Background
Crohn’s disease (CD) and ulcerative colitis (UC) are the two primary and most prevalent forms of inflammatory bowel disease (IBD). Characterized by inappropriate and exacerbated immune responses within the gastrointestinal tract, the nature and location of the inflammatory changes defines each respective disease. Despite years of research, the exact cause of IBD is still relatively poorly understood. Nevertheless, it is clear that many factors, including genetic and epigenetic predispositions, the gut microbiota content, and environmental injuries or exposures, contribute significantly to the disease process. Linkage analysis and genome wide association studies (GWAS) have uncovered over 100 loci that have significant association with IBD [
1-
3]. Cytokines (IL-17), cytokine receptors (IL-23 receptor), and bacterial response elements (CARD15/NOD2, ATG16L1) are just a few of the pathways that have been found to be either mutated or otherwise altered in IBD patients and mouse models of chronic intestinal inflammation [
2-
5].
MicroRNAs (miRNAs), single stranded RNA molecules of 19–25 nucleotides, are poised to make significant contributions to defining the multifactorial etiology and pathobiology of IBD. Initially discovered in the early 1990s, this novel class of noncoding RNAs regulates gene expression post-transcriptionally to repress translation and/or promote mRNA degradation [
6-
9]. The biological footprint of miRNAs is widespread; over 30% of the genome is predicted to be actively regulated by miRNAs and studies have shown that miRNAs are involved in the control of a variety of normal cellular events including differentiation, organogenesis, and metabolism [
10,
11]. Further, aberrant expression of miRNAs has been associated with a growing number of disease states, including cancer and autoimmune diseases [
12-
17].
However, the consequence of how the alterations in miRNA expression occur and contribute to disease pathobiology remains mostly intangible at the moment. Understanding the role of miRNAs in the regulation of inflammation is an area of importance that may have broad significance to understanding the pathogenesis of IBD as well as a number of other diseases. Several studies have identified miRNAs that are associated with IBD in intestinal tissues and peripheral blood specimens [
18-
24]. In a previous study involving the interleukin-10 knockout (IL-10−/−) mice, a mouse model of chronic intestinal inflammation, we demonstrated selective dysregulation of miRNAs in colon tissue and peripheral blood leukocytes [
25]. We therefore postulated that miRNA expression would be similarly disrupted in IBD patients with the possibility that some of the same miRNAs might have altered expression. The purpose of this study was to identify differentially expressed miRNAs that could selectively discriminate CD from UC and healthy controls using colon, blood, and saliva specimens. A further goal was to determine how the tissue microenvironment affected miRNA expression.
Discussion
In the present study into the miRNA regulatory network of IBD, we were able to identify miRNAs (miR-19a, miR-21, miR-31, miR-101, miR-146a, and miR-375) that had statistically significant altered expression in pooled colon biopsy samples with several other miRNAs just outside the significance threshold (miR-26a, miR-142-3p, miR-155, miR-223, and miR-494;
p ≤ 0.1). Of these miRNAs, miR-21, miR-31, miR-146a, and miR-375 have been identified in previous studies examining miRNA expression in CD and UC colon biopsies [
18-
21]. Further, these miRNAs are implicated in inflammation and cancer with miR-21 and miR-375 functioning as oncomiRs while miR-146a has been reported to be a tumor suppressor [
39-
41]. miR-31 is more complicated in that it has been reported to be both an oncomiR in lung cancer, and a tumor suppressor in breast cancer [
42,
43]. The increased expression of miR-146a and possibly miR-31 seem to be attempts to rein in the chronic inflammatory response in IBD. Previous studies that established miR-146a functionally opposing the effects of miR-155 via regulation of inflammatory signaling proteins, particularly NF-κB, support this [
44-
47]. miR-155 overexpression in mice results in a myeloproliferative disorder [
44]. Conversely, miR-146a knockout mice develop hyperinflammatory and immunoproliferative disorders [
45-
47]. This suggests that, in IBD, the balance between these miRNAs with seemingly opposing functions is tilted and begs the question as to why the pro-inflammatory miRNAs win.
