. A single asterisk indicates p < 0.05 [G]. 3.4. Protection of miR-21–/– mice against DSS-induced colitis is lost after antibiotic treatment In order to ascertain the extent to which the intestinal microbiota was influencing the disease severities between the different groups, WT and miR-21–/– mice were put on a course of oral antibiotics prior to DSS treatment. WT and miR-21–/– control mice had different weight loss levels, blood and diarrhea scores, as seen before, with miR-21–/– mice again showing reduced susceptibility to DSS-induced colitis [Figure 5A–D]. However, interestingly, these differences between WT and miR-21–/– mice were lost following antibiotic pre-treatment, as can be seen in the DAI score [Figure 5A–D]. The difference in colon length between the WT and miR-21–/– mice was also lost following antibiotic pre-treatment [Figure 5E]. Strikingly, miR-21–/– mice displayed a significantly greater loss of mobility than their WT counterparts upon pre-treatment with antibiotics, and this led to a significantly higher morbidity in this group [Figure 5F and G]. Figure 5. View largeDownload slide Protection of miR-21–deficient mice against DSS-induced colitis is lost after antibiotic treatment. WT and miR-21–/– mice aged between 8 and 12 weeks were given a 4-way antibiotics mix ad libitum for 2 weeks. The mice were then weighed on Day 0 and given 3% DSS [w/v] ad libitum for 7 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. The mice were sacrificed on Day 8 and their colons measured [E]. In addition, the mice were scored for mobility loss [F] and morbidity [G]. Data represent mean values ± SEM pooled from two experiments of n = ≥5 mice per group. Student’s t-test was used to calculate the p values, and the Chi squared test was used to generate the Kaplan–Meier survival curve. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001 for comparisons between WT and miR-21–/–, except in panel G where they compare WT [ABX] and miR-21–/– [ABX]. Figure 5. View largeDownload slide Protection of miR-21–deficient mice against DSS-induced colitis is lost after antibiotic treatment. WT and miR-21–/– mice aged between 8 and 12 weeks were given a 4-way antibiotics mix ad libitum for 2 weeks. The mice were then weighed on Day 0 and given 3% DSS [w/v] ad libitum for 7 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. The mice were sacrificed on Day 8 and their colons measured [E]. In addition, the mice were scored for mobility loss [F] and morbidity [G]. Data represent mean values ± SEM pooled from two experiments of n = ≥5 mice per group. Student’s t-test was used to calculate the p values, and the Chi squared test was used to generate the Kaplan–Meier survival curve. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001 for comparisons between WT and miR-21–/–, except in panel G where they compare WT [ABX] and miR-21–/– [ABX]. 3.5. Altered composition of the miR-21–/– microbiota We employed 16S rRNA sequencing of fecal samples to investigate the possibility of differences in the composition of the microbiota between WT and miR-21–/– mice. After quality filtering and length trimming, an average of 37 313 [±7123 SD] 16S rRNA high-quality sequences were generated per sample. The average number of OTUs per sample was 355 [±16 SD]. The top 25 most abundant OTUs were compared between the two groups, with each OTU being given its phylum name followed by its genus name [e.g. Actinobacterium: Bifidobacterium] [Figure 6A]. There were considerable differences between the two groups, with the miR-21–/– group displaying a higher proportion of (i) Actinobacterium: Bifidobacterium; (ii) Firmicutes: Coprococcus; (iii) Firmicutes: Lactonifactor; (iv) Firmicutes; Oscilibacter; (v) Firmicutes: Clostridium sensu stricto; (vi) Firmicutes: Anaerosporobacter; and (vii) Firmicutes: Catonella; and a lower proportion of (i) Firmicutes: Clostridium XIVa; (ii) Firmicutes: Tannerella; (iii) Firmicutes: Dorea; (iv) Bacteroidetes: Barnesiella; and (v) Bacteroidetes: Prevotella, relative to WT mice [Figure 6A]. Beta diversity is represented by a Principal Coordinates Analysis [PCoA], performed using all 16S rRNA reads clustered at 97% similarity [Figure 6B]. Clear phylogenetic separation between the WT and miR-21 knockout groups was evident [ANOSIM p value = 0.001]. There were also several differences apparent at various taxonomic levels: at the phylum level, Proteobacteria [q = 0.00079935] were present in higher proportions in the miR-21–/– microbiota. At the family level, the miR-21–/– samples contained higher proportions of Bifidobacteriaceae [q = 0.01181564] and Peptostreptococcaceae [q = 0.14586726], and reduced proportions of Verrucomicrobiaceae [q = 0.07853655] and Bacteroidaceae [q = 0.05746273]. At the genus level, the WT samples group contained higher proportions of Enterorhabdus [q = 0.00036569], and reduced proportions of Odoribacter [q = 7.896e-05] and Bifidobacterium [q = 0.00143734] [Figure 6C–E]. Figure 6. View largeDownload slide View largeDownload slide 16S rRNA analysis of WT and miR-21–deficient mice fecal microbiota. The 16S rRNA reads were compared with the RDP database, and WT and miR-21–/– microbiota samples were compared for OTU analysis [A], beta diversity [B] and taxa analysis at phylum [C], family [D] and genus [E] levels. Data is representative of n = ≥9. Figure 6. View largeDownload slide View largeDownload slide 16S rRNA analysis of WT and miR-21–deficient mice fecal microbiota. The 16S rRNA reads were compared with the RDP database, and WT and miR-21–/– microbiota samples were compared for OTU analysis [A], beta diversity [B] and taxa analysis at phylum [C], family [D] and genus [E] levels. Data is representative of n = ≥9. 4. Discussion Over the past decade there has been an increasing appreciation of the crucial roles played by miRNAs in a wide variety of cellular processes. Among the biggest discoveries in that time was the role of miRNAs in inflammatory processes, in which they regulate both pro- and anti-inflammatory pathways.20 miR-21 is an extensively studied molecule with a known association with human IBD.6,8 With this in mind, we generated a knockout mouse to study the disease. Our initial finding that miR-21 was pathological in the DSS mouse model of colitis was surprising, as miR-21 is widely regarded as an anti-inflammatory mediator. These findings were supported by Shi et al., who described the same phenotype and suggested that miR-21 mechanistically influenced the disease by allowing the intestinal epithelial barrier to become more permeable via enhancing epithelial apoptosis and targeting the tight junction protein RhoB.6,8 These results indicate to us that miR-21 overexpression in IBD is unlikely to be simply a bystander effect caused by an inflammatory milieu, as has been suggested,21 but an active component of disease. Another field of study that has exploded in recent years is the growing understanding that the gut microbiota greatly influences human health and disease. Microbial dysbiosis has been found to play a significant contribution in a number of DSS-induced colitis phenotypes exhibited by transgenic mice with alterations in immune function. One such study demonstrated that the inflammasome component NLRP6 was required for intact host–microbe interaction, and that the exacerbated colitis phenotype exhibited by NLRP6–/– mice in response to treatment with DSS could be transferred to WT mice by co-housing.22,23 As miR-21 is a known regulator of the immunological activity of the bacterial sensor TLR4, we sought to ascertain the significance of the gut microbiota in our mice, and we reasoned that this might be a crucial aspect of the disease phenotype. First, we observed that, following co-housing of mice, WT mice co-housed with miR-21–/– mice displayed a reduced colitis phenotype when compared with the WT control group, which was evident in the fecal occult blood and in the diarrhea score. This indicated that the mice lacking miR-21 had an altered microbiota compared with their WT counterparts, which offered protection against the development of colitis; it also indicated that this protection was transferable. There is a sparse but growing literature regarding the potential of miRNAs to modulate the gut microbiota, but given the variety of roles in the regulation of bacterial sensing, it is likely that miRNAs play an indirect role. In one interesting study, conducted by Lui et al., the authors demonstrated that host miRNAs, including miR-21, are present in mouse feces and that they can be taken up by commensal bacteria to alter their gene expression. In addition, this study showed a significant microbial dysbiosis was present in Dicer knockout mice, which lack all functional miRNAs, and that this dysbiosis could be rescued by re-introducing synthetic miRNAs to the mice.11 To further investigate the protective microbiota observed in the miR-21–/– mice, we colonized GF mice with the fecal microbiota of WT or miR-21–/– mice prior to a DSS challenge. In this experiment, mice colonized with a miR-21–/– fecal microbiota exhibited protection from DSS-induced colitis relative to the mice colonized with the WT microbiota. A colonoscopy confirmed that the miR-21–/––colonized mice displayed fewer signs of inflammation. Colonoscopy is the current gold standard in IBD diagnosis, and it is becoming increasingly utilized in studies modelling the disease.24 It had been previously been reported that miR-21 expression was downregulated in the caecum of GF mice relative to that in conventional controls, implying that the caecal microbiota impacts on its expression.12 In order to establish whether miR-21 is induced in the DSS colitis model due to increased penetration of the epithelium by commensal bacteria and subsequently enhanced immune activation, miR-21 expression levels were compared between GF mice given DSS or water. It was observed that miR-21 expression remained enhanced in these mice upon DSS treatment. This was interesting and pointed to a mechanism of induction more related to the epithelial damage caused by DSS administration.25 As a final confirmation that the microbiota of miR-21–/– mice confers protection against DSS-induced colitis, we determined the consequence of depleting this microbiota by administering antibiotics to mice prior to challenge with DSS. Both WT-depleted and miR-21–/––depleted mice displayed increased detectable fecal occult blood and higher diarrhea scores earlier in the experiment. However, interestingly, the protective phenotype previously observed in the miR-21–/– mouse was lost following antibiotic treatment, as demonstrated in each of the disease scores and in the overall DAI score. Antibiotic depletion of the microbiota has been shown to increase intestinal haemorrhaging and mortality in mice subsequently treated with DSS.26,27 These effects have been demonstrated to be due to a weakened intestinal barrier and the absence of tonic microbial signals to the epithelium, respectively. Our results demonstrated an increase in haemorrhaging [as measured by fecal occult blood], in keeping with these studies. Most striking of all the observations in this experiment was the highly significant increase in sickness behaviour [measured by loss of mobility] and associated mortality displayed in the miR-21–/– mice treated with antibiotics. This effect has previously been described in WT mice at slightly later time points, but interestingly we saw mortality as early as Day 5 in the miR-21 –/– mice, indicating that in the absence of microbial signaling, miR-21 appears to play a protective role in this disease model. This finding confirms that the microbiota