Dysregulation of Inflammasome Priming and Activation by MicroRNAs in Human Immune-Mediated Diseases

Inflammasomes are endogenous multimeric protein complexes that mediate the cleavage of proinflammatory cytokines, especially IL-1β and IL-18, in response to a diverse set of stimuli such as infections, tissue damage, and metabolic imbalances (1, 2). Mutations in genes encoding the inflammasome sensor molecules NLRP3 and pyrin cause autoinflammatory disorders involving an excessive activation of innate immune responses, namely cryopyrin-associated periodic syndrome and familial Mediterranean fever (3). Moreover, changes in inflammasome activity and closely related dysregulation of IL-1β and IL-18 expression levels have been linked to cancer (4) and various human disorders involving autoimmune processes, such as multiple sclerosis (MS) (5) and rheumatic diseases (6).

MicroRNAs (miRNAs) are small noncoding RNAs of a length of ∼22 nt that act as posttranscriptional regulators of gene expression by binding complementary target sites in other transcripts (7). Several miRNAs can target mRNA transcripts of genes coding for components of inflammasome complexes or for regulators of the inflammasomes. Moreover, dysregulations of these miRNAs cause disease-like phenotypes in cellular and rodent models.

In the past 7 years, these mechanisms have been studied and important functional insights have been gained. This review aims to provide an overview of this emerging research field on miRNA–inflammasome interactions and to elucidate the potential of inflammasome-targeting miRNAs as biomarkers and for the development of novel therapeutic approaches for autoimmune diseases.

Several canonical inflammasomes named after their NLRP1, NLRP3, NLRC4, AIM2, and pyrin subunits and noncanonical inflammasomes named based on their caspases caspase-4, caspase-5, and caspase-8 have been described (1, 2). The common function of the canonical inflammasomes is the proteolytic activation of IL-1β and IL-18 and their secretion (2). Noncanonical inflammasomes have been proposed to realize additional functions such as the mediation of programmed cell death (pyroptosis) and the initiation of NLRP3 assembly (8, 9). Moreover, the inflammasomes control the microbiome in humans, both by hindering invasion of pathogens and by regulating the composition of commensal bacteria (10). In mouse models, deficiency of inflammasome components enhances the susceptibility to inflammatory bowel diseases (IBD) (10). Furthermore, several therapies that are currently in use to treat autoimmune disorders, such as IFN-β and glatiramer acetate in MS, glucocorticoids in systemic lupus erythematosus (SLE), and the TNF-α inhibitor infliximab in rheumatic diseases, partially affect the regulation of inflammasomes (11–14).

The NLRP3 inflammasome, first described by Tschopp and colleagues (15) in 2002, is currently best understood. It is composed of the sensor molecule NLRP3, the adapter protein ASC that enhances complex formation, and of pro–caspase-1. Upon assembly, the activated inflammasome converts pro–caspase-1 into caspase-1, which cleaves pro–IL-1β and pro–IL-18 into mature proinflammatory cytokines IL-1β and IL-18 (15). The proinflammatory cytokines are then secreted (16).

The inflammasome components ASC and pro–caspase-1 are widely expressed in many tissues and in cell types of both the hematopoietic and the nonhematopoietic lineages (17). However, the inflammasome sensor molecules exhibit a more differentiated expression pattern, suggesting that in distinct microenvironments, various inflammatory stimuli are sensed and transduced, eliciting context-dependent responses (17). In the brain, and in particular the microglia, NLRP3 mRNA levels are among the highest throughout human tissues (18, 19). In resting cells, the expression of NLRP3 is not sufficient for inflammasome assembly, and a priming step leading to the upregulation of NLRP3 and pro–IL-1β expression is required. Once the cellular NLRP RNA level reaches an activating threshold, inflammasome assembly can be triggered (20). Pathways that prime NLRP3 and pro–IL-1β expression and activate inflammasome assembly have been excellently described in other reviews and are visualized in Fig. 1 (1, 2). One of the most effective mechanisms to mitigate the expression of NLRP3 is mediated by miRNAs, molecules that reduce NLRP3 protein expression through posttranscriptional downregulation of NLRP3 mRNA, as will be outlined in more detail in sections four and five.

miRNAs are ∼22-nt long noncoding RNAs that posttranscriptionally regulate the expression of genes by binding to complementary sequences in target transcripts (7). In the current release 22 of the database miRBase, 2883 known human mature miRNAs are listed (21). Experimental disruptions of miRNA genes amount to diverse abnormal phenotypes, including defects in the thymus, hematopoietic lineages, and brain, altered susceptibility to infections, and autoimmune-mediated diseases (7). For example, the miRNAs miR-20b-5p, miR-30a-5p, and miR-146a-5p are dysregulated in immunological and neurologic pathways in MS (22), and the miRNA levels in peripheral blood cells of MS patients change in response to IFN-β treatment (23).

