MicroRNA effects on gut homeostasis: therapeutic implications for inflammatory bowel disease

Abstract

Inflammatory bowel disease (IBD) spans a range of chronic conditions affecting the gastrointestinal tract which are marked by intermittent flare-ups and remissions. IBD results from microbial dysbiosis or a defective mucosal barrier in the gut that triggers an inappropriate immune response in a genetically susceptible person, altering the immune-microbiome axis. In this review, we discuss the regulatory roles of microRNAs, small non-coding RNAs with gene regulatory functions, in the stability and maintenance of the gut immune-microbiome axis and detail the challenges and recent advances in the use of microRNAs as putative therapeutic agents for treating IBD.

Inflammatory bowel disease: a perspective

Conventionally considered a disease of the West, inflammatory bowel disease (IBD) is now a global disorder whose geographical spread closely follows that of industrialization. With a compounding prevalence close to 0.5% in developed countries [1] and more than 6 million annual cases worldwide, IBD is predicted to become a major health and policy concern across the world [1]. The first reported case of IBD can be dated back to the mid-19^th century with the possibility that Charles Darwin himself was one of its undiagnosed victims [1]. However, IBD pathogenesis still remains incompletely understood, fostering an active field of research, due to its multifactorial nature comprising genetic as well as environmental factors that together challenge a simple interpretation of the disease [2]. These compounded factors can bring about microbial dysbiosis or a compromised epithelial barrier that disturbs gut homeostasis. Subsequently, activation of innate and adaptive immunity ensues which can lead to chronic conditions with recurring inflammation that define IBD. The two broad categories of IBD are Crohn’s disease (CD) and ulcerative colitis (UC). CD is characterized by a patchy, transmural inflammation that can affect any part of the gastrointestinal (GI) tract. In contrast, UC is limited to the colon with inflammation mostly occurring within the mucosa and submucosa (Figure 1A) [2]. In addition to the recurrent flare-ups, extra-intestinal inflammations in organs such as the kidney or eye, together with secondary complications such stenosis or colitis-associated colorectal cancers can further deteriorate the quality of life in IBD patients [2].

Figure 1: Intestinal barrier as a first line of defense against inflammation and IBD.

(A) Ulcerative colitis (UC) mostly affects the colon while Crohn’s disease (CD) can affect any part of the gastrointestinal tract. (B) The small intestine has small finger-like projections called villi that progressively grow shorter and wider down its length. In contrast, (C) the large intestine is completely devoid of villi and has a comparatively flat epithelium. (D) Junctional complexes between adjacent epithelial cells comprise tight junctions (TJ), adherens junctions (AJ), and desmosomes (DSM) that constitute a selectively-permeable system, giving rise to intestinal barrier function that provides the first line of defense against inflammatory agents present within the intestinal lumen. Intestinal epithelial cell (IEC), intraepithelial lymphocyte (IEL), macrophage (MP), dendritic cell (DC), antimicrobial peptide (AMP). Figure created with BioRender.com.

Existing treatments for IBD include immunomodulatory drugs such as Ozanimod, biologic therapies that target inflammatory cytokines such as TNF-α and IL-12/IL-23, or microbiome restoration which target various aspects of IBD pathogenesis, from immune dysfunction to microbial dysbiosis [2]. However, a considerable fraction of patients fails to respond to these therapies or exhibits impaired response over time, highlighting an unmet need for new effective therapies.

MicroRNAs (miRNAs) have emerged as potential biomarkers that can be targeted for therapeutic intervention in various diseases including IBD [3]. Briefly, miRNAs are evolutionarily conserved, 20-22-nucleotide-long, non-coding RNAs that activate the RNA interference (RNAi) pathway to induce gene silencing via mRNA translational inhibition or degradation in a sequence-specific way [4]. Two important enzymes involved in miRNA biogenesis are Dicer and Drosha which are commonly targeted in various studies aiming to assess the functional importance of miRNAs in biological pathways of interest [4]. In this review, we discuss the roles of miRNAs in the maintenance of gut homeostasis and highlight the challenges and recent advances in miRNA-based therapeutical interventions in preclinical models of IBD.

MiRNAs in intestinal barrier function

The mammalian intestinal epithelium is a single layer of cells that keeps the largest corps of immune cells in the body from striking against trillions of commensal microorganisms found within a healthy adult [5]. Its unique structure and cellular composition vary across intestinal lengths and differ greatly between the small and large intestines [6]. The small intestine exhibits small finger-like epithelial projections into the lumen called villi that progressively grow shorter and wider down its length. In contrast, the large intestine is completely devoid of villi and harbors a comparatively flat epithelium (Figure 1B, 1C) [6]. Across the epithelium, there are invaginations called the ‘crypts of Lieberkühn’ where LGR5⁺ intestinal stem cells (ISCs) reside. Intestinal epithelial cells (IECs) together with junctional complexes create a selectively permeable structure called the intestinal barrier (IB) that maintains gut homeostasis. Accordingly, loss of IB function is associated with IBD and other GI diseases [5]. Although it is undecided if IB dysfunction is causal in these conditions, it is proposed as a putative therapeutic target in many of them [5].