of miR-21–/– mice is protective in DSS-induced colitis, and that the protection demonstrated by the miR-21–/– mice is, at least in part, dependent on the presence of the microbiota. Having shown that the microbiota of the miR-21–/– mouse determined the outcome of DSS-induced colitis, 16S rRNA sequencing was employed to compare the fecal microbial profiles of miR-21–/– and WT mice. The top 25 most abundant OTUs in WT and miR-21–/– mice were compared and several significant differences were noted. Several of the observed differences pointed towards a potentially protective microbial signature in the miR-21–/– mouse, including a generally higher abundance of genera within the Firmicute phylum, which correlates with protection against IBD, while there was a lesser abundance of Bacteroidetes and Prevotella, which correlates with development of IBD.28,29 The higher abundance in the gram-positive probiotic genera Actinobacterium and Bifidobacterium can also be interpreted as evidence of a protective microbiota, as several studies have indicated a reduction in the severity of DSS-induced disease in mice treated with members of these genera.30–32 In order to assess whether or not these populations shared similar species, beta diversity was assessed and found to differ between WT and miR-21–/– fecal bacterial populations. This supported the case for an altered, protective microbiota being present in the miR-21–/– intestinal compartment. We next performed a taxa analysis to compare the relative proportions of the different bacterial groupings at various taxonomic levels. Surprisingly, at the phylum level there was an increased presence of Proteobacteria in the miR-21–/– mice, which would run counter to the idea of a protective microbiota. This phylum of gram-negative bacteria generally increase in proportion in IBD, with a commensurate decrease in Firmicutes.33,34 Further down the taxa, at family level the fecal microbiota of the mice lacking miR-21 display an increased proportion of Bifidobacteriaceae and Peptostreptococcaceae relative to the WT microbiota, and at genus level it displays an increase in Bifidobacterium and Odoribacter, all of which correlate with a healthy microbiota according to various studies.35,36 We examined the fecal microbiota, because this is routinely performed in the field, and also because the fecal microbiota was implicated in the co-housing and GF colonization studies that we carried out. However, it should be noted that the fecal microbiota might differ from that at individual mucosal surfaces throughout the gastrointestinal tract. This has been shown to be important in studies using microbial signatures as a biomarker for intestinal disease; thus, further investigation may be required in order to elucidate whether there are important differences in this model.37 Overall, these data confirmed that there was an altered microbiota in the miR-21–/– mouse, with some of these alterations being in accordance with previous studies indicating protective associations in relation to IBD and mouse models of colitis. An interesting aspect of this work worth considering is the potential alteration of other mediators present in the feces of miR-21–/– mice, which might impact the disease phenotype. These could include cytokines, antibodies, secreted epithelial factors or indeed other miRNAs, all of which have been demonstrated to be present in mouse feces and to have a transferable impact on disease outcomes.11 Secretory IgA [sIgA] has been shown to be an important mediator of microbial composition, a factor in microbial transferability, and a marker for susceptibility in IBD models.38,39 We measured sIgA levels in the feces of WT and miR-21–/– mice, and while there appeared to be reduced levels of sIgA in the feces of the miR-21–/– mice, this difference was not statistically significant [Supplementary figure S2]. We were unable to detect differences in fecal cytokines [data not shown]. We also measured the inflammatory markers MPO activity and IL-1β expression in all mice following DSS-induced colitis, but found little consistent evidence that these mechanisms may be influencing our observed phenotype [Supplementary figure S3]. This is somewhat in keeping with the counterintuitive nature of miR-21’s established role as an anti-inflammatory molecule in immune signaling.21 The mechanisms by which miR-21 may be influencing microbial composition have yet to be determined, but given that miR-21 has been demonstrated to be an important regulator of immune and epithelial barrier function, it is likely that modulation of these important mediators of gut homeostasis is involved. For instance, the enhanced barrier integrity displayed in miR-21–/– mice demonstrated in previous studies via increase in the expression of the miR-21 target RhoB may play a role in shaping the microbiota by creating a more favorable niche for colitis-protective bacteria.6,8 miRNAs have also been reported to alter the expression of specific mucins in the gut, which may also have a bearing on the niche occupied by these bacteria.40,41 It is also possible that miR-21 impacts the commensal bacteria more directly, by targeting secreted bactericidal mediators: for example, miR-21 has been shown to target CAMP and DEFB4A in humans.42 T-cells play an important role in immune homeostasis in the gut, and have a crucial role in shaping the gut microbiota.43 miR-21 has been demonstrated to impact the differentiation and activation of different T-cell subsets via modulation of key genes such as IL-12p35 and Stat3. For instance, in different inflammatory contexts, miR-21 has been shown to promote Th17 cells, to negatively regulate Treg cells and to promote memory T-cell activation.44–46 In murine models of colitis, miR-21 deficiency has been shown to skew T-cells towards Th1 responses.47 It would be interesting to explore the potential contribution of T-cell responses to shaping the microbiota in this system. These are a few of the possible mechanisms by which miR-21 may regulate the shape of the intestinal microbiota. It may be that one or all of them are involved in this system, and certainly it is likely that the altered barrier permeability demonstrated to be present in the miR-21–/– mouse by other groups is a factor. It is also possible that miR-21 may directly influence bacterial gene expression following secretion into the lumen in exosomes, as has recently been demonstrated.11 Clearly further studies are required in order to elucidate the pathways by which miR-21 may be exerting its influence on the microbiota. Another interesting aspect of the findings presented here are the implications of microbial alteration in the divergent phenotype displayed by miR-21–/– mice when subjected to the T-cell–dependent 2,4,6-trinitrobenzene sulfonic acid [TNBS] colitis and T-cell transfer colitis models. It has been previously reported that miR-21 deletion exacerbates colitis in these models through aberrant skewing of T-cell responses. As stated previously, T-cells help shape the microbiota, and the microbiota helps shape T-cell responses in inflammatory disease.43 Accordingly, it would be interesting to explore the microbial contribution to disease onset and progression in this model, and whether or not microbial manipulation [e.g. antibiotic depletion] might affect these phenotypes.47 There is a huge body of literature on miR-21’s role in many forms of cancer. miR-21 is upregulated in nearly all solid tumors and is a bone fide oncomiR, influencing many cancer-related processes such as cell division, apoptosis and metastasis.21,48 In the context of IBD, miR-21 is upregulated in colitis-associated cancer [CAC] and has also been shown to have a negative role in the azoxymethane [AOM]/DSS model of CAC, with miR-21–/– mice showing reduced susceptibility to the disease.7 Our results are interesting in this regard, because disruption to the microbiota has been reported to be involved in carcinogenesis through various mechanisms, including immune–microbiota crosstalk.49 This is particularly relevant in the case of CAC, because the breakdown of microbial tolerance and homeostasis in the gut leads to inflammation and subsequent tumorigenesis. It is therefore possible that miR-21’s impact on the intestinal microbiome may contribute to disease pathogenesis alongside its direct role in tumor formation and persistence. In our model, we have seen that miR-21 deletion gives rise to a microbiota that correlates with healthy individuals and protects from disease initiation, but does not remedy established disease [Supplemental figure S1]. Fecal microbial transplantation [FMT] therapy for IBD is in its infancy, with several studies reporting mixed efficacy for UC patients.50 A recent large study has shown efficacy using a multi-donor approach to maximize bacterial diversity, without identifying any particularly beneficial microbes.51 Future approaches for FMT and other microbial therapeutics may rely on a more detailed knowledge of which microbes are responsible for beneficial effects. Our findings are interesting in this context, because the mix of microbes and other factors present in the miR-21–/– mice have prophylactic value but appear less therapeutically beneficial. A deeper exploration of the composition of the transferable microbiota could provide valuable insights into this difference and inform future work in this area. As mentioned earlier, it would be very interesting to explore the microbial contribution to miR-21–/– disease phenotypes in other models, because this would further inform us of the potential efficacy of fecal transplant using a similar fecal bacterial signature. miRNAs are currently being targeted for therapeutic use in inflammatory disease. A recent study by Wang et al. has demonstrated that inhibition of the pro-inflammatory miRNA miR-223 can lessen disease severity in the DSS model of colitis.52 It would be interesting to see whether miR-21 inhibition would have a similar outcome in the DSS model. miR-21 inhibition is currently being developed for therapeutic use in a rare kidney disease, and has potential for use in IBD, though further work is required in order to tease out miR-21’s role in the various aspects of the disease [immune, microbiota, etc.], as highlighted by the divergence in its role in different models of disease.47,53 Our study highlights the need to consider the impact on the microbiota of any study carried out to this effect, as this may have an important bearing on the function of any such therapeutic intervention. In summary, our findings demonstrate that miR-21 expression is deleterious in the DSS model of IBD, and that the intestinal microbiota is crucial to the progression of this disease. Indeed, our findings identify a novel role for miR-21 in regulating the composition of the intestinal microbiota, and in the resulting development of intestinal inflammation. Funding This work was supported by a Starting Investigator Research Grant from Science Foundation Ireland [SFI] [grant number 11/SIRG/B2099] awarded to SCC. Conflict of Interest The authors have no competing interests to declare. Author Contributions SCC conceived ideas and oversaw the research programme. DGJ carried out the research. SCC and DGJ analysed data and wrote the manuscript. MAW and CM performed experiments. MR analysed histology. CT and EE assisted with germ-free experiments. RCR and PC assisted with analysis of 16S data. LAON funded creation of the miR-21–/– strain and provided invaluable guidance during the course of the project. Supplementary Data Supplementary data are available at ECCO-JCC Online. Acknowledgments The authors