The first step of canonical miRNA biogenesis is transcription into primary miRNAs (pri-miRNAs) by RNA polymerase II (24, 25). Pri-miRNAs form double-stranded hairpin structures with a characteristic loop. In a two-step process, the pri-miRNA is first cleaved into a precursor miRNA by Drosha/DGCR8 and additional RNA-binding proteins, translocated into the cytoplasm by exportin-5, and subsequently processed by Dicer/TRBP to form a short miRNA duplex (26–28). The resulting duplex is incorporated into an Argonaute protein and unwound so that only one strand (i.e., the mature miRNA) is kept in the RNA-induced silencing complex, which targets transcripts with matching sequences for degradation and prevents translation into proteins (25, 29, 30). Depending on which miRNA duplex strand is employed, the suffix “-3p” or “-5p” is used to designate the precise mature miRNA molecule (31, 32). Moreover, imprecise miRNA cleavage during their biogenesis as well as trimming, tailing, and editing have been described to produce isoforms of mature miRNAs with differing sequences (33).

Pairing of the miRNA seed region (nucleotides at position 2–7 from the 5′ end) and complementary sites preferably within 3′ untranslated regions (3′ UTRs) of target mRNAs mediates translational repression and destabilization of the bound transcripts (34). Additional matches of the transcript and the miRNA can augment the seed match as described by Bartel (7) in 2018. In humans, mRNA destabilization by deadenylation and subsequent exonucleolytic decay dominates the regulatory effect of miRNAs, whereas the contribution from direct translational repression is relatively small (35–39). As of today, more than 400,000 regulatory interactions are listed in miRTarBase (release 7.0), a database collecting miRNA–target interactions from the literature (40). Other databases, such as miRWalk2.0, can be used to predict miRNA–target interactions based on different algorithms that, for example, employ sequence complementarity, binding energy, and sequence conservation (41).

The strongest evidence for miRNA–target interactions is usually provided by reporter assays, followed by Western blots and quantitative real-time PCR (qPCR) analyses. Reporter assays detect changes in the expression level of a hybrid construct consisting of a 3′ UTR of a putative target gene and a reporter (e.g., luciferase) in response to miRNA overexpression or inhibition (42). This method provides reliable evidence for direct interactions. Approximately 7600 of the 400,000 interactions listed in miRTarBase are supported by reporter assays. Analyzing changes in RNA and protein expression upon modulation of miRNA levels with qPCR and Western blots can be used to validate putative functional interactions as well, although this approach does not prove direct miRNA–target interactions as it also captures secondary effects (42). Novel highly parallel approaches for the identification of miRNA–target interactions include cross-linking and immunoprecipitation methods coupled with next-generation sequencing (43–45) and posttranscriptional regulatory element sequencing (46). More than 350,000 of the 400,000 miRTarBase-listed interactions have only been identified by next-generation sequencing methods, indicating the need for further validation.

In 2012, two studies that were published in The Journal of Immunology identified miR-223-3p as the first human miRNA directly targeting NLRP3 (47, 48). A role for this interaction in the differentiation of monocytes to macrophages was proposed, as miR-223-3p levels decrease during differentiation, whereas NLRP3 protein levels increase (47). NLRP3 mRNA levels correlate negatively with miR-223-3p expression in myeloid cells (48). Interestingly, among human innate immune cells, miR-223-3p expression is highest in neutrophils and lowest in dendritic cells (48), hinting at a mechanism for why dendritic cells have a relatively low threshold for responding to NLRP3 stimuli (48). Moreover, a miRNA from EBV, a common virus that targets B cells and is considered essential for the development of the neurodegenerative disease MS (49), ebv-miR-BART15-3p, regulates NLRP3 expression (47). Infected B cells can secrete ebv-miR-BART15-3p via exosomes and inhibit NLRP3 activity in noninfected cells, which may be a mechanism for the virus to evade host defense (47). Interestingly, miR-223-3p and ebv-miR-BART15-3p hybridize with NLRP3 mRNA at the same site, and a common polymorphism, rs10802501, is present in the target site and could influence the binding of both miRNAs (47).