MiRNAs in the regulation of IECs

The intestinal epithelium comprises a number of specialized IECs such as enterocytes (for nutrient absorption), goblet cells (for mucus production), enteroendocrine cells (for secreting hormones), Microfold (M) cells (for facilitating antigen uptake), and chemosensory cells called Tuft cells (Figure 1B, 1C) [7]. There are also Paneth cells in the small intestine—the large intestine equivalents are the deep crypt cells—that secrete proteins such as epidermal growth factor (EGF), maintaining the stem cell niche to facilitate intestinal repair and regeneration [8] (Figure 1B, 1C). In addition, Paneth cells secrete antimicrobial agents such as defensins and express genes such as NOD2, ATG16L1 and IRGM1 that are dysregulated in CD patients [6]. Similarly, defects in mucus production due to goblet cell dysfunction have been associated with enhanced susceptibility to developing UC [6].

Many recent studies have reported on the direct roles of certain miRNAs in the regulation and function of IECs. For example, IEC-specific deletion of miR-802 in mice enhanced the antimicrobial properties of Paneth cells while increasing the proliferation and turnover of IECs [9]. These effects were reported to be mediated by Tmed9 expression, a direct target of miR-802 which supported Wnt and defensin secretion in Paneth cells [9]. Similarly, another study reported that IEC-specific transgenic expression of miR-195 in mice reduced the number of Paneth and tuft cells, which increased the vulnerability of IB to endotoxins by directly repressing tuft cell-specific double cortin-like kinase 1 (Dlck1) [10]. Moreover, the roles of miRNAs in the regulation of goblet cell differentiation and function are well-documented such that gut-specific deletion of Dicer in mice results in 50% fewer goblet cells than in control mice [11]. The same study also showed that miR-375, whose reduced expression has been associated with UC [12], promotes goblet cell differentiation by modulating the expression of the effector protein Gob5. In another study, miR-429 downregulated goblet cell-secreted mucin MUC2 by inhibiting myristoylated alanine-rich protein kinase C substrate (MARCKS) in human cell lines [13]. Similarly, miR-16 suppressed the proliferation and growth of IECs in rat jejunal crypts by targeting Cyclin D, regulating the G1/S transition [14]. Moreover, miR-99b inhibited enterocyte migration along the crypt-villus axis by targeting the milk fat globule-EGF factor 8 (Mfg-e8) mRNA in mice [15].

Certain miRNAs have also been reported to play an important role in the highly dynamic process of intestinal regeneration. For example, the HIPPO pathway—which regulates epithelial regeneration and is frequently compromised in IBD [16]—has been reported to be a direct target of miR-31. Specifically, MiR31^−/− mice showed impaired epithelial regeneration following dextran sodium sulfate (DSS)-induced colitis [16]. Mechanistically, miR-31 targets the Tgf-β pathway which in turn derepresses Wnt expression, facilitating epithelial regeneration following irradiation [17]. Similarly, miR-143/145 directly targets insulin-like growth factor binding protein 5 (Igfbp5) to activate insulin-like growth factor (IGF), signaling which helps in intestinal regeneration following DSS-induced colitis in mice [18]. In contrast, miR-381-3p opposes epithelial regeneration after hypoxia/reoxygenation injury in rat IEC-6 cells by inhibiting nuclear receptor-related protein 1 (Nurr1) from the proliferative pathway [19]. MiRNAs can also regulate IEC plasticity during acute intestinal injuries that often deplete the ISC pool in crypts. IEC plasticity helps restore the ISC pool through dedifferentiation, given that post-mitotic progenitor cells such enteroendocrine cells or even the terminally differentiated Paneth cells, can revert back to being ISCs (which form the basis for intestinal regeneration) [20]. For example, a recent report showed that miR-802^−/− mice exhibited enhanced Paneth cell plasticity by upregulating Wnt signaling due to derepression of an miR-802 target, Tmed9 [9]. Collectively, these studies identify potential miRNAs that might be targeted to maintain IEC-mediated gut homeostasis in pathological conditions such as in IBD.

MiRNAs in epithelial barrier integrity

In addition to their roles in IEC differentiation, proliferation, and repair, miRNAs have also been implicated in the regulation of intercellular junctions between IECs controlling epithelial permeability, which is frequently found impaired in IBD [21]. Tight junctions (TJ), adherens junctions (AJ), and desmosomes come together to make junctional complexes such that TJs are the closest to the lumen followed by AJs and then desmosomes (Figure 1D) [21]. TJ is the primary determinant of paracellular permeability harboring ‘pore’ and ‘leak’ pathways. Specifically, the pore pathway is highly selective and is mainly regulated by claudins together with cingulin, while the leak pathway is less selective and is regulated by occludin and Zonula-occuladens-1 (ZO1) [22]. Several miRNAs can modulate epithelial barrier function by targeting the members of the pore and leak pathways. Specifically, miR-423-5p was recently reported to compromise epithelial barrier function by targeting Claudin 5 mRNA; mice administered miR-423-5p inhibitor showed improved symptoms of DSS-induced colitis [23]. In addition, miR-223 [24] and miR155-5p [25] can directly inhibit CLDN8 and CLDN16, respectively, in human cell lines (Figure 2A). Also, CLDN2 was reported to be a direct target of miR-16 and miR-125b [26] and was upregulated by miR-34a/c-5p and miR-29b-3p, which directly target the transcriptional repressor of CLDN 2, SNAIL, in human cell lines [27]. Moreover miR-320a was overexpressed in CD patients and it enhanced the expression of JAM-A, an important TJ protein facilitating membrane apposition [28], in T84 cells [29]. OCCLUDIN (from the leak pathway) was downregulated by direct targeting of miR155-5p [25] and miR-122a [30], and upregulated by miR-21, which directly targeted the gene encoding RHO-associated protein kinase 1 (ROCK1) in human cell lines [31]. ZO1 was inhibited by miR-675 [32] and upregulated by miR-34a/c-5p and miR-29b-3p, via targeting of the transcription factor SNAIL in human cell lines [27]. Altogether, the pore and leak pathways are direct targets of miRNA-based TJ regulation, such that the miRNAs supporting barrier function are often downregulated in gastrointestinal diseases such as IBD [23][25][26][29].