wish to acknowledge the technical assistance provided by the Comparative Medicine Unit, Trinity Biomedical Sciences Institute, in maintenance of animal colonies. References 1. Maloy KJ , Powrie F . Intestinal homeostasis and its breakdown in inflammatory bowel disease . Nature 2011 ; 474 : 298 – 306 . Google Scholar CrossRef Search ADS PubMed 2. Bartel DP . MicroRNAs: genomics, biogenesis, mechanism, and function . Cell 2004 ; 116 : 281 – 97 . Google Scholar CrossRef Search ADS PubMed 3. Bartel DP . MicroRNAs: target recognition and regulatory functions . Cell 2009 ; 136 : 215 – 33 . Google Scholar CrossRef Search ADS PubMed 4. Chapman CG , Pekow J . The emerging role of miRNAs in inflammatory bowel disease: a review . Therap Adv Gastroenterol 2015 ; 8 : 4 – 22 . Google Scholar CrossRef Search ADS PubMed 5. Carthew RW , Sontheimer EJ . Origins and mechanisms of miRNAs and siRNAs . Cell 2009 ; 136 : 642 – 55 . Google Scholar CrossRef Search ADS PubMed 6. Shi C , Liang Y , Yang J et al. MicroRNA-21 knockout improve the survival rate in DSS induced fatal colitis through protecting against inflammation and tissue injury . PLoS One 2013 ; 8 : e66814 . Google Scholar CrossRef Search ADS PubMed 7. Shi C , Yang Y , Xia Y et al. Novel evidence for an oncogenic role of microRNA-21 in colitis-associated colorectal cancer . Gut 2016 ; 65 : 1470 – 81 . Google Scholar CrossRef Search ADS PubMed 8. Yang Y , Ma Y , Shi C et al. Overexpression of miR-21 in patients with ulcerative colitis impairs intestinal epithelial barrier function through targeting the Rho GTPase RhoB . Biochem Biophys Res Commun 2013 ; 434 : 746 – 52 . Google Scholar CrossRef Search ADS PubMed 9. Zhang L , Shen J , Cheng J , Fan X . MicroRNA-21 regulates intestinal epithelial tight junction permeability . Cell Biochem Funct 2015 ; 33 : 235 – 40 . Google Scholar CrossRef Search ADS PubMed 10. Sheedy FJ , Palsson-McDermott E , Hennessy EJ et al. Negative regulation of TLR4 via targeting of the proinflammatory tumor suppressor PDCD4 by the microRNA miR-21 . Nat Immunol 2010 ; 11 : 141 – 7 . Google Scholar CrossRef Search ADS PubMed 11. Liu S , da Cunha AP , Rezende RM et al. The host shapes the gut microbiota via fecal microRNA . Cell Host Microbe 2016 ; 19 : 32 – 43 . Google Scholar CrossRef Search ADS PubMed 12. Singh N , Shirdel EA , Waldron L , Zhang RH , Jurisica I , Comelli EM . The murine caecal microRNA signature depends on the presence of the endogenous microbiota . Int J Biol Sci 2012 ; 8 : 171 – 86 . Google Scholar CrossRef Search ADS PubMed 13. Dalmasso G , Nguyen HT , Yan Y et al. Microbiota modulates host gene expression via MicroRNAs . Gastroenterology 2011 ; 140 : e19293 . 14. Smith P , Mangan NE , Walsh CM et al. Infection with a helminth parasite prevents experimental colitis via a macrophage-mediated mechanism . J Immunol 2007 ; 178 : 4557 – 66 . Google Scholar CrossRef Search ADS PubMed 15. Schmieder R , Edwards R . Quality control and preprocessing of metagenomic datasets . Bioinformatics 2011 ; 27 : 863 – 4 . Google Scholar CrossRef Search ADS PubMed 16. Edgar RC . Search and clustering orders of magnitude faster than BLAST . Bioinformatics 2010 ; 26 : 2460 – 1 . Google Scholar CrossRef Search ADS PubMed 17. Wang Q , Garrity GM , Tiedje JM , Cole JR . Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy . Appl Environ Microbiol 2007 ; 73 : 5261 – 7 . Google Scholar CrossRef Search ADS PubMed 18. Caporaso JG , Kuczynski J , Stombaugh J et al. QIIME allows analysis of high-throughput community sequencing data . Nat Meth 2010 ; 7 : 335 – 6 . Google Scholar CrossRef Search ADS 19. Bradley PP , Priebat DA , Christensen RD , Rothstein G . Measurement of cutaneous inflammation: estimation of neutrophil content with an enzyme marker . J Invest Dermatol 1982 ; 78 : 206 – 9 . Google Scholar CrossRef Search ADS PubMed 20. Baltimore D , Boldin MP , O’Connell RM , Rao DS , Taganov KD . MicroRNAs: new regulators of immune cell development and function . Nat Immunol 2008 ; 9 : 839 – 45 . Google Scholar CrossRef Search ADS PubMed 21. Sheedy FJ . Turning 21: induction of miR-21 as a key switch in the inflammatory response . Front Immunol 2015 ; 6 : 19 . Google Scholar CrossRef Search ADS PubMed 22. Elinav E , Strowig T , Kau AL et al. NLRP6 inflammasome regulates colonic microbial ecology and risk for colitis . Cell 2011 ; 145 : 745 – 57 . Google Scholar CrossRef Search ADS PubMed 23. Wlodarska M , Thaiss CA , Nowarski R et al. NLRP6 inflammasome orchestrates the colonic host–microbial interface by regulating goblet cell mucus secretion . Cell 2014 ; 156 : 1045 – 59 . Google Scholar CrossRef Search ADS PubMed 24. Beltzer A , Kaulisch T , Bluhmki T , Schoenberger T , Stierstorfer B , Stiller D . Evaluation of quantitative imaging biomarkers in the DSS colitis model . Mol Imaging Biol 2016 ; 18 : 697 – 704 . Google Scholar CrossRef Search ADS PubMed 25. Perše M , Cerar A . Dextran sodium sulphate colitis mouse model: traps and tricks . J Biomed Biotechnol 2012 ; 2012 : 718617 . Google Scholar CrossRef Search ADS PubMed 26. Rakoff-Nahoum S , Paglino J , Eslami-Varzaneh F , Edberg S , Medzhitov R . Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis . Cell 2004 ; 118 : 229 – 41 . Google Scholar CrossRef Search ADS PubMed 27. Hernández-Chirlaque C , Aranda CJ , Ocón B et al. Germ-free and antibiotic-treated mice are highly susceptible to epithelial injury in DSS colitis . J Crohns Colitis 2016 ; 10 : 1324 – 35 . Google Scholar CrossRef Search ADS PubMed 28. Bloom SM , Bijanki VN , Nava GM et al. Commensal bacteroides species induce colitis in host–genotype-specific fashion in a mouse model of inflammatory bowel disease . Cell Host Microbe 2011 ; 9 : 390 – 403 . Google Scholar CrossRef Search ADS PubMed 29. Natividad JM , Pinto-Sanchez MI , Galipeau HJ et al. Ecobiotherapy rich in firmicutes decreases susceptibility to colitis in a humanized gnotobiotic mouse model . Inflamm Bowel Dis 2015 ; 21 : 1883 – 93 . Google Scholar CrossRef Search ADS PubMed 30. Toumi R , Soufli I , Rafa H , Belkhelfa M , Biad A , Touil-Boukoffa C . Probiotic bacteria Lactobacillus and Bifidobacterium attenuate inflammation in dextran sulfate sodium–induced experimental colitis in mice . Int J Immunopathol Pharmacol 2014 ; 27 : 615 – 27 . Google Scholar CrossRef Search ADS PubMed 31. Osman N , Adawi D , Ahrne S , Jeppsson B , Molin G . Modulation of the effect of dextran sulfate sodium–induced acute colitis by the administration of different probiotic strains of Lactobacillus and Bifidobacterium . Dig Dis Sci 2004 ; 49 : 320 – 7 . Google Scholar CrossRef Search ADS PubMed 32. Srutkova D , Schwarzer M , Hudcovic T et al. Bifidobacterium longum CCM 7952 promotes epithelial barrier function and prevents acute DSS-induced colitis in strictly strain-specific manner . PLoS One 2015 ; 10 : e0134050 . Google Scholar CrossRef Search ADS PubMed 33. Mukhopadhya I , Hansen R , El-Omar EM , Hold GL . IBD—what role do Proteobacteria play ? Nat Rev Gastroenterol Hepatol 2012 ; 9 : 219 – 30 . Google Scholar CrossRef Search ADS PubMed 34. Matsuoka K , Kanai T . The gut microbiota and inflammatory bowel disease . Semin Immunopathol 2015 ; 37 : 47 – 55 . Google Scholar CrossRef Search ADS PubMed 35. Morgan XC , Tickle TL , Sokol H et al. Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment . Genome Biol 2012 ; 13 : R79 . Google Scholar CrossRef Search ADS PubMed 36. Nagy-Szakal D , Hollister EB , Luna RA et al. Cellulose supplementation early in life ameliorates colitis in adult mice . PLoS One 2013 ; 8 : e56685 . Google Scholar CrossRef Search ADS PubMed 37. Parthasarathy G , Chen J , Chen X et al. Relationship between microbiota of the colonic mucosa vs feces and symptoms, colonic transit, and methane production in female patients with chronic constipation . Gastroenterology 2016 ; 150 : 367 – 79.e1 . Google Scholar CrossRef Search ADS PubMed 38. Reinshagen M . [Immunoglobulin A coating identifies colitogenic bacteria in inflammatory bowel disease] . Z Gastroenterol 2015 ; 53 : 53 . Google Scholar CrossRef Search ADS PubMed 39. Palm NW , de Zoete MR , Cullen TW et al. Immunoglobulin A coating identifies colitogenic bacteria in inflammatory bowel disease . Cell 2014 ; 158 : 1000 – 10 . Google Scholar CrossRef Search ADS PubMed 40. Mo JS , Alam KJ , Kim HS , Lee YM , Yun KJ , Chae SC . MicroRNA 429 regulates mucin gene expression and secretion in murine model of colitis . J Crohns Colitis 2016 ; 10 : 837 – 49 . Google Scholar CrossRef Search ADS PubMed 41. Donaldson GP , Lee SM , Mazmanian SK . Gut biogeography of the bacterial microbiota . Nat Rev Microbiol 2016 ; 14 : 20 – 32 . Google Scholar CrossRef Search ADS PubMed 42. Liu PT , Wheelwright M , Teles R et al. MicroRNA-21 targets the vitamin D–dependent antimicrobial pathway in leprosy . Nat Med 2012 ; 18 : 267 – 73 . Google Scholar CrossRef Search ADS PubMed 43. Zhang H , Sparks JB , Karyala SV , Settlage R , Luo XM . Host adaptive immunity alters gut microbiota . ISME J 2015 ; 9 : 770 – 81 . Google Scholar CrossRef Search ADS PubMed 44. Li S , Fan Q , He S , Tang T , Liao Y , Xie J . MicroRNA-21 negatively regulates Treg cells through a TGF-β1/Smad-independent pathway in patients with coronary heart disease . Cell Physiol Biochem 2015 ; 37 : 866 – 78 . Google Scholar CrossRef Search ADS PubMed 45. Murugaiyan G , da Cunha AP , Ajay AK et al. MicroRNA-21 promotes Th17 differentiation and mediates experimental autoimmune encephalomyelitis . J Clin Invest 2015 ; 125 : 1069 – 80 . Google Scholar CrossRef Search ADS PubMed 46. Smigielska-Czepiel K , van den Berg A , Jellema P et al. Dual role of miR-21 in CD4+ T-cells: activation-induced miR-21 supports survival of memory T-cells and regulates CCR7 expression in naive T-cells . PLoS One 2013 ; 8 : e76217 . Google Scholar CrossRef Search ADS PubMed 47. Wu F , Dong F , Arendovich N , Zhang J , Huang Y , Kwon JH . Divergent influence of microRNA-21 deletion on murine colitis phenotypes . Inflamm Bowel Dis 2014 ; 20 : 1972 – 85 . Google Scholar CrossRef Search ADS PubMed 48. Krichevsky AM , Gabriely G . miR-21: a small multi-faceted RNA . J Cell Mol Med 2009 ; 13 : 39 – 53 . Google Scholar CrossRef Search ADS PubMed 49. Schwabe RF , Jobin C . The microbiome and cancer . Nat Rev Cancer 2013 ; 13 : 800 – 12 . Google Scholar CrossRef Search ADS PubMed 50. Lopez J , Grinspan A . Fecal microbiota transplantation for inflammatory bowel disease . Gastroenterol Hepatol 2016 ; 12 : 374 – 9 . 51. Paramsothy S , Kamm MA , Kaakoush NO et al. Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial . Lancet 2017 ; 389 : 1218 – 28 . Google Scholar CrossRef Search ADS PubMed 52. Wang H , Chao K , Ng SC et al. Pro-inflammatory miR-223 mediates the cross-talk between the IL23 pathway and the intestinal barrier in inflammatory bowel disease . Genome Biol 2016 ; 17 : 58 . Google Scholar CrossRef Search ADS PubMed 53. Gomez IG , MacKenna DA , Johnson BG et al. Anti-microRNA-21 oligonucleotides prevent Alport nephropathy progression by stimulating metabolic pathways . J Clin Invest 2015 ; 125 : 141 – 56 . Google Scholar CrossRef Search ADS PubMed Copyright © 2018 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Crohn's and Colitis Oxford University Press

Loss of MicroRNA-21 Influences the Gut Microbiota, Causing Reduced Susceptibility in a Murine Model of Colitis

Journal of Crohn's and Colitis , Volume Advance Article (7) – Mar 28, 2018

Loading next page...
 
/lp/ou_press/loss-of-microrna-21-influences-the-gut-microbiota-causing-reduced-sQeeYopZcD
Publisher
Oxford University Press
Copyright
Copyright © 2018 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1873-9946
eISSN
1876-4479
D.O.I.