We additionally identified two miRNAs, miR-19a and miR-101, that have not previously been associated with IBD. miR-19a is a member of the miR-17-92 cluster and has been shown to be overexpressed in T-cell acute lymphoblastic leukemia and multiple myeloma where it was revealed that miR-19a negatively regulates the expression of CYLD and SOCS-1 respectively to promote cell survival and pathogenesis [
48,
49]. miR-101 has been associated with inflammation as a negative regulator of inducible costimulator (ICOS) and cancer cell stemness as a negative regulator of the corepressor C-terminal binding protein-2 (CtBP2) [
50,
51]. Once again, two miRNAs with seemingly opposing targets, miR-19a is pro-inflammatory while miR-101 is anti-proliferative, are both elevated in IBD.
Further, we were able to demonstrate that CD was associated with the differential expression of 10 miRNAs in a statistically significant manner (9 elevated, 1 decreased) while 6 miRNAs were elevated in a statistically significant manner in UC when comparing matched endoscopically uninvolved colon biopsies with endoscopically involved colon biopsies. Of these miRNAs, miR-21, miR-101, miR-142-5p, miR-155, and miR-223 had overlapping expression patterns between CD and UC. miR-146a was elevated in a statistically significant manner only in UC in the paired analysis while miR-19a, miR-31, miR-142-3p, miR-375, and miR-494 were altered in a statistically significant manner in CD.
Taken together, our analysis of the pooled colon samples versus the matched colon biopsies provides crucial but distinct information regarding the role of miRNAs in IBD. The former analysis suggests that a panel of miRNAs (miR-19a, miR-21, miR-31, miR-101, miR-146a, and miR-375) may be used as markers to identify and discriminate between CD and UC. The latter analysis provides more mechanistic information regarding miRNA involvement in disease pathology. Namely, it points to the underlying pathways that the miRNAs are altering to cause or promote disease. As an example, the overexpression of miR-31 and miR-146a, both of which target CD40L (predicted or experimentally confirmed), in the matched samples suggests that the CD40:CD40L costimulatory pathway plays an important role in promoting localized inflammation; CD40 and CD40L are elevated in IBD [
52]. With further studies, we may even be able to distinguish patients at higher risk of colon cancer based on miRNA analysis.
As a proof of concept, we analyzed miRNA expression in the saliva of IBD patients. This new frontier of RNA diagnostics in oral fluids is being pioneered in the detection of cancer [
33,
53-
56]. Although salivary miRNAs are emerging as a novel class of biomarkers in cancer, they have yet to be examined in other non-cancer diseases. In our study, we were able to detect differences in miRNA expression in the oral fluid from IBD patients. miR-21 in particular is a ubiquitous miRNA associated with many diseases and is an oncomiR [
57,
58]. Likewise, miR-31 and miR-101 have been implicated in cancer as well with the former being described as an oncomiR while the latter behaves as a tumor suppressor [
43,
59,
60]. Finally, our results identified a potentially new IBD associated gene. In a subset of CD patients,
RC3H1 expression was reduced in statistically significant manner in a subset of the endoscopically involved versus endoscopically uninvolved matched colony biopsy pairs. In mouse models deficient for
Rc3h1, either the sanroque missense mutation (
Rc3h1
san/san
) mouse or a genetrap knockout (
Rc3h1
gt/gt
) mouse, an acute small intestinal inflammation develops among other phenotypes [
50,
61,
62]. Roquin-1 is involved in the post-transcriptional regulation of mRNA. Roquin-1 localizes to P bodies and stress granules and plays a role in regulating mRNA turnover, and has been shown to regulate the expression of T cell coactivators, specifically ICOS and OX40 as well as the cytokines TNFα and IL-17. Given this role and the number of altered IBD-related miRNAs predicted or confirmed to target Roquin-1 (Additional file
1: Table S3), the importance of Roquin-1 in intestinal inflammation and human IBD necessitates further study.
This study also affirms the use of the IL-10−/− mouse as an effective model of IBD as miRNAs identified as differentially expressed in the CD and UC human samples were shared with the IL-10−/− mice [
25]. Although there are frequently difficulties in going from bench to bedside, at least in this instance it seems that the IL-10−/− mouse model faithfully recreates some aspects of IBD.
Competing interests
Dr. Graham is a consultant for RedHill Biopharma regarding novel therapies for Crohn’s disease.
Authors’ contributions
JSS, JRK, study concept and design; TA, BA, JH, HS, ARO, DYG, acquisition of samples and assessment of clinical disease; JSS acquisition of data; JSS, JRK, CS, analysis and interpretation of data; JSS, JRK, CS, statistical analysis; JSS, JRK, drafting of the manuscript; JSS, JRK, TA, BA, JH, HS, ARO, DYG, critical revision of the manuscript. All authors read and approved the final manuscript.