Key Figure, Figure 2: MiRNAs implicated in the regulation of gut homeostasis in mice and/or humans.

Shown are: (A) MiRNAs modulating intestinal barrier function; (B) MiRNAs implicated in the crosstalk between the microbiota and the intestinal epithelium; (C) MiRNAs implicated in the regulation and function of intestinal epithelial cells (IECs), toll-like receptors (TLRs), and the innate immune system; (D) MiRNAs modulating the crosstalk between innate and adaptive immune responses, as well as their presumed functions. Black lines represent reported indirect interactions whereas green lines represent reported direct activating interactions, and solid red lines represent reported direct inhibitive interactions. The dotted red line represents reported indirect inhibitive interactions. antimicrobial peptides (AMPs). Figure created with BioRender.com.

MiRNAs in gut microbial homeostasis

The human GI tract contains more than 40 trillion microorganisms including bacteria and virus-like particles [7]. The human gut microbiome is inherited from the host’s mother, after which it matures until the age of three, and remains relatively stable thereafter [7]. It also plays an important role in mucosal homeostasis by outcompeting pathogenic invaders for nutrients. Moreover, microbiota are also crucial in promoting tolerogenic signaling and maintenance of epithelial integrity by regulating TJ proteins [33]. Microbial dysbiosis is associated with different GI diseases including IBD [34]. While dietary habits define the core microbiota, the functions of miRNAs within the diet can also shape the gut microbiome (Figure 2B). For example, miRNAs-containing ginger exosome-like nanoparticles (GELN) have been reported to alleviate DSS-induced colitis in mice by altering the gut microbiota, wherein miR-7267-3p has promoted the growth of Lactobacillaceae and Bacteroidales S24-7 and at the same time, inhibited that of Clostridiaceae. Specifically, miR-7267-3p targeted various genes in Lactobacillus rhamnosus (LGG) including monooxygenase, resulting in LGG growth and increased secretion of indole-3-carboxaldehyde (I3A) in GELN-treated mice, as compared to controls. I3A subsequently promoted IL-22 secretion in mouse colonic mucus and enhanced barrier function [35]. Interestingly, miRNAs from the host can also shape and alter the microbiota by directly modulating bacterial growth. For example, in a pioneering study, human and mouse hosts were reported to utilize miRNAs produced by IECs to regulate the gene expression and growth of certain bacteria, including Escherichia coli, in the gut [36]. Specifically, IEC-specific deletion of Dicer resulted in altered gut microbiome and increased the susceptibility to DSS-induced colitis in mice [36]. Another study reported that miR-30d ameliorated experimental autoimmune encephalomyelitis (mouse model of multiple sclerosis) by altering the gut microbiome, possibly by supporting the growth of Akkermansia muciniphilia via upregulation of bacterial lactase [37]. Similarly, oral administration of miR-142a-3p was reported to alleviate DSS-induced colitis in mice by promoting the growth of Lactobacillus reuteri—an important microbiome genus for gut homeostasis; although inhibition of miR-142a-3p in a cell line did not reduce the production of inflammatory cytokines, treating mice with antibiotics along with miR-142a-3p inhibitor abrogated the protective effects of miR-142a-3p, suggesting that miR-142a-3p worked by altering the microbiome via increased Lactobacillus reuteri numbers. [38]. In a separate study, miR-21 promoted gut dysbiosis by inhibiting the growth of Lactobacillus reuteri [39]. However, the mechanisms by which certain miRNAs can influence bacterial growth and control gut microbial dysbiosis are only beginning to be understood. Conversely, even microbiota can change the miRNA content in the intestine [40]. For example, commensal bacteria induced the expression of miR-21-5p in human IECs which in turn increased epithelial permeability, possibly by downregulating TJ proteins CLAUDIN-4 and OCCLUDIN [40]. Altogether, these studies demonstrate the close interdependence between gut homeostasis and microbiota while highlighting the roles of certain miRNAs in modulating this balance.