10.1093/ecco-jcc/jjy038
Publisher site
See Article on Publisher Site

Abstract

Abstract Background and Aims microRNAs regulate gene expression and influence the pathogenesis of human diseases. The present study investigated the role of microRNA-21 [miR-21] in the pathogenesis of intestinal inflammation, because miR-21 is highly expressed in inflammatory bowel disease. Inflammatory bowel disease is associated with intestinal barrier dysfunction and an altered gut microbiota. Recent studies have demonstrated that host microRNAs can shape the microbiota. Thus, we determined the influence of miR-21 on the gut microbiota and observed the subsequent impact in a dextran sodium sulphate [DSS]-induced colitis model. Methods The influence of miR-21 on the gut microbiota and inflammation was assessed in wild-type [WT] and miR-21–/– mice, in co-housed mice, following antibiotic depletion of the microbiota, or by colonization of germ-free [GF] mice with fecal homogenate, prior to DSS administration. We carried out 16S rRNA sequencing on WT and miR-21–/– mice to dissect potential differences in the gut microbiota. Results miR-21–/– mice have reduced susceptibility to DSS-induced colitis compared with WT mice. Co-housing conferred some protection to WT mice, while GF mice colonized with fecal homogenate from miR-21–/– were protected from DSS colitis compared with those colonized with WT homogenate. Further supporting a role for the microbiota in the observed phenotype, the protection afforded by miR-21 depletion was lost when mice were pre-treated with antibiotics. The 16S rRNA sequencing revealed significant differences in the composition of WT and miR-21–/– intestinal microbiota. Conclusions These findings suggest that miR-21 influences the pathogenesis of intestinal inflammation by causing propagation of a disrupted gut microbiota. MicroRNA, microbiota, inflammation 1. Introduction Inflammatory bowel disease [IBD], which includes Crohn’s disease [CD] and Ulcerative colitis [UC], is a lifelong chronic inflammation of the gastrointestinal tract. Inflammatory bowel disease is a complex and multifactorial disease of unknown etiology; however, it is generally believed to arise from a complex interaction between genetic factors, environmental factors and the host immune system. It is thought that abnormal interaction of the gut microbiota with the underlying mucosal immune system leads to an aberrant immune activation and chronic inflammation.1 Indeed, IBD has been shown to have an association with an altered gut microbiota [dysbiosis], and patients with IBD have an impaired intestinal barrier, which either contributes to the disease or arises as a result of inflammation, subsequently exacerbating the disease. Whether microbial dysbiosis causes impaired barrier function and development of inflammation or whether dysbiosis is a consequence of an altered barrier remains to be determined. Recently, there has been increasing evidence that this inflammation is associated with the altered expression of microRNA-21 [miR-21]. miRNAs are short ~22-nucleotide non-coding RNAs that act as post-transcriptional regulators of mRNA function by targeting specific RNAs for destruction or by repressing their translation.2–4 miRNAs regulate many cellular processes in response to a wide variety of stimuli, including processes associated with the progression of IBD, such as barrier function, mucin secretion, apoptosis and immune cell activation and function.5 However, although altered miRNA expression profiles have been identified in serum and tissue from IBD patients, the mechanistic basis underlying some of these associations is not yet understood.4 miR-21 is overexpressed in both IBD patients and experimental models of IBD.6–9 It has several reported functions that may impact on the disease, including roles in apoptosis and cytoskeletal rearrangement in the context of the barrier.8,9 It also has functions in both innate and adaptive immunity. In particular, miR-21 has been shown to modulate the responses of macrophages to bacterially derived Toll-like receptor [TLR] agonists, including lipopolysaccharides [LPSs].10 Recently, the role of miRNAs in interactions between the host and its microbiota has begun to emerge, with the miRNA profile being dependent on the presence of an endogenous microbiota; furthermore, host miRNAs have been shown to influence the activity of the gut microbiota.11–13 Given these observations and the ever-increasing evidence implicating the intestinal microbiota in the pathogenesis of IBD, we questioned whether miR-21 expression may influence the composition of the microbiota, and whether this was contributing to the pathogenesis of IBD. Here we demonstrated that miR-21 plays a pathological role in the development of intestinal inflammation. We showed that mice deficient in miR-21 [miR-21–/–] are less susceptible to DSS-induced colitis compared with wild-type [WT] mice. We demonstrated a protective microbiota in the miR-21–/– mice that contributed to the observed reduced susceptibility in the DSS-induced colitis model. Furthermore, using 16s rRNA sequencing analysis we confirmed alterations in the miR-21–/– microbial composition that were indicative of protection, including a reduced abundance of the phylum Bacteroidetes and an increased abundance of Firmicutes. Interestingly, analysis also revealed an increased abundance of the protective Clostridia classes upon loss of miR-21. Taken together, this work identified a novel role for miR-21 in shaping the gut microbiota and in the subsequent development of intestinal inflammation such as IBD. 2. Materials and Methods 2.1. Animals miR-21–/– mice were developed by Taconic Artemis using a Cre/lox approach. Briefly, miR-21 was modified by the insertion of two loxP sites that enable excision of the floxed miR-21 segment through Cre-mediated recombination. Chimeric offspring were backcrossed onto the C57BL/6J background for a total of eight generations. Homozygous deletion of miR-21 was confirmed by PCR genotyping. Homozygous miR-21–/– mice and WT littermates were used for animal studies. Animals were maintained in ventilated cages at 21 ± 1°C, humidity 50 ± 10% and with a 12 h-light/12 h-dark light cycle under specific pathogen-free conditions, in line with Irish and European Union regulations. Food and water were available ad libitum throughout all of the experiments. All experiments were subject to ethical approval by the Trinity College Dublin’s Animal Research Ethics Committee [AREC] and were carried out in accordance with the Irish Health Products Regulatory Authority. Germ-free [GF] animal experiments were carried out at the Weizmann Institute, Revolt, Israel, according to The Institutional Animal Care and Use Committee [IACUC]. 2.2. DSS-induced experimental colitis Colitis was induced in age- and sex-matched WT and miR-21–/– mice. Mice were weighed at the start of the trial. To induce acute experimental colitis, mice were administered 2.5% [w//v] dextran sodium sulphate [DSS, 36–50 kDa; MP Biomedicals, UK] in their drinking water ad libitum for 5 days [unless otherwise stated in the figure legend]. To assess experimental colitis and repair, a recovery model was implemented; specifically, mice were administered 2.5% [w/v] DSS in their drinking water ad libitum for 5 days followed by 5 days normal water. Control mice were given normal drinking water throughout. In all colitis models, mice were checked daily for morbidity and body weight was recorded. Each mouse was scored daily for pathological features, including stool consistency [diarrhea score], presence of fecal occult [blood score], and weight loss. Individual scores were combined to generate the Disease Activity Index [DAI], which was calculated daily for each mouse. The maximum DAI score was 12, based on assigning a 1–4 scoring system for each parameter: Score 0, no weight loss, normal stool and no blood; Score 1, 1–3% weight loss; Score 2, 3–6% weight loss, loose stool [a loose stool was defined as the formation of a stool that readily becomes paste upon handling] and blood visible in stool; Score 3, 6–9% weight loss; and Score 4, >9% weight loss, diarrhea and gross bleeding. At the end of the trial, mice were euthanized by CO2 asphyxiation. The colons were removed and measured as an indication of colonic inflammation. Subsequently, colonic sections were taken and stored in 10% formalin for histological analysis, or snap-frozen in liquid nitrogen and stored at –80°C. 2.3. Co-housing experiments Prior to induction of DSS experimental colitis, sex-matched WT and miR-21–/– mice aged 4 weeks were co-housed for a minimum of 4 weeks, to allow natural transfer of the fecal microbiota. 2.4. Colonization of germ-free mice with fecal microbiota A homogenate of frozen feces from WT or miR-21–/– mice was prepared in sterile PBS under anaerobic conditions and subsequently filtered through a sterile 70 μm strainer to remove soil particles. For colonization of GF mice, each mouse was administered 100 μL fecal suspension containing 20 mg feces by oral gavage. Mice were then allowed to rest for 9 days to ensure colonization. The same dose was used to mimic fecal microbial transplant in conventional mice. 2.5. Colonoscopy Colonoscopy was performed on experimental mice using a high‐resolution mouse video endoscopic system [Carl Storz, Tuttlingen, Germany]. The severity of colitis was scored in a blinded manner using MEICS [Murine Endoscopic Index of Colitis Severity], which is based on five parameters: granularity of mucosal surface, vascular pattern, translucency of the colon mucosa, visible fibrin, and stool consistency. 2.6. Antibiotic treatment of mice Mice were put on a 4-way antibiotic treatment course ad libitum for 2 weeks prior to DSS treatment. The course of treatment consisted of vancomycin [0.5 g/L], ampicillin [1 g/L], kanamycin [1 g/L] and metronidazole [1 g/L] in their drinking water. 2.7. Histology Sections from the distal colon of each mouse were fixed in 10% buffered formalin and analysed using haematoxylin and eosin [H&E] staining. H&E-stained sections of the distal colon were blind scored to ascertain the extent of colitis, using a previously established protocol.14 A combined score of inflammatory cell infiltration and tissue damage was determined as follows: (i) cell infiltration [Score 0, occasional inflammatory cells in the lamina propria [LP]; Score 1, increased infiltrate in the LP, predominantly at the base of crypts; Score 2, confluence of inflammatory infiltrate extending into the mucosa; Score 3, transmural extension of infiltrate]; (ii) tissue damage (Score 0, no mucosal damage; Score 1, partial [up to 50%] loss of crypts in large areas; Score 2, partial to total [50–100%] loss of crypts in large areas, epithelium intact; Score 3, total loss of crypts in large areas and epithelium lost). 2.9. qRT-PCR Total RNA was isolated from colonic sections using the RNeasy PlusMini kit [QIAGEN] and quantified using a Nanodrop 2000 UV-visible spectrophotometer. cDNA was prepared using 20–100 ng/mL total RNA by RT-PCR using a high capacity cDNA reverse transcription kit [Applied Biosystems], according to the manufacturer’s instructions. Real-time qPCR for miR-21 and IL-1β expression was performed on cDNA using Taqman probes. qPCR was performed on a 7900 HT Fast Real-Time PCR System [Applied Biosystems] using KAPA FAST Master Mix High ROX [Kapa Biosystems]. Fold changes in expression were calculated by the Delta Delta Ct method, using mouse Rps18 or RNU6B as an endogenous control for mRNA expression. 2.10. 