Meanwhile, several further miRNAs have been found to directly target transcripts of NLRP3 and thereby inhibit inflammasome priming in vitro and in vivo in human and animal studies (Table I). Among them are miR-7-5p and miR-30e-5p, and treatment with mimics of these miRNAs alleviates Parkinson disease (PD)-like phenotypes in mouse models involving accumulation of α-synuclein (50–52). The NLRP3 inflammasome and IL-1β production can be activated by α-synuclein, which drives neurodegeneration (53). NLRP3 is also a direct target of miR-20b-5p, upregulation of which alleviates Mycobacterium tuberculosis–induced inflammation in murine lung tissue (54). Additionally, miR-22-3p downregulates NLRP3 expression, and this miRNA is decreased in injured coronary arterial endothelial cells in a rat model of high-fat diet–induced coronary heart disease, whereas upregulation of miR-22-3p exerts a protective effect on cell survival (55). The expression of miR-133b-3p, which also targets NLRP3, is reduced in the nasal mucosa of a mouse model of allergic rhinitis (56). Allergic symptoms of the mice, immune cell infiltration, cytokine levels, and NLRP3 inflammasome protein expression were reduced in the nasal mucosa after administration of modified miR-133b-3p mimics, indicating that miR-133b-3p downregulation may contribute to the pathogenesis of allergic rhinitis, whereas its upregulation shows therapeutic potential (56). In a mouse model of prosopalgia, miR-186-5p is downregulated, and injection of miRNA mimics could alleviate neuropathic pain behavior by downregulating NLRP3 expression and NLRP3 inflammasome activity (57). The levels of miR-495-3p are decreased in mice with myocardial ischemia/reperfusion injury, and by targeting NLRP3, treatment with miR-495-3p mimics alleviates inflammation (58). The predicted secondary structure of the NLRP3 mRNA 3′ UTR and the proposed binding sites of the nine miRNAs that so far have been shown to target NLRP3 are displayed in Fig. 2.

In regard to miRNA regulation, NLRP3 is the best investigated inflammasome component. Beyond that, two caspases are directly targeted by miRNAs. In human epithelial cells, overexpression of miR-214-3p downregulates caspase-1, and the miRNA is downregulated in cataract lens anterior capsular tissues (59). MiR-371a-5p, miR-373-3p, and miR-543-3p have been proposed to enhance TNF-α–induced cell death by targeting the noncanonical caspase-8 inflammasome, and they are overexpressed in hepatocellular carcinoma tissue (60). However, the miRNA-driven regulation of further human inflammasome components, such as other NLR sensor molecules, AIM2, caspase-4, and caspase-5, has so far not been investigated in specific experimental studies. Caspase-11, the murine ortholog of human caspase-4 and caspase-5, was shown to be indirectly regulated by miR-34a, miR-451, and miR-874 (61).

Inflammasomes are regulated by various upstream pathways. In several recent studies, it has been demonstrated that miRNAs that target these upstream pathways alter the expression levels of components of the NLRP3 inflammasome. As presented in Fig. 1 and listed in Table II, some miRNAs target more than one inflammasome regulator and likewise, a regulator can be downregulated by several miRNAs. The miRNA-mediated modulation of inflammasome regulation was investigated in very heterogeneous research contexts, such as gout, depression, and alcoholic liver disease.

Several miRNAs negatively regulate the NF-κB–mediated priming of the NLRP3 inflammasome by targeting a wide range of transducers of TLR, IL-1R, and TNFR signaling. Signals from both TLR4 and IL-1R are transduced via MyD88, IRAK1, and IRAK4, TRAF6, IKK, and finally NF-κB, which then induces NLRP3 and pro–IL-1β transcription (2, 62). In a study on inflammation of gout, miR-302b-3p was upregulated after induction of THP-1–derived macrophages using monosodium urate (MSU). Overexpression of miR-302b-3p directly downregulates IRAK4 and thereby inhibits NLRP3 expression (63). Another research group investigated the function of miR-146a-5p in response to MSU stimulation (64). The mRNA and protein expression of TRAF6, IRAK1 and downstream NLRP3, ASC, and caspase-1 were upregulated in MSU-induced bone marrow–derived macrophages of miR-146a-5p knockout mice as compared with wild-type mice. The miR-146a-5p–deficient mice developed a more severe phenotype of gouty arthritis (64). Two further members of the TLR4 and IL-1R signaling pathways are targeted by miRNAs: MyD88 is downregulated by miR-155-5p overexpression in rat prefrontal cortex, and this alleviates depressive-like behaviors (65), and IKKγ (gene symbol: IKBKG) expression is reduced by miR-34c-5p. In response to stimulation with damage-associated molecular patterns, miR-34c-5p is highly expressed in PBMC (66). Through this mechanism, miR-34c-5p was proposed to limit aberrant sterile acute or chronic inflammation (66).