MiRNAs in pattern recognition receptor signaling

The gut microbiome is contained in a highly dynamic environment which necessitates the timely recognition of and response to intestinal damage and other pathogens by pattern recognition receptors (PRRs) through their interactions with pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs). Toll-like receptors (TLRs) and nod-like receptors (NLRs) are two widely-studied classes of PRRs. TLRs are membrane-bound and interact directly with microbes to identify PAMPs and DAMPs, while NLRs are cytosolic and do the same in an intracellular fashion [41]. Most IECs, in addition to dendritic cells (DCs) and macrophages, express NLRs and TLRs [42]. Once stimulated, TLRs and NLRs interact with their respective adaptor molecules to initiate downstream signaling that, depending on the type of TLR or NLR involved, leads to the activation of NF-κB, interferon regulatory factor (IRF), and mitogen-activated protein kinase (MAPK) pathways responsible for transcriptionally upregulating proinflammatory genes, growth factors, and cytokines [43]. Most of these pathways are often dysregulated in IBD [44]. In fact, NOD2—an NLR which is expressed in DCs, macrophages, monocytes, and Paneth cells—was the first locus to be associated with adult IBD [45]. These observations have increasingly projected PRR pathways as potential drug targets in such inflammatory disorders [41]. MiRNAs harboring gene regulatory roles in PRR have thus been considered as candidate therapeutic targets. A recent study showed that miR-146a—an important, evolutionarily conserved, anti-inflammatory miRNA [46]—targeted the adaptor molecule receptor interacting serine/threonine kinase 2 (Ripk2) from the Nod2 signaling pathway to mediate its anti-inflammatory functions; indeed, its absence is associated with severe forms of DSS-induced colitis in mice [47]. In addition to its role in Nod2 signaling, miR-146a is also implicated in the regulation of γδ-T cell plasticity by downregulating the Nod1 expression in mice [48]. MiR-146a has also been reported to downregulate the adaptor molecules, IL-1 receptor associated kinase (IRAK1) and tumor necrosis factor-associated factor 6 (TRAF6) from the TLR pathway in human cell lines [49][50]. Relevant to IBD, miR-146a can target IRAK1 and limit TLR signaling in IECs and provide protection from bacteria-induced epithelial damage in mice [51]. Moreover, NOD2 expression was also reported to be suppressed by direct targeting of other miRNAs such as miR-10a [52], miR-320 [53], miR-192, miR-495, miR-512, and miR-671 [54] that restrained the release of inflammatory cytokines in human cell lines, and their expression is dysregulated in IBD patients [52][53][54]. Similarly, TLR4—a major endotoxin receptor expressed by major immune cells and enterocytes in small and large intestines [42]—was reported to be translationally suppressed by the miRNA let-7i [55] and was negatively regulated by miR-21 [56] in a MyD88- and NF-κB-dependent manner in human cell lines [55][56]. While the protein expression of TLR2, which primarily recognizes gram-positive bacteria and is mainly expressed in innate immune cells [57] and enterocytes in the proximal colon [42], was downregulated by miR-195 in human macrophage cultures [58]. Similarly, miR-149 negatively regulates the inflammatory response in murine macrophage cultures by downregulating the expression MyD88 downstream of all TLRs, with the exception of TLR3 [59]. Collectively, these studies demonstrate the importance of PRR signaling from IECs and myeloid cells in mobilizing appropriate responses to injuries or infections in the gut, and in identifying miRNAs that can potentially be targeted in case of delayed or inappropriate responses, as see in IBD.

Complementing their roles in PRR signaling, miRNAs are also involved in the transcriptional regulation of cytokines in response to PRR stimulation and downstream signal transduction [60] that eventually leads to the activation of innate and adaptive immunity, as discussed below.

MiRNAs in intestinal immune homeostasis

Gut mucosal immunity functions as the last sentinel to intestinal disequilibria resulting from persistent microbial dysbiosis or a defective mucosal barrier. Certain miRNAs exhibit important, multifaceted roles in regulating mucosal immunity to reinstate and maintain intestinal homeostasis.

MiRNAs in innate immunity

Activation of the innate immune system involves a complex web of signaling pathways (Box 1 and Figure 3) many of which are simultaneously modulated by individual miRNAs (Figure 2C). For example, miR-223—a new putative biomarker for IBD [61] with anti-inflammatory roles—has been shown to limit intestinal inflammation in mouse models by constraining Nlrp3 inflammasome in mice [62]. MiR-223 has also been implicated in neutrophil functions via the inhibition of IL-18-mediated neutrophil extracellular trap (NET) formation in human neutrophil cultures [63]. Moreover, miR-223^−/− mice develop more severe colitis than wildtype mice due to stronger pro-inflammatory response in macrophages (MPs) and DCs [64]. Also, miR-155—a pro-inflammatory microRNA with multiple roles in the maintenance of gut immune homeostasis (Figure 2)—can promote the accumulation of innate lymphoid cells (ILC2) in response to IL-33 (or alarmin) stimulation by protecting mice against apoptosis [65]. It has also helped maintain IFN-γ expression in patient-derived NK cells by targeting suppressor of cytokine signaling 1 (SOCS1) [66]. However, unlike miR-223, miR-155 can promote NET formation via peptidylarginine deaminase (Pad4) upregulation in mice [67]. Activated MPs in mice have been reported to express miR-155 in an NF-κB-dependent manner, mounting a pro-inflammatory response during the initial stages of inflammation. This initial pro-inflammatory response was accomapnied by a gradual build-up of another important IBD-related miRNA miR-146, which opposed NF-κB expression in later stages of inflammation that further attenuated miR-155-mediated expression of pro-inflammatory cytokines [68]. In addition, miR-146 has been reported to exert its anti-inflammatory function via multiple pathways in the innate immune system: it has suppressed ILC2 proliferation and function (as measured by IL-5 secretion) in mice [69], downregulated NET formation and neutrophil senescence associated with chronic inflammation in mice [70], decreased IFN-γ expression in human NK cells [71], suppressed the expression of MHC-II and pro-inflammatory cytokines such as IL-6 from human and mouse DCs and macrophages [47][72][73], and suppressed MP activation in mice, by directly targeting IRF5 [74]. Accordingly, miR-146a^−/− mice are resistant to DSS-induced colitis [47]. Another multifaceted miRNA, miR-150, can repress the cell lytic functions of mouse NK cells by downregulating the expression of pore-forming cytolytic protein perforin-1 [75]. Furthermore, miR-150^−/− mice exhibit reduced expression of TGF-β resulting in impaired intraepithelial lymphocyte production and differentiation [76], as well as enhanced B1-mediated humoral immunity via the derepression of transcription factor c-Myb [77]. In addition to such multifaceted miRNAs, other miRNAs have also been identified as individually controlling the generation, differentiation, and proliferation of innate immune cells. For example, miR-183 [78] and miR-20a [79] have inhibited human NK cell cytolytic function by targeting NK cell surface receptor proteins. Furthermore, miR-34a [80] and let-7i [81] have supported the activation and maturation of human and mouse DCs, respectively. Also, miR-24 have promoted the polarization of human MPs to the so-called “M1-like” pro-inflammatory subtype [82] while let-7c has polarized mouse MPs towards the anti-inflammatory “M2-like” subtype [83]. Together, such studies further highlight the versatility of certain miRNAs in the modulation of innate immunity in homeostatic and disease conditions.