16S sequencing and analysis To determine the gut microbiota composition of WT and miR-21–/– mice, 16S rRNA sequence analysis was performed on fecal samples from each cohort. Fecal samples were snap-frozen in liquid nitrogen, and DNA extraction was performed using a MOBIO PowerLyser DNA extraction kit according to the manufacturers’ instructions. The 16S rRNAgene [V3–V4 region] was PCR amplified, following the Illumina 16S Sample Preparation Guide, and the resulting amplicons were sequenced on the Illumina MiSeq platform using v3 sequencing chemistry with 2 × 250 bp paired-end reads. Sequences were further filtered on the basis of quality [removal of low-quality nucleotides at the 3' end, and removal of windows of 20 nucleotides with a low average quality] and length [removal of sequences with <200 nucleotides] with prinseq-lite, and joined using fastq-join [https://expressionanalysis.github.io/ea-utils/].15 The sequences were clustered with 97% identity level [calculated in operational taxonomic units; OTUs], using the closed-reference usearch v7.0 algorithm,16 and the cluster sequences assigned using the RDP’s Classifier.17 Alpha and Beta-diversity was determined using QIIME.18 Data were statistically analyzed by Adonis for beta-diversity analysis. Statistical differences between multiple samples were estimated using the Kruskal–Wallis test, and the False discovery rate [FDR, qvalue] control based on the Benjamini–Hochberg procedure was used to correct for multiple testing with the R statistical package [https://www.r-project.org/]. 2.11. Fecal sIgA detection Fecal pellets were taken, weighed and homogenized in protease inhibition buffer. Homogenate was centrifuged at 16 000g for 5 min at 4°C before IgA was measured using an IgA ELISA kit [BD Pharmingen] in accordance with the manufacturer’s instructions. 2.12. Myeloperoxidase activity assay Colon sections were snap-frozen in liquid nitrogen before homogenization, and myeloperoxidase [MPO] activity was assessed using a slightly modified version of a previously described assay.19 The modifications are as follows: 3,3',5,5'-tetramethylbenzidine [TMB] was used a substrate for the assay, and MPO activity was expressed in arbitrary units [AUs]. 2.13. Statistical analysis Numerical results are given as arithmetic means ± standard error of the means. Statistical differences were analysed by GraphPad Prism 5.0 statistical software [GraphPad Software Inc., San Diego, USA] to perform Student’s t-tests or the Chi squared test for the Kaplan–Meier survival curve. A p-value of <0.05 [p ≤ 0.05] was considered statistically significant. 3. Results 3.1. miR-21 deletion is protective in DSS-induced colitis We initially compared WT and miR-21–/– mice in an acute model of DSS-induced colitis. It was evident that miR-21–deficient mice were significantly protected from DSS-induced colitis, as evidenced by significantly reduced weight loss [Figure 1A], strikingly reduced levels of occult blood in the stool [Figure 1B] and a reduction in diarrhea score [Figure 1C]. Indeed, miR-21–/– mice had a significantly reduced DAI score [Figure 1D], indicative of reduced susceptibility to disease. The colons of miR-21–/– mice were also significantly longer [Figure 1E]. miR-21 expression was elevated in DSS-treated WT mice when compared with control animals [Figure 1F]. In a DSS-recovery model, in which mice received normal drinking water following 5 days treatment with 2.5% DSS, miR-21–/– mice again demonstrated protection from the DSS-induced colitis, recovering faster and showing reduced disease symptoms for each parameter [Figure 2A– D]. While there was no significant difference in colon length between WT and miR-21–/– mice treated with DSS, there was a significant difference between WT control mice and WT mice treated with DSS, which was absent between the miR-21–/– groups [Figure 2E]. Histological analysis was performed to assess the inflammation in the colon of WT and miR-21–/– mice post-DSS treatment. Control mice displayed no inflammation, whereas WT mice treated with DSS had significantly more inflamed colons than miR-21–/– mice, as assessed by a cumulative histology score [Figure 2F] generated from crypt damage score and inflammation score [Figure 2G and 2H]. Representative images show this difference in inflammation [Figure 2I]. Figure 1. View largeDownload slide Loss of miR-21 reduces susceptibility to DSS-induced colitis. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index score [DAI] [D]. On Day 5, the mice were sacrificed and their colons measured [E]. miR-21 expression in WT mice post-DSS treatment was compared with that in water controls [F]. [A–E] n = 5 mice per group, mean values ± SEM are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001. [F] miR-21 expression was assessed relative to RNU6B in mice treated with water [n = 2] or 2.5% DSS for 5 days [n = 4]. Mean values ± SD are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 1. View largeDownload slide Loss of miR-21 reduces susceptibility to DSS-induced colitis. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index score [DAI] [D]. On Day 5, the mice were sacrificed and their colons measured [E]. miR-21 expression in WT mice post-DSS treatment was compared with that in water controls [F]. [A–E] n = 5 mice per group, mean values ± SEM are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001. [F] miR-21 expression was assessed relative to RNU6B in mice treated with water [n = 2] or 2.5% DSS for 5 days [n = 4]. Mean values ± SD are presented; p values were calculated using an unpaired t-test. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 2. View largeDownload slide View largeDownload slide miR-21–deficient mice were found to be protected compared with WT mice in a DSS-induced colitis recovery model. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days, followed by normal drinking water for a further 5 days. Control mice were given normal drinking water throughout. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index [DAI] score [D]. On Day 10, the mice were sacrificed and their colons measured [E]. Colitis severity of H&E-stained colonic tissues was assessed by the combined histological score [F] of tissue disruption [crypt damage score; 0–3, according to the severity of mucosal and crypts damages] [G] and colon cellular infiltration [inflammation score; 0–3, according to the extent of inflammation throughout the intestinal wall] [H]. Representative microscopic pictures of H&E-stained colon sections [×100 magnification] from water and 2.5% DSS–treated WT and miR-21–/– mice are shown [I]. Data represent mean values ± SEM of n = 3 mice per group [n = 2 in the miR-21–/– water control group]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 2. View largeDownload slide View largeDownload slide miR-21–deficient mice were found to be protected compared with WT mice in a DSS-induced colitis recovery model. WT and miR-21–/– mice aged between 8 and 12 weeks were weighed on Day 0 and given 2.5% DSS [w/v] ad libitum for 5 days, followed by normal drinking water for a further 5 days. Control mice were given normal drinking water throughout. Weight [A], blood score [B] and diarrhea score [C] were determined daily throughout the experiment and combined to generate a disease activity index [DAI] score [D]. On Day 10, the mice were sacrificed and their colons measured [E]. Colitis severity of H&E-stained colonic tissues was assessed by the combined histological score [F] of tissue disruption [crypt damage score; 0–3, according to the severity of mucosal and crypts damages] [G] and colon cellular infiltration [inflammation score; 0–3, according to the extent of inflammation throughout the intestinal wall] [H]. Representative microscopic pictures of H&E-stained colon sections [×100 magnification] from water and 2.5% DSS–treated WT and miR-21–/– mice are shown [I]. Data represent mean values ± SEM of n = 3 mice per group [n = 2 in the miR-21–/– water control group]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. 3.2. Co-housing confers protection to WT mice against DSS-induced colitis To determine any influence of miR-21 expression on the composition and activities of the gut microbiota and subsequent development of DSS-induced colitis, WT and miR-21–/– mice were co-housed for 4 weeks prior to DSS treatment. Control mice co-housed with mice of the same genotype displayed a similar disease phenotype to that observed in previous experiments [Figure 3A, B]. However, there was a slight reduction in disease severity in WT mice that had been co-housed with miR-21–/– [CH-WT], as indicated by a reduced DAI score compared with WT housed with other WT mice [Figure 3A, C]. Although there was no reduction in weight loss [Figure 3D], CH-WT had reduced diarrhea and blood scores compared with WT controls [Figure 3E, F]. Furthermore, the colon lengths of the CH-WT mice were longer than those of the WT controls [Figure 3G]. Although subtle, these results suggested that natural transfer of the fecal microbiota from miR-21–/– mice offers some protection against the severity of DSS-induced colitis. As such, the miR-21–/– fecal microbiota appears to display a less colitogenic phenotype compared with WT fecal microbiota, and can protect against development of colitis in recipient WT mice. We next attempted to determine whether the miR-21–/– fecal microbiota could also influence the course of established colitis [Supplementary figure S1]. To do this, WT mice were given DSS for 5 days prior to inoculation with either PBS [control] or a homogenate of miR-21–/– feces [prepared in the same way as for the GF colonization model]. On the day of inoculation, DSS was removed and mice received normal drinking water ad libitum for a further 5 days. Mice were scored daily throughout the model, as before, for disease parameters and DAI generated. No difference in the rates of recovery between control and miR-21–/– fecal homogenate– treated mice was observed, other than reduced shortening of colons in miR-21–/– fecal homogenate–treated mice [Supplementary figure S1E]. Figure 3. View largeDownload slide WT mice co-housed with miR-21–deficient mice were found to be protected from DSS-induced colitis. WT and miR-21–/– mice aged 4–6 weeks were co-housed for 4 weeks with either mice of the opposite genotype [e.g. WT mouse co-housed with miR-21–/– = CH-WT] or other mice of the same genotype [e.g. WT mouse co-housed with another WT = WT] before being weighed on Day 0 and given 3.5% DSS [w/v] ad libitum for 7 days. Weight, diarrhea score and blood score was determined daily and combined to generate a disease activity index [DAI] score [A]. Individual plots of selected groups’ DAI, weight loss, diarrhea score and blood score are shown [B–F]. On Day 7, the mice were sacrificed and their colons measured [G]. Data represent mean values ± SEM of n = 4 mice per group [except miR-21–/– control group, for which n = 3]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. Figure 3. View largeDownload slide WT mice co-housed with miR-21–deficient mice were found to be protected from DSS-induced colitis. WT and miR-21–/– mice aged 4–6 weeks were co-housed for 4 weeks with either mice of the opposite genotype [e.g. WT mouse co-housed with miR-21–/– = CH-WT] or other mice of the same genotype [e.g. WT mouse co-housed with another WT = WT] before being weighed on Day 0 and given 3.5% DSS [w/v] ad libitum for 7 days. Weight, diarrhea score and blood score was determined daily and combined to generate a disease activity index [DAI] score [A]. Individual plots of selected groups’ DAI, weight loss, diarrhea score and blood score are shown [B–F]. On Day 7, the mice were sacrificed and their colons measured [G]. Data represent mean values ± SEM of n = 4 mice per group [except miR-21–/– control group, for which n = 3]. Student’s t-test was used to calculate p values. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. 3.3. Colonization of GF mice with miR-21 fecal microbiota offers protection against DSS-induced colitis Having observed protection being partially transferred to WT mice via co-housing, we wished to explore whether or not this effect could be seen using a GF setting. GF mice were colonized with fecal homogenates from WT or miR-21–/– animals for 9 days prior to DSS treatment. Upon DSS treatment, mice colonized with WT fecal homogenate exhibited significantly greater weight loss, significantly higher occult blood scores and higher diarrhea scores [Figure 4A–C]. This culminated in significantly higher DAI scores [Figure 4D]. Overall colitis severity was assessed by colonoscopy, where scores were combined from multiple disease parameters [Figure 4E]. Mice colonized with WT homogenate had a higher disease score than their miR-21–/– colonized counterparts [Figure 4E] and this can be seen in representative images [Figure 4F]. Interestingly, miR-21 expression is induced by DSS in the colons of GF mice [Figure 4G]. Figure 4. View largeDownload slide Colonization of germ-free [GF] mice with the fecal microbiota of miR-21–deficient mice protected them against DSS-induced colitis. Germ-free Swiss–Webster mice were colonized with fecal preparations from WT or miR-21–/– mice by oral gavage. Nine days later, mice were given 3% DSS [w/v] ad libitum for 13 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. On Day 13, the mice were anesthetized and underwent a colonoscopy using the ColoView system. Colons were blind scored for a number of clinical parameters that were combined to give a colitis severity score [E]. Representative images are shown [F]. Data are representative of two independent experiments, presented as mean values ± SEM of n = 5 mice per group. Student’s t-test was used to calculate the p values for A–E. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. miR-21 expression relative to RNU6B was assessed by qPCR in colonic tissue from GF mice given 1.5% DSS ad libitum for 10 days; data represent mean values ± SD of n = 3 mice per group. Student’s t-test was used to calculate the p values. A single asterisk indicates p < 0.05 [G]. Figure 4. View largeDownload slide Colonization of germ-free [GF] mice with the fecal microbiota of miR-21–deficient mice protected them against DSS-induced colitis. Germ-free Swiss–Webster mice were colonized with fecal preparations from WT or miR-21–/– mice by oral gavage. Nine days later, mice were given 3% DSS [w/v] ad libitum for 13 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. On Day 13, the mice were anesthetized and underwent a colonoscopy using the ColoView system. Colons were blind scored for a number of clinical parameters that were combined to give a colitis severity score [E]. Representative images are shown [F]. Data are representative of two independent experiments, presented as mean values ± SEM of n = 5 mice per group. Student’s t-test was used to calculate the p values for A–E. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01. miR-21 expression relative to RNU6B was assessed by qPCR in colonic tissue from GF mice given 1.5% DSS ad libitum for 10 days; data represent mean values ± SD of n = 3 mice per group. Student’s t-test was used to calculate the p values. A single asterisk indicates p < 0.05 [G]. 3.4. Protection of miR-21–/– mice against DSS-induced colitis is lost after antibiotic treatment In order to ascertain the extent to which the intestinal microbiota was influencing the disease severities between the different groups, WT and miR-21–/– mice were put on a course of oral antibiotics prior to DSS treatment. WT and miR-21–/– control mice had different weight loss levels, blood and diarrhea scores, as seen before, with miR-21–/– mice again showing reduced susceptibility to DSS-induced colitis [Figure 5A–D]. However, interestingly, these differences between WT and miR-21–/– mice were lost following antibiotic pre-treatment, as can be seen in the DAI score [Figure 5A–D]. The difference in colon length between the WT and miR-21–/– mice was also lost following antibiotic pre-treatment [Figure 5E]. Strikingly, miR-21–/– mice displayed a significantly greater loss of mobility than their WT counterparts upon pre-treatment with antibiotics, and this led to a significantly higher morbidity in this group [Figure 5F and G]. Figure 5. View largeDownload slide Protection of miR-21–deficient mice against DSS-induced colitis is lost after antibiotic treatment. WT and miR-21–/– mice aged between 8 and 12 weeks were given a 4-way antibiotics mix ad libitum for 2 weeks. The mice were then weighed on Day 0 and given 3% DSS [w/v] ad libitum for 7 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. The mice were sacrificed on Day 8 and their colons measured [E]. In addition, the mice were scored for mobility loss [F] and morbidity [G]. Data represent mean values ± SEM pooled from two experiments of n = ≥5 mice per group. Student’s t-test was used to calculate the p values, and the Chi squared test was used to generate the Kaplan–Meier survival curve. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001 for comparisons between WT and miR-21–/–, except in panel G where they compare WT [ABX] and miR-21–/– [ABX]. Figure 5. View largeDownload slide Protection of miR-21–deficient mice against DSS-induced colitis is lost after antibiotic treatment. WT and miR-21–/– mice aged between 8 and 12 weeks were given a 4-way antibiotics mix ad libitum for 2 weeks. The mice were then weighed on Day 0 and given 3% DSS [w/v] ad libitum for 7 days. Weight [A], blood score [B] and diarrhea score [C] were determined daily and combined to generate a disease activity index [DAI] score [D]. The mice were sacrificed on Day 8 and their colons measured [E]. In addition, the mice were scored for mobility loss [F] and morbidity [G]. Data represent mean values ± SEM pooled from two experiments of n = ≥5 mice per group. Student’s t-test was used to calculate the p values, and the Chi squared test was used to generate the Kaplan–Meier survival curve. A single asterisk indicates p < 0.05; two asterisks indicate p < 0.01; three asterisks indicate p < 0.001 for comparisons between WT and miR-21–/–, except in panel G where they compare WT [ABX] and miR-21–/– [ABX]. 3.5. Altered composition of the miR-21–/– microbiota We employed 16S rRNA sequencing of fecal samples to investigate the possibility of differences in the composition of the microbiota between WT and miR-21–/– mice. After quality filtering and length trimming, an average of 37 313 [±7123 SD] 16S rRNA high-quality sequences were generated per sample. The average number of OTUs per sample was 355 [±16 SD]. The top 25 most abundant OTUs were compared between the two groups, with each OTU being given its phylum name followed by its genus name [e.g. Actinobacterium: Bifidobacterium] [Figure 6A]. There were considerable differences between the two groups, with the miR-21–/– group displaying a higher proportion of (i) Actinobacterium: Bifidobacterium; (ii) Firmicutes: Coprococcus; (iii) Firmicutes: Lactonifactor; (iv) Firmicutes; Oscilibacter; (v) Firmicutes: Clostridium sensu stricto; (vi) Firmicutes: Anaerosporobacter; and (vii) Firmicutes: Catonella; and a lower proportion of (i) Firmicutes: Clostridium XIVa; (ii) Firmicutes: Tannerella; (iii) Firmicutes: Dorea; (iv) Bacteroidetes: Barnesiella; and (v) Bacteroidetes: Prevotella, relative to WT mice [Figure 6A]. Beta diversity is represented by a Principal Coordinates Analysis [PCoA], performed using all 16S rRNA reads clustered at 97% similarity [Figure 6B]. Clear phylogenetic separation between the WT and miR-21 knockout groups was evident [ANOSIM p value = 0.001]. There were also several differences apparent at various taxonomic levels: at the phylum level, Proteobacteria [q = 0.00079935] were present in higher proportions in the miR-21–/– microbiota. At the family level, the miR-21–/– samples contained higher proportions of Bifidobacteriaceae [q = 0.01181564] and Peptostreptococcaceae [q = 0.14586726], and reduced proportions of Verrucomicrobiaceae [q = 0.07853655] and Bacteroidaceae [q = 0.05746273]. At the genus level, the WT samples group contained higher proportions of Enterorhabdus [q = 0.00036569], and reduced proportions of Odoribacter [q = 7.896e-05] and Bifidobacterium [q = 0.00143734] [Figure 6C–E]. Figure 6. View largeDownload slide View largeDownload slide 16S rRNA analysis of WT and miR-21–deficient mice fecal microbiota. The 16S rRNA reads were compared with the RDP database, and WT and miR-21–/– microbiota samples were compared for OTU analysis [A], beta diversity [B] and taxa analysis at phylum [C], family [D] and genus [E] levels. Data is representative of n = ≥9. Figure 6. View largeDownload slide View largeDownload slide 16S rRNA analysis of WT and miR-21–deficient mice fecal microbiota. The 16S rRNA reads were compared with the RDP database, and WT and miR-21–/– microbiota samples were compared for OTU analysis [A], beta diversity [B] and taxa analysis at phylum [C], family [D] and genus [E] levels. Data is representative of n = ≥9. 4. Discussion Over the past decade there has been an increasing appreciation of the crucial roles played by miRNAs in a wide variety of cellular processes. Among the biggest discoveries in that time was the role of miRNAs in inflammatory processes, in which they regulate both pro- and anti-inflammatory pathways.20 miR-21 is an extensively studied molecule with a known association with human IBD.6,8 With this in mind, we generated a knockout mouse to study the disease. Our initial finding that miR-21 was pathological in the DSS mouse model of colitis was surprising, as miR-21 is widely regarded as an anti-inflammatory mediator. These findings were supported by Shi et al., who described the same phenotype and suggested that miR-21 mechanistically influenced the disease by allowing the intestinal epithelial barrier to become more permeable via enhancing epithelial apoptosis and targeting the tight junction protein RhoB.6,8 These results indicate to us that miR-21 overexpression in IBD is unlikely to be simply a bystander effect caused by an inflammatory milieu, as has been suggested,21 but an active component of disease. Another field of study that has exploded in recent years is the growing understanding that the gut microbiota greatly influences human health and disease. Microbial dysbiosis has been found to play a significant contribution in a number of DSS-induced colitis phenotypes exhibited by transgenic mice with alterations in immune function. One such study demonstrated that the inflammasome component NLRP6 was required for intact host–microbe interaction, and that the exacerbated colitis phenotype exhibited by NLRP6–/– mice in response to treatment with DSS could be transferred to WT mice by co-housing.22,23 As miR-21 is a known regulator of the immunological activity of the bacterial sensor TLR4, we sought to ascertain the significance of the gut microbiota in our mice, and we reasoned that this might be a crucial aspect of the disease phenotype. First, we observed that, following co-housing of mice, WT mice co-housed with miR-21–/– mice displayed a reduced colitis phenotype when compared with the WT control group, which was evident in the fecal occult blood and in the diarrhea score. This indicated that the mice lacking miR-21 had an altered microbiota compared with their WT counterparts, which offered protection against the development of colitis; it also indicated that this protection was transferable. There is a sparse but growing literature regarding the potential of miRNAs to modulate the gut microbiota, but given the variety of roles in the regulation of bacterial sensing, it is likely that miRNAs play an indirect role. In one interesting study, conducted by Lui et al., the authors demonstrated that host miRNAs, including miR-21, are present in mouse feces and that they can be taken up by commensal bacteria to alter their gene expression. In addition, this study showed a significant microbial dysbiosis was present in Dicer knockout mice, which lack all functional miRNAs, and that this dysbiosis could be rescued by re-introducing synthetic miRNAs to the mice.11 To further investigate the protective microbiota observed in the miR-21–/– mice, we colonized GF mice with the fecal microbiota of WT or miR-21–/– mice prior to a DSS challenge. In this experiment, mice colonized with a miR-21–/– fecal microbiota exhibited protection from DSS-induced colitis relative to the mice colonized with the WT microbiota. A colonoscopy confirmed that the miR-21–/––colonized mice displayed fewer signs of inflammation. Colonoscopy is the current gold standard in IBD diagnosis, and it is becoming increasingly utilized in studies modelling the disease.24 It had been previously been reported that miR-21 expression was downregulated in the caecum of GF mice relative to that in conventional controls, implying that the caecal microbiota impacts on its expression.12 In order to establish whether miR-21 is induced in the DSS colitis model due to increased penetration of the epithelium by commensal bacteria and subsequently enhanced immune activation, miR-21 expression levels were compared between GF mice given DSS or water. It was observed that miR-21 expression remained enhanced in these mice upon DSS treatment. This was interesting and pointed