Box 1: Activation of the innate immune system in the gut of mice and/or humans.

The intestinal epithelium engages in a constant, continuous interaction with commensal and dietary microbes within the alimentary tract. IECs within the epithelium form a physical barrier between these potential pathogens and the largest immune compartment in the body, and also play a functional role in mobilizing appropriate immune responses should a breach happen. Polarized IECs secrete IL-8 upon TLR2 apical stimulation, and TLR3, TLR5 or TLR9 basolateral stimulation[119][120][121][122]. IL-8 is the primary cytokine recruiting neutrophils to sites of damage or infection [123]. IECs also secrete CXCL2 [124] and CXCL5 [125], which recruit other polymorphonuclear leukocytes (PMNs) such as eosinophils and basophils [126]. Antimicrobial peptides such as defensins released during degranulation by neutrophils and Paneth cells in the epithelium have activating effects on resident myeloid DCs in the lamina propria or Peyer’s patches (PP) [127]. Under homeostatic conditions, DCs sample the lumen for potential pathogens via transepithelial dendrites, internalize antigens and present them to T cells in secondary lymphoid organs (SLO) such as mesenteric lymph nodes or PP, activating tolerogenic immune responses by generating FoxP3⁺ Treg cells [128]. However, when encountering PAMPs or DAMPs, DCs release pro-inflammatory cytokines and massively migrate to SLOs to initiate adaptive immune responses.

DCs [129], IECs [130] and PMNs [131][132] also play important roles by recruiting monocytes to the affected sites in the epithelium, releasing chemokines such as MIP-1α. These monocytes are polarized and can differentiate into pro-inflammatory macrophages via IFN-γ released from the NK cells or Th1 cells [133]. These are highly bactericidal, secrete pro-inflammatory cytokines, and phagocytose cellular debris along with aging and apoptotic neutrophils, as well as IECs. IECs are also involved in the activation of the TCR-expressing intraepithelial lymphocytes (IELs), type b—the more common of the two types of IELs, IEL-a and IEL-b (Figure 1) [134]. Most IELs are CD8⁺ with cytolytic phenotypes similar to CD8⁺ cytotoxic T cells. However, IEL-b, unlike IEL-a, do not recognize antigens presented by antigen presenting cells via polymorphic MHC and are instead primed directly by IECs [134].

DCs [135][136] and IECs [137][138] also release t IL-7, IL-12, IL-15, and IL-18 that activate the IFN- γ-producing group 1 innate lymphoid cells (ILC1s), including NK cells [139]. ILCs are the innate counterpart to helper T cells and lack antigen receptors; they are quicker in reacting to perceived threats [140]. While ILC1 cells are similar to—albeit, less cytolytic than—NK cells, ILC2 cells are the innate counterpart of the Th2 cells, whereas ILC3 cells are that of Th17 cells [141]. When Tuft cell-derived cytokine IL-25 activates ILC2 cells in the lamina propia [142], they secrete Th2 cytokines IL-4 and IL-5 -- important in B1 cell activation [143][144].

Figure 3: Crosstalk among the three layers of defense against intestinal inflammation that comprise intestinal barrier function, and the innate and adaptive immune systems in mice and/or humans.

Intestinal epithelial cells (IECs) recognize potential pathogens via pattern recognition receptors such as the toll-like receptors (TLRs). TLRs in turn induce the expression of cytokines such as IL-8 that directly recruit the innate arm of the immune system. In the case of persistent inflammation, antigen presenting cells (APCs) such as the dendritic cells (DCs) and macrophages (MPs) initiate the activation of the adaptive immune system. Black lines represent reported indirect interactions, whereas green lines represent reported direct activating interactions. Also, DCs and MPs are invariably involved in the activation of adaptive immunity, except for B2 cells. Hence, the arrows connecting DCs, MPs and adaptive immune cells are omitted for simplicity. Antimicrobial peptides (AMPs). Figure created with BioRender.com.

MiRNAs in adaptive immunity

With its repertoire of immune cells, the innate immune system is capable of independently resolving intestinal infection or injury in a timely manner [68]. However, in case of chronic, persistent infection, activation of the antigen-specific adaptive immune system is triggered (Box 2 and Figure 3).

Box 2: Activation of the adaptive immune system in the gut of mice and/or humans.