to a mechanism of induction more related to the epithelial damage caused by DSS administration.25 As a final confirmation that the microbiota of miR-21–/– mice confers protection against DSS-induced colitis, we determined the consequence of depleting this microbiota by administering antibiotics to mice prior to challenge with DSS. Both WT-depleted and miR-21–/––depleted mice displayed increased detectable fecal occult blood and higher diarrhea scores earlier in the experiment. However, interestingly, the protective phenotype previously observed in the miR-21–/– mouse was lost following antibiotic treatment, as demonstrated in each of the disease scores and in the overall DAI score. Antibiotic depletion of the microbiota has been shown to increase intestinal haemorrhaging and mortality in mice subsequently treated with DSS.26,27 These effects have been demonstrated to be due to a weakened intestinal barrier and the absence of tonic microbial signals to the epithelium, respectively. Our results demonstrated an increase in haemorrhaging [as measured by fecal occult blood], in keeping with these studies. Most striking of all the observations in this experiment was the highly significant increase in sickness behaviour [measured by loss of mobility] and associated mortality displayed in the miR-21–/– mice treated with antibiotics. This effect has previously been described in WT mice at slightly later time points, but interestingly we saw mortality as early as Day 5 in the miR-21 –/– mice, indicating that in the absence of microbial signaling, miR-21 appears to play a protective role in this disease model. This finding confirms that the microbiota of miR-21–/– mice is protective in DSS-induced colitis, and that the protection demonstrated by the miR-21–/– mice is, at least in part, dependent on the presence of the microbiota. Having shown that the microbiota of the miR-21–/– mouse determined the outcome of DSS-induced colitis, 16S rRNA sequencing was employed to compare the fecal microbial profiles of miR-21–/– and WT mice. The top 25 most abundant OTUs in WT and miR-21–/– mice were compared and several significant differences were noted. Several of the observed differences pointed towards a potentially protective microbial signature in the miR-21–/– mouse, including a generally higher abundance of genera within the Firmicute phylum, which correlates with protection against IBD, while there was a lesser abundance of Bacteroidetes and Prevotella, which correlates with development of IBD.28,29 The higher abundance in the gram-positive probiotic genera Actinobacterium and Bifidobacterium can also be interpreted as evidence of a protective microbiota, as several studies have indicated a reduction in the severity of DSS-induced disease in mice treated with members of these genera.30–32 In order to assess whether or not these populations shared similar species, beta diversity was assessed and found to differ between WT and miR-21–/– fecal bacterial populations. This supported the case for an altered, protective microbiota being present in the miR-21–/– intestinal compartment. We next performed a taxa analysis to compare the relative proportions of the different bacterial groupings at various taxonomic levels. Surprisingly, at the phylum level there was an increased presence of Proteobacteria in the miR-21–/– mice, which would run counter to the idea of a protective microbiota. This phylum of gram-negative bacteria generally increase in proportion in IBD, with a commensurate decrease in Firmicutes.33,34 Further down the taxa, at family level the fecal microbiota of the mice lacking miR-21 display an increased proportion of Bifidobacteriaceae and Peptostreptococcaceae relative to the WT microbiota, and at genus level it displays an increase in Bifidobacterium and Odoribacter, all of which correlate with a healthy microbiota according to various studies.35,36 We examined the fecal microbiota, because this is routinely performed in the field, and also because the fecal microbiota was implicated in the co-housing and GF colonization studies that we carried out. However, it should be noted that the fecal microbiota might differ from that at individual mucosal surfaces throughout the gastrointestinal tract. This has been shown to be important in studies using microbial signatures as a biomarker for intestinal disease; thus, further investigation may be required in order to elucidate whether there are important differences in this model.37 Overall, these data confirmed that there was an altered microbiota in the miR-21–/– mouse, with some of these alterations being in accordance with previous studies indicating protective associations in relation to IBD and mouse models of colitis. An interesting aspect of this work worth considering is the potential alteration of other mediators present in the feces of miR-21–/– mice, which might impact the disease phenotype. These could include cytokines, antibodies, secreted epithelial factors or indeed other miRNAs, all of which have been demonstrated to be present in mouse feces and to have a transferable impact on disease outcomes.11 Secretory IgA [sIgA] has been shown to be an important mediator of microbial composition, a factor in microbial transferability, and a marker for susceptibility in IBD models.38,39 We measured sIgA levels in the feces of WT and miR-21–/– mice, and while there appeared to be reduced levels of sIgA in the feces of the miR-21–/– mice, this difference was not statistically significant [Supplementary figure S2]. We were unable to detect differences in fecal cytokines [data not shown]. We also measured the inflammatory markers MPO activity and IL-1β expression in all mice following DSS-induced colitis, but found little consistent evidence that these mechanisms may be influencing our observed phenotype [Supplementary figure S3]. This is somewhat in keeping with the counterintuitive nature of miR-21’s established role as an anti-inflammatory molecule in immune signaling.21 The mechanisms by which miR-21 may be influencing microbial composition have yet to be determined, but given that miR-21 has been demonstrated to be an important regulator of immune and epithelial barrier function, it is likely that modulation of these important mediators of gut homeostasis is involved. For instance, the enhanced barrier integrity displayed in miR-21–/– mice demonstrated in previous studies via increase in the expression of the miR-21 target RhoB may play a role in shaping the microbiota by creating a more favorable niche for colitis-protective bacteria.6,8 miRNAs have also been reported to alter the expression of specific mucins in the gut, which may also have a bearing on the niche occupied by these bacteria.40,41 It is also possible that miR-21 impacts the commensal bacteria more directly, by targeting secreted bactericidal mediators: for example, miR-21 has been shown to target CAMP and DEFB4A in humans.42 T-cells play an important role in immune homeostasis in the gut, and have a crucial role in shaping the gut microbiota.43 miR-21 has been demonstrated to impact the differentiation and activation of different T-cell subsets via modulation of key genes such as IL-12p35 and Stat3. For instance, in different inflammatory contexts, miR-21 has been shown to promote Th17 cells, to negatively regulate Treg cells and to promote memory T-cell activation.44–46 In murine models of colitis, miR-21 deficiency has been shown to skew T-cells towards Th1 responses.47 It would be interesting to explore the potential contribution of T-cell responses to shaping the microbiota in this system. These are a few of the possible mechanisms by which miR-21 may regulate the shape of the intestinal microbiota. It may be that one or all of them are involved in this system, and certainly it is likely that the altered barrier permeability demonstrated to be present in the miR-21–/– mouse by other groups is a factor. It is also possible that miR-21 may directly influence bacterial gene expression following secretion into the lumen in exosomes, as has recently been demonstrated.11 Clearly further studies are required in order to elucidate the pathways by which miR-21 may be exerting its influence on the microbiota. Another interesting aspect of the findings presented here are the implications of microbial alteration in the divergent phenotype displayed by miR-21–/– mice when subjected to the T-cell–dependent 2,4,6-trinitrobenzene sulfonic acid [TNBS] colitis and T-cell transfer colitis models. It has been previously reported that miR-21 deletion exacerbates colitis in these models through aberrant skewing of T-cell responses. As stated previously, T-cells help shape the microbiota, and the microbiota helps shape T-cell responses in inflammatory disease.43 Accordingly, it would be interesting to explore the microbial contribution to disease onset and progression in this model, and whether or not microbial manipulation [e.g. antibiotic depletion] might affect these phenotypes.47 There is a huge body of literature on miR-21’s role in many forms of cancer. miR-21 is upregulated in nearly all solid tumors and is a bone fide oncomiR, influencing many cancer-related processes such as cell division, apoptosis and metastasis.21,48 In the context of IBD, miR-21 is upregulated in colitis-associated cancer [CAC] and has also been shown to have a negative role in the azoxymethane [AOM]/DSS model of CAC, with miR-21–/– mice showing reduced susceptibility to the disease.7 Our results are interesting in this regard, because disruption to the microbiota has been reported to be involved in carcinogenesis through various mechanisms, including immune–microbiota crosstalk.49 This is particularly relevant in the case of CAC, because the breakdown of microbial tolerance and homeostasis in the gut leads to inflammation and subsequent tumorigenesis. It is therefore possible that miR-21’s impact on the intestinal microbiome may contribute to disease pathogenesis alongside its direct role in tumor formation and persistence. In our model, we have seen that miR-21 deletion gives rise to a microbiota that correlates with healthy individuals and protects from disease initiation, but does not remedy established disease [Supplemental figure S1]. Fecal microbial transplantation [FMT] therapy for IBD is in its infancy, with several studies reporting mixed efficacy for UC patients.50 A recent large study has shown efficacy using a multi-donor approach to maximize bacterial diversity, without identifying any particularly beneficial microbes.51 Future approaches for FMT and other microbial therapeutics may rely on a more detailed knowledge of which microbes are responsible for beneficial effects. Our findings are interesting in this context, because the mix of microbes and other factors present in the miR-21–/– mice have prophylactic value but appear less therapeutically beneficial. A deeper exploration of the composition of the transferable microbiota could provide valuable insights into this difference and inform future work in this area. As mentioned earlier, it would be very interesting to explore the microbial contribution to miR-21–/– disease phenotypes in other models, because this would further inform us of the potential efficacy of fecal transplant using a similar fecal bacterial signature. miRNAs are currently being targeted for therapeutic use in inflammatory disease. A recent study by Wang et al. has demonstrated that inhibition of the pro-inflammatory miRNA miR-223 can lessen disease severity in the DSS model of colitis.52 It would be interesting to see whether miR-21 inhibition would have a similar outcome in the DSS model. miR-21 inhibition is currently being developed for therapeutic use in a rare kidney disease, and has potential for use in IBD, though further work is required in order to tease out miR-21’s role in the various aspects of the disease [immune, microbiota, etc.], as highlighted by the divergence in its role in different models of disease.47,53 Our study highlights the need to consider the impact on the microbiota of any study carried out to this effect, as this may have an important bearing on the function of any such therapeutic intervention. In summary, our findings demonstrate that miR-21 expression is deleterious in the DSS