The adaptive immune system is equipped with cell-mediated and antibody-mediated immunity, as enforced by T cells and B cells, respectively. Conventional DCs in the absence of any perceived threat, maintain gut homeostasis by inducing the differentiation of naïve T cells into FoxP3⁺ Treg cells. With persistent inflammation, DCs activate the inflammatory response in SLOs by activating CD8⁺ and CD4⁺ T cells. Naïve CD8⁺ (cytotoxic) T cells, when presented with antigens and stimulated by DCs in SLOs, are activated and home to the affected regions in the gut where they release proinflammatory cytokines together with apoptosis-inducing cytotoxins such as perforins and granzymes. Naïve CD4⁺ T cells, upon TCR stimulation, are activated to first undergo clonal expansion followed by differentiation into helper T cells—Th1, Th2, Th17 or follicular Tfh cells—depending on the cytokine profile of the antigen-presenting DCs [145]. After the initial expansion phase, T cells (both CD4⁺ and CD8⁺) effectively mount an inflammatory response against the existing infection and resolve it within days, after which more than 90% of them undergo sudden apoptosis, marking the contraction phase. The remaining, long-lasting T cells (both CD4⁺ and CD8⁺) mark a third phase, where more specific, effective and faster T cell responses occur, forming the basis of memory T cell function [146][147]. As part of the adaptive B cell repertoire, B2 cells in SLOs are activated in a two-stage process whereby antigen (soluble as well as membrane-bound) recognition by B2 cells is followed by their activation from follicular Tfh cells in germinal centers. Once activated, B2 cells can undergo antibody class switch recombination to mount a specific antibody response followed by maturation into highly antigen-specific plasma or memory B2 cells via their interactions with helper T cells in the presence of Th17 cytokine IL-21 [148]. The highly specific antibodies from plasma cells opsonize pathogens in the gut for phagocytosis.

Once the source of inflammation is cleared from the mucosa, macrophages play a major role in re-establishing gut homeostasis. Tissue-resident macrophages, which are intrinsically anti-inflammatory, promote tissue repair and homeostasis, releasing TGF-β and IL-10. These cytokines further drive the transition of monocyte-derived pro-inflammatory macrophages toward pro-repair, anti-inflammatory macrophages [44]. However, during chronic inflammation, cytokines from anti-inflammatory macrophages are outcompeted by pro-inflammatory cytokines, resulting in continued recruitment of neutrophils, which can exacerbate inflammation, such as in IBD.

Similarly to what has been observed in innate immunity, adaptive immunity is also extensively modulated by miRNAs (Figure 2D). Of note, here again we focus on multifunctional miRNAs regulating multiple components of the adaptive immune system simultaneously. MiR-155, playing multiple roles in innate immunity has also been reported to regulate adaptive immunity; its expression was directly correlated with the numbers of Th1, Th17, and Treg cells in mice [84]. However, while miR-155 promoted IL-17 production in mouse Th17 cells, it failed to promote the release of TGF-β and IL-10 in Treg cells [84], despite the fact that miR-155 is under the influence of Treg transcription factor Foxp3 [85]. In fact, miR-155 inhibition was reported to confer protective effects in DSS-induced colitis in mice by targeting Th17-mediated inflammation [86][87]. It has also been implicated in the regulation of humoral immunity by promoting the production of high-affinity IgG antibodies by B2 cells in the mouse germinal center [88]. Similarly, the anti-inflammatory miRNA, miR-146a, with multiple roles in innate immunity was also reported to limit Tfh cell numbers and hence, humoral immunity in the mouse germinal center, possibly by targeting the transcription factor Stat1 [89]. MiR-146a can also inhibit Th17 cells in mice [47]. In addition, a cluster of miRNAs, miR-17-92, with known oncogenic roles can support T cell-dependent antibody responses by promoting Tfh cell differentiation in mouse germinal centers [90]. From this cluster, MiR-17 and MiR-19b can promote Th1 cell pro-inflammatory responses and inhibit Treg cell differentiation in mice [91]. The miR-17-92 cluster has also been reported to control the effects on Th17 cell-mediated inflammation by targeting macrophage IL-6 expression—a cytokine responsible for promoting Th17 differentiation as opposed to Treg cells—during bacterial infection in mice [92]. A recent study showed that miR-92a from the miR-17-92 cluster operated in a T cell-intrinsic manner to promote Th17 differentiation and suppress Treg cells via direct inhibition of transcription factor Foxo1 in mice [93].

The inflammatory cytokine IL-12, which promotes Th1 cell differentiation, is a direct target of miR-21; miR-21^−/− mice had increased concentrations of Th1 cell-derived IFN-γ and Cxcl9 and decreased Th2 cell-derived IL-4 and Ccl17 [94]. Similarly, miR-29 was reported to oppose Th1 cell differentiation by targeting the Th1-specific transcription factor T-bet in mice [95]. MiRNAs can also prolong inflammation by enhancing the stability of T cells. For example, miR-148a has been reported to prolong inflammation by promoting the stability of Th1 cells via inhibited expression of proapoptotic gene Bim in the colon of colitic mice [96]. In contrast, miR-210 expression in mouse CD4⁺ T cells can promote Th1 and Th17 cell differentiation by directly targeting Stat6 and Lyn mRNAs [97]. By contrast, miR-219a-5p and miR-20b suppressed Th17 cell-mediated inflammation in mice by inhibiting the transcription factors Rorγt and Stat3 [98][99].

Although discussed individually, there is constant crosstalk among each of the individual miRNA components within the GI tract (Figure 3). This adds to the complex and multifactorial nature of IBD, which we posit might be putatively managed by the pleiotropic actions of certain miRNAs, as discussed below.

Therapeutic potential of miRNAs in IBD

The existing treatments for IBD involve aminosalicylates, steroids, and immunomodulatory drugs that manage primary inflammation [2]. However, in cases when such interventions are met with failure, novel biologic therapies that target cytokines or integrins are employed [2]. Although highly effective, many patients eventually develop therapeutic non-response to biologic therapies too. For example, patients receiving biologic therapies involving anti-TNF antibody treatment are prone to develop resistance within just a year of therapeutic initiation, due to the activation of alternative signaling pathways [2]. While there have been efforts to minimize such alternative pathway-mediated non-responses via molecules that regulate multiple cytokine pathways simultaneously [2], miRNAs with their inherent multi-target nature might provide a ready alternative to be tested.