model of IBD, and that the intestinal microbiota is crucial to the progression of this disease. Indeed, our findings identify a novel role for miR-21 in regulating the composition of the intestinal microbiota, and in the resulting development of intestinal inflammation. Funding This work was supported by a Starting Investigator Research Grant from Science Foundation Ireland [SFI] [grant number 11/SIRG/B2099] awarded to SCC. Conflict of Interest The authors have no competing interests to declare. Author Contributions SCC conceived ideas and oversaw the research programme. DGJ carried out the research. SCC and DGJ analysed data and wrote the manuscript. MAW and CM performed experiments. MR analysed histology. CT and EE assisted with germ-free experiments. RCR and PC assisted with analysis of 16S data. LAON funded creation of the miR-21–/– strain and provided invaluable guidance during the course of the project. Supplementary Data Supplementary data are available at ECCO-JCC Online. Acknowledgments The authors wish to acknowledge the technical assistance provided by the Comparative Medicine Unit, Trinity Biomedical Sciences Institute, in maintenance of animal colonies. References 1. Maloy KJ , Powrie F . Intestinal homeostasis and its breakdown in inflammatory bowel disease . Nature 2011 ; 474 : 298 – 306 . Google Scholar CrossRef Search ADS PubMed 2. Bartel DP . MicroRNAs: genomics, biogenesis, mechanism, and function . Cell 2004 ; 116 : 281 – 97 . Google Scholar CrossRef Search ADS PubMed 3. Bartel DP . MicroRNAs: target recognition and regulatory functions . Cell 2009 ; 136 : 215 – 33 . Google Scholar CrossRef Search ADS PubMed 4. Chapman CG , Pekow J . The emerging role of miRNAs in inflammatory bowel disease: a review . Therap Adv Gastroenterol 2015 ; 8 : 4 – 22 . Google Scholar CrossRef Search ADS PubMed 5. Carthew RW , Sontheimer EJ . Origins and mechanisms of miRNAs and siRNAs . Cell 2009 ; 136 : 642 – 55 . Google Scholar CrossRef Search ADS PubMed 6. Shi C , Liang Y , Yang J et al. MicroRNA-21 knockout improve the survival rate in DSS induced fatal colitis through protecting against inflammation and tissue injury . PLoS One 2013 ; 8 : e66814 . Google Scholar CrossRef Search ADS PubMed 7. Shi C , Yang Y , Xia Y et al. Novel evidence for an oncogenic role of microRNA-21 in colitis-associated colorectal cancer . Gut 2016 ; 65 : 1470 – 81 . Google Scholar CrossRef Search ADS PubMed 8. Yang Y , Ma Y , Shi C et al. Overexpression of miR-21 in patients with ulcerative colitis impairs intestinal epithelial barrier function through targeting the Rho GTPase RhoB . Biochem Biophys Res Commun 2013 ; 434 : 746 – 52 . Google Scholar CrossRef Search ADS PubMed 9. Zhang L , Shen J , Cheng J , Fan X . MicroRNA-21 regulates intestinal epithelial tight junction permeability . Cell Biochem Funct 2015 ; 33 : 235 – 40 . Google Scholar CrossRef Search ADS PubMed 10. Sheedy FJ , Palsson-McDermott E , Hennessy EJ et al. Negative regulation of TLR4 via targeting of the proinflammatory tumor suppressor PDCD4 by the microRNA miR-21 . Nat Immunol 2010 ; 11 : 141 – 7 . Google Scholar CrossRef Search ADS PubMed 11. Liu S , da Cunha AP , Rezende RM et al. The host shapes the gut microbiota via fecal microRNA . Cell Host Microbe 2016 ; 19 : 32 – 43 . Google Scholar CrossRef Search ADS PubMed 12. Singh N , Shirdel EA , Waldron L , Zhang RH , Jurisica I , Comelli EM . The murine caecal microRNA signature depends on the presence of the endogenous microbiota . Int J Biol Sci 2012 ; 8 : 171 – 86 . Google Scholar CrossRef Search ADS PubMed 13. Dalmasso G , Nguyen HT , Yan Y et al. Microbiota modulates host gene expression via MicroRNAs . Gastroenterology 2011 ; 140 : e19293 . 14. Smith P , Mangan NE , Walsh CM et al. Infection with a helminth parasite prevents experimental colitis via a macrophage-mediated mechanism . J Immunol 2007 ; 178 : 4557 – 66 . Google Scholar CrossRef Search ADS PubMed 15. Schmieder R , Edwards R . Quality control and preprocessing of metagenomic datasets . Bioinformatics 2011 ; 27 : 863 – 4 . Google Scholar CrossRef Search ADS PubMed 16. Edgar RC . Search and clustering orders of magnitude faster than BLAST . Bioinformatics 2010 ; 26 : 2460 – 1 . Google Scholar CrossRef Search ADS PubMed 17. Wang Q , Garrity GM , Tiedje JM , Cole JR . Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy . Appl Environ Microbiol 2007 ; 73 : 5261 – 7 . Google Scholar CrossRef Search ADS PubMed 18. Caporaso JG , Kuczynski J , Stombaugh J et al. QIIME allows analysis of high-throughput community sequencing data . Nat Meth 2010 ; 7 : 335 – 6 . Google Scholar CrossRef Search ADS 19. Bradley PP , Priebat DA , Christensen RD , Rothstein G . Measurement of cutaneous inflammation: estimation of neutrophil content with an enzyme marker . J Invest Dermatol 1982 ; 78 : 206 – 9 . Google Scholar CrossRef Search ADS PubMed 20. Baltimore D , Boldin MP , O’Connell RM , Rao DS , Taganov KD . MicroRNAs: new regulators of immune cell development and function . Nat Immunol 2008 ; 9 : 839 – 45 . Google Scholar CrossRef Search ADS PubMed 21. Sheedy FJ . Turning 21: induction of miR-21 as a key switch in the inflammatory response . Front Immunol 2015 ; 6 : 19 . Google Scholar CrossRef Search ADS PubMed 22. Elinav E , Strowig T , Kau AL et al. NLRP6 inflammasome regulates colonic microbial ecology and risk for colitis . Cell 2011 ; 145 : 745 – 57 . Google Scholar CrossRef Search ADS PubMed 23. Wlodarska M , Thaiss CA , Nowarski R et al. NLRP6 inflammasome orchestrates the colonic host–microbial interface by regulating goblet cell mucus secretion . Cell 2014 ; 156 : 1045 – 59 . Google Scholar CrossRef Search ADS PubMed 24. Beltzer A , Kaulisch T , Bluhmki T , Schoenberger T , Stierstorfer B , Stiller D . Evaluation of quantitative imaging biomarkers in the DSS colitis model . Mol Imaging Biol 2016 ; 18 : 697 – 704 . Google Scholar CrossRef Search ADS PubMed 25. Perše M , Cerar A . Dextran sodium sulphate colitis mouse model: traps and tricks . J Biomed Biotechnol 2012 ; 2012 : 718617 . Google Scholar CrossRef Search ADS PubMed 26. Rakoff-Nahoum S , Paglino J , Eslami-Varzaneh F , Edberg S , Medzhitov R . Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis . Cell 2004 ; 118 : 229 – 41 . Google Scholar CrossRef Search ADS PubMed 27. Hernández-Chirlaque C , Aranda CJ , Ocón B et al. Germ-free and antibiotic-treated mice are highly susceptible to epithelial injury in DSS colitis . J Crohns Colitis 2016 ; 10 : 1324 – 35 . Google Scholar CrossRef Search ADS PubMed 28. Bloom SM , Bijanki VN , Nava GM et al. Commensal bacteroides species induce colitis in host–genotype-specific fashion in a mouse model of inflammatory bowel disease . Cell Host Microbe 2011 ; 9 : 390 – 403 . Google Scholar CrossRef Search ADS PubMed 29. Natividad JM , Pinto-Sanchez MI , Galipeau HJ et al. Ecobiotherapy rich in firmicutes decreases susceptibility to colitis in a humanized gnotobiotic mouse model . Inflamm Bowel Dis 2015 ; 21 : 1883 – 93 . Google Scholar CrossRef Search ADS PubMed 30. Toumi R , Soufli I , Rafa H , Belkhelfa M , Biad A , Touil-Boukoffa C . Probiotic bacteria Lactobacillus and Bifidobacterium attenuate inflammation in dextran sulfate sodium–induced experimental colitis in mice . Int J Immunopathol Pharmacol 2014 ; 27 : 615 – 27 . Google Scholar CrossRef Search ADS PubMed 31. Osman N , Adawi D , Ahrne S , Jeppsson B , Molin G . Modulation of the effect of dextran sulfate sodium–induced acute colitis by the administration of different probiotic strains of Lactobacillus and Bifidobacterium . Dig Dis Sci 2004 ; 49 : 320 – 7 . Google Scholar CrossRef Search ADS PubMed 32. Srutkova D , Schwarzer M , Hudcovic T et al. Bifidobacterium longum CCM 7952 promotes epithelial barrier function and prevents acute DSS-induced colitis in strictly strain-specific manner . PLoS One 2015 ; 10 : e0134050 . Google Scholar CrossRef Search ADS PubMed 33. Mukhopadhya I , Hansen R , El-Omar EM , Hold GL . IBD—what role do Proteobacteria play ? Nat Rev Gastroenterol Hepatol 2012 ; 9 : 219 – 30 . Google Scholar CrossRef Search ADS PubMed 34. Matsuoka K , Kanai T . The gut microbiota and inflammatory bowel disease . Semin Immunopathol 2015 ; 37 : 47 – 55 . Google Scholar CrossRef Search ADS PubMed 35. Morgan XC , Tickle TL , Sokol H et al. Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment . Genome Biol 2012 ; 13 : R79 . Google Scholar CrossRef Search ADS PubMed 36. Nagy-Szakal D , Hollister EB , Luna RA et al. Cellulose supplementation early in life ameliorates colitis in adult mice . PLoS One 2013 ; 8 : e56685 . Google Scholar CrossRef Search ADS PubMed 37. Parthasarathy G , Chen J , Chen X et al. Relationship between microbiota of the colonic mucosa vs feces and symptoms, colonic transit, and methane production in female patients with chronic constipation . Gastroenterology 2016 ; 150 : 367 – 79.e1 . Google Scholar CrossRef Search ADS PubMed 38. Reinshagen M . [Immunoglobulin A coating identifies colitogenic bacteria in inflammatory bowel disease] . Z Gastroenterol 2015 ; 53 : 53 . Google Scholar CrossRef Search ADS PubMed 39. Palm NW , de Zoete MR , Cullen TW et al. Immunoglobulin A coating identifies colitogenic bacteria in inflammatory bowel disease . Cell 2014 ; 158 : 1000 – 10 . Google Scholar CrossRef Search ADS PubMed 40. Mo JS , Alam KJ , Kim HS , Lee YM , Yun KJ , Chae SC . MicroRNA 429 regulates mucin gene expression and secretion in murine model of colitis . J Crohns Colitis 2016 ; 10 : 837 – 49 . Google Scholar CrossRef Search ADS PubMed 41. Donaldson GP , Lee SM , Mazmanian SK . Gut biogeography of the bacterial microbiota . Nat Rev Microbiol 2016 ; 14 : 20 – 32 . Google Scholar CrossRef Search ADS PubMed 42. Liu PT , Wheelwright M , Teles R et al. MicroRNA-21 targets the vitamin D–dependent antimicrobial pathway in leprosy . Nat Med 2012 ; 18 : 267 – 73 . Google Scholar CrossRef Search ADS PubMed 43. Zhang H , Sparks JB , Karyala SV , Settlage R , Luo XM . Host adaptive immunity alters gut microbiota . ISME J 2015 ; 9 : 770 – 81 . Google Scholar CrossRef Search ADS PubMed 44. Li S , Fan Q , He S , Tang T , Liao Y , Xie J . MicroRNA-21 negatively regulates Treg cells through a TGF-β1/Smad-independent pathway in patients with coronary heart disease . Cell Physiol Biochem 2015 ; 37 : 866 – 78 . Google Scholar CrossRef Search ADS PubMed 45. Murugaiyan G , da Cunha AP , Ajay AK et al. MicroRNA-21 promotes Th17 differentiation and mediates experimental autoimmune encephalomyelitis . J Clin Invest 2015 ; 125 : 1069 – 80 . Google Scholar CrossRef Search ADS PubMed 46. Smigielska-Czepiel K , van den Berg A , Jellema P et al. Dual role of miR-21 in CD4+ T-cells: activation-induced miR-21 supports survival of memory T-cells and regulates CCR7 expression in naive T-cells . PLoS One 2013 ; 8 : e76217 . Google Scholar CrossRef Search ADS PubMed 47. Wu F , Dong F , Arendovich N , Zhang J , Huang Y , Kwon JH . Divergent influence of microRNA-21 deletion on murine colitis phenotypes . Inflamm Bowel Dis 2014 ; 20 : 1972 – 85 . Google Scholar CrossRef Search ADS PubMed 48. Krichevsky AM , Gabriely G . miR-21: a small multi-faceted RNA . J Cell Mol Med 2009 ; 13 : 39 – 53 . Google Scholar CrossRef Search ADS PubMed 49. Schwabe RF , Jobin C . The microbiome and cancer . Nat Rev Cancer 2013 ; 13 : 800 – 12 . Google Scholar CrossRef Search ADS PubMed 50. Lopez J , Grinspan A . Fecal microbiota transplantation for inflammatory bowel disease . Gastroenterol Hepatol 2016 ; 12 : 374 – 9 . 51. Paramsothy S , Kamm MA , Kaakoush NO et al. Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial . Lancet 2017 ; 389 : 1218 – 28 . Google Scholar CrossRef Search ADS PubMed 52. Wang H , Chao K , Ng SC et al. Pro-inflammatory miR-223 mediates the cross-talk between the IL23 pathway and the intestinal barrier in inflammatory bowel disease . Genome Biol 2016 ; 17 : 58 . Google Scholar CrossRef Search ADS PubMed 53. Gomez IG , MacKenna DA , Johnson BG et al. Anti-microRNA-21 oligonucleotides prevent Alport nephropathy progression by stimulating metabolic pathways . J Clin Invest 2015 ; 125 : 141 – 56 . Google Scholar CrossRef Search ADS PubMed Copyright © 2018 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Crohn's and ColitisOxford University Press

Published: Mar 28, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off