MiRNAs are being increasingly identified as faithful diagnostic markers in IBD, as they are differentially expressed in different patient samples relative to healthy controls (table 1). This disease-related, differential expression of miRNAs in patients might be targeted to treat IBD in two major ways: (i) by inhibiting upregulated miRNAs using antagomirs and (ii) by replacing downregulated miRNAs via miRNA mimetics. On the one hand, antagomirs are synthetic single-stranded RNA molecules that inhibit overexpressed target miRNAs by binding to them in a perfectly complementary way [100]. On the other hand, miRNA mimetics (or miRNA agomirs) are synthetic double-stranded miRNAs that simulate the function of downregulated target miRNAs. Both antagomirs and agomirs require chemically-modified nucleotides that confer enhanced stability and protection from cellular nucleases. Such modifications are necessary to help miRNA antagomirs or agomirs perform their respective tasks once delivered within a target cell [100]. The ideal delivery of such miRNA-based drugs into a cell ensures the protection of miRNA from endogenous RNases while also preventing adverse reactions from innate immune responses against the RNA cargo. With recent advances in RNA delivering technologies [100][101] and chemical modifications that confer enhanced protection to nucleotides [102], miRNA-based drugs might potentially become mainstream in alleviating multifactorial diseases such as IBD.

Table 1:

Differentially expressed miRNAs and their putative roles in human or mouse inflammatory boweI disease pathogenesis

miRNA	Disease (Change in expression)	Sample type	Target protein	Reported effects/References
miR-301a	UC (↑)	Tissue	BTG1	miR-301 downregulated the expression of E-cadherin (CDH1), responsible for barrier integrity in human IECs in a BTG anti-proliferation factor 1 (BTG1)-dependent manner [109].
miR-206a	UC (↑)	Tissue	A3AR	miR-206a inhibited adenosine A3 receptor (A3AR) that induced the expression of pro-inflammatory factors such as NF-κB, IL-1β in human cells [110].
miR-31	UC (↑)	Tissue	IL-13RA1	miR-31 directly targeted anti-inflammatory expression of IL13 and IL13RA1 mRNAs in human IECs [111].
miR-138-5p	UC (↑)	Tissue	CLDN19*,	miR-138-5p is predicted to directly target CLDN19 from tight junctions that are important in intestinal barrier integrity [112].
miR-142-5p	UC (↑)	Tissue	SOCS1	miR-142-5p inhibited the expression of SOCS1 and upregulated pro-inflammatory cytokine IL-8 in human IECs [113].
miR-378-3p	UC (↓)	Tissue	IL-33	miR-378-3p directly targeted IL33 mRNA in human IECs [114].
miR-301a	UC (↑), CD (↑)	Blood	SNIP1	SNIP1 identified as a functional target of miR-310a; SNIP1 knockdown in patient-derived CD4+ T cells promoted Th17 cell differentiation via upregulation of IL17 and TNF-α [115].
miR-143	CD (↑)	Tissue	ATG2B	miR-143 directly targeted autophagy related 2B (ATG2B) mRNA, inhibiting autophagy and enhancing the expression of pro-inflammatory IL-8, TNF-α, and IFN-γ in human IECs [116]
miR-196	CD (↑)	Tissue	IRGM	miR-196 expression correlated with downregulation of the protective variant of immunity related GTPase M (IRGM) gene in CD patients [117].
miR-130a-3p	CD (↑)	Tissue	Atg16l1	miR-130-3p inhibited autophagy in mouse macrophages in vitro by targeting autophagy-related protein 16-like 1 (ATG16L1). Reduced autophagy in macrophages can lead to prolonged inflammation [103].
miR-144-3p	CD (↑)	Blood	N/A	Yet to be defined [118].
miR-15a-5p, miR-16-5p, miR-24-3p	CD (↑)	Feces	N/A	Yet to be defined [3].
miR-10a-5p, miR-141-3p, let-7g-5p, miR-375	CD (↓)	Feces	N/A	Yet to be defined [3].

CD: Crohn’s disease; UC: Ulcerative colitis; ↑: upregulation; ↓: downregulation; IEC: intestinal epithelial cell

^*Predicted targets

Emerging preclinical studies in mouse models of IBD support the feasibility of miRNA-based therapies. For example, the overexpression of miR-146a via the miR-146a mimetic recently ameliorated DSS-colitis in mice by limiting IL-17 signaling [47]. Similarly, miR-31 reduced the severity of DSS-induced colitis by targeting the 3’ untranslated regions of Il17 mRNA in mice [16]. MiR-142a-3p can ameliorate murine colitis by promoting the growth of Lactobacillus reuteri [38]. Moreover, miRNA antagomirs for miR-155 [86][87], miR-130a-3p [103], miR-31 [104] have been reported to improve murine colitis symptoms by targeting the Th1/Th17 axis, while an miR-214 antagomir ameliorated murine colitis by suppressing the NF-κB pathway [105]. Also, an miR-423-5p antagomir enhanced barrier function and ameliorated mouse DSS-induced colitis in mice via the derepression of Claudin 5 from the pore pathway [23]. These pre-clinical studies in mice underscore the potential of certain miRNAs as possible therapeutic targets in IBD. However, the scope and efficacy of miRNA therapeutics in IBD patients remains to be rigorously established.

Concluding remarks

IBD is projected to become a major health and policy concern in the near future necessitating concerted efforts to build a complete understanding of IBD etiology and focusing on environmental as well as genetic factors that maintain gut homeostasis: from microbiota to the intestinal epithelium, to PRR signaling, and immune activation. Certain miRNAs, with their multi-target nature, are increasingly being associated with IBD and are being recognized as putative targets for therapeutical interventions based on pre-clinical studies. While the multi-target nature of miRNAs has been proposed as a potential solution to the observed alternative pathway-mediated non-response in IBD patients, it can also cause unwanted side effects in patients by targeting other unknown pathways—a major concern related to miRNA-based therapies. This is also reflected in the fact that while siRNA drugs (siRNA being the second and highly-specific of the two kinds of RNAi effector molecules (the first: miRNAs)) have 60 candidates under trial with 2 already being approved by the USA FDA, only 10 are miRNAs, none of which have moved to phase 3 trials [106]. The successful application of miRNA-based drugs will require a complete understanding of all possible pathways that candidate miRNAs can target to minimize off-target effects. Such off-target effects might also be reduced by tissue-specific, targeted delivery of miRNAs using vehicles such as surface-engineered extracellular vesicles [107] or 3D-scaffolds such as nanoparticle hydrogels [100]. Although many questions remain (see outstanding questions), future studies should leverage the advances in miRNA drug delivery together with the power of algorithms in accelerating miRNA target prediction and validation [108] to harness the full potential of candidate miRNA therapeutics in IBD.

Outstanding questions.

• Which miRNAs mediate or modulate the complex crosstalk across different components of gut homeostasis? So far, miR-146 and miR-155 have been reported to modulate multiple pathways affecting gut homeostasis. Identifying other miRNAs that might also be clinically relevant can accelerate the putative applicability of miRNA-based therapeutics.

• How can we efficiently predict and validate possible miRNA targets (specific as well as non-specific), and their therapeutic potential, as well as better predict possible benefits and/or side effects? This is vital for minimizing potential risks involved in preclinical and clinical trials.

• Can miRNA profiles be used to accurately predict remissions in IBD patients and help improve their quality of life? Can this also help lower the probability of progression of colitis-associated colorectal cancers in IBD patients?

• How long-lasting are the effects of miRNA-based therapies? What are the best ways and routes of administration for such putative therapies?

Highlights.

• IBD pathogenesis is poorly understood due to its multifactorial nature comprising genetic as well as environmental factors, making miRNAs promising therapeutic candidates.

• MiRNAs can modulate multiple aspects of mammalian gut homeostasis (various at a time, simultaneously), ranging from intestinal barrier function, to microbial homeostasis, to damage or pathogen recognition, activating innate and adaptive immune responses.

• MiRNAs have been recently implicated in shaping the gut microbiota of mice and/or humans, where intestinal epithelial cell miRNAs can promote the growth of certain bacteria and vice versa.

• Emerging preclinical studies in mouse models of IBD are demonstrating the potential of candidate miRNAs-based therapies, such as those involving antagomirs (miRNA inhibitors) and agomirs (miRNA mimietics).

Significance.

The existing treatments for inflammatory bowel disease can lead to acquired resistance via alternative signaling pathways. With their multiple roles in gut homeostasis, certain microRNAs might constitute another candidate therapeutic approach.

Glossary

Stenosis

pathological abnormality marked by narrowing of blood vessels due to lesions, resulting in pain and inflammation

RNA interference (RNAi)

conserved post-transcriptional mechanism of suppressing the expression of double-stranded RNA in a sequence-specific way that can be activated by two effector small RNA molecules: microRNA (miRNA) and small interfering RNA (siRNA)

Dicer and Drosha

Part of RNAse III family; Drosha and Dicer cleave kb-long primary miRNAs and 60–90 nucleotide-long pre-miRNAs, respectively, giving rise to 20–22 nucleotide-long mature miRNAs

LGR5⁺ intestinal stem cells (ISCs)

most rapidly proliferating adult stems cells residing in intestinal crypts and marked by the G-protein coupled receptor LGR5

Tight junctions (TJs)

multi-protein junctional complexes forming a tight seal between adjacent intestinal epithelial cells; they monitor paracellular transport

Pore and leak pathways

in epithelial tight junctions, determine paracellular transport. The pore pathway is charge and size selective (< 8 Å), unlike the leak pathway

Adherens junctions (AJs)

Positioned immediately below the tight junction, AJs are characterized by intercellular E-cadherin chains that are crucial for the initiation and maintenance of intercellular adhesion

Desmosomes

mediate strong intercellular adhesion between adjacent cells and provide firm support to the epithelium during extreme mechanical stress

Pathogen-associated molecular patterns (PAMPs)

evolutionarily conserved small molecular motifs associated with pathogens or microbes; in general, recognized by a host’s immune system

Danger-associated molecular patterns (DAMPs)

molecules released by cells to signal injury or infection and mount appropriate immune response

Pattern recognition receptors (PRRs)

proteins recognizing specific molecular motifs on the surface of pathogens, damaged or senescent cells, or apoptotic host cells

Toll-like receptors (TLRs)

single-pass membrane-spanning receptors expressed by dendritic cells, macrophages, and most intestinal epithelial cells; help recognize conserved molecular motifs on microbes to activate innate immunity in the host

NOD-like receptors (NLRs)

cytosolic receptors recognizing molecular signatures associated with pathogens or dying cells that enter the cell via phagocytosis