IL-1R2, a receptor that inhibits IL-1 signaling in a competitive manner (67), is downregulated by miR-383-3p (68). In myocardial tissues of a rat model of atherosclerosis, miR-383-3p levels are reduced, whereas IL-1R2 expression is increased. Overexpression of miR-383-3p decreases the protein levels of caspase-1 and the secretion of IL-1β and IL-18 in homocysteine-induced endothelial injury in rat coronary arteries (68).

Binding of TNF-α to TNFRSF1A followed by downstream TRAF2/SphK1/S1P signaling constitutes another NF-κB–activating and NLRP3 inflammasome-priming pathway (69). Administration of miR-330-3p mimics attenuates fructose-induced upregulation of SphK1, NF-κB, NLRP3, pro–IL-1β, and IL-1β protein levels, whereas miR-330-3p inhibition has the opposite effect in a proximal tubule epithelial cell line from adult human kidney (HK-2 cells) (70). By targeting SphK1, miR-330-3p thus inhibits inflammasome priming (70).

The receptor IFNAR responds to IFN-α and IFN-β and activates the JAK/STAT/ISGF3 pathway, which is able to drive NLRP3 expression (71). JAK1 is targeted by miR-9-5p, and this miRNA is increased in human primary peripheral blood monocytes stimulated with oxidized low density lipoproteins indicating its role as a regulator of inflammation (72, 73). Overexpression of miR-9-5p inhibits NLRP3 inflammasome priming in atherosclerosis inflammation models using human THP-1–derived macrophages and primary peripheral blood monocytes (72).

During mitochondrial stress, excessive reactive oxygen species (ROS), which can be reduced by superoxide dismutases (SOD), are generated, and they induce TXNIP, an NLRP3-activating factor (74). Fructose induces miR-377-3p expression in kidney cortex and glomeruli of rats as well as in a mouse podocyte cell line (75). Upregulation of miR-377-3p can inhibit SOD1 and SOD2 protein levels and thereby raise ROS levels and induce elevated NLRP3, ASC, caspase, and supernatant IL-1β protein levels in cultured mouse podocytes, implying a role for miR-377-3p in metabolic syndrome (75). The NLRP3-activating factor TXNIP is inhibited by several individual miRNAs. Expression levels of miR-17-5p in rat brains and rat insulin-producing cells with endoplasmic reticulum stress (76, 77), of miR-20a-5p in rat fibroblast-like synoviocytes in adjuvant-induced arthritis (78), and of miR-148a-3p in mouse hepatocytes in alcoholic hepatitis (79) are decreased and correlate with upregulation of TXNIP and activation of the NLRP3 inflammasome. Upregulation of the miRNAs with mimics reversed these effects. The chemokine receptor CXCR4 was proposed to stabilize TXNIP (80). In mice with sciatic nerve injury, overexpression of miR-23a-3p downregulates CXCR4, inhibits the activation of the NLRP3 inflammasome and subsequent IL-1β secretion, and alleviates neuropathic pain behavior (80).

Two further miRNAs are known to modulate inflammasome regulation. In an atherosclerosis model of human aortic endothelial cells treated with oxidized low density lipoproteins, miR-30c-5p is downregulated, leading to the upregulation of FOXO3 (81). The administration of miR-30c-5p mimics decreases FOXO3 expression, and the knockdown of FOXO3 downregulates NLRP3, caspase-1, and IL-1β mRNA and protein levels, indicating a protective effect of the miRNA (81). The promotor of miR-145a-5p can be hypermethylated during atherosclerotic plaque formation, which leads to a decreased expression of miR-145a-5p, an upregulation of its target genes CD137 and NFATc1, and this induces the activation of the NLRP3 inflammasome. An elevation of the miRNA levels has a protective effect on vascular smooth muscle cells (82).

Of the described miRNAs, miR-9-5p (83) and miR-155-5p (84) also directly target NFKB1 mRNA, and miR-7-5p (85) downregulates RELA expression. Both are subunits of NF-κB. However, for these interactions, the dysregulation of the inflammasomes was not investigated in response to miRNA dysregulation.

At least five of the above mentioned NLRP3 inflammasome-targeting miRNAs and seven miRNAs that modulate inflammasome regulation are associated with autoimmune diseases (Table III). In the following, miRNAs involved in MS, rheumatoid arthritis (RA), SLE, IBD, and type 1 diabetes (T1D) will be addressed. The most prominent roles for inflammasome-regulating miRNAs in autoimmune diseases have been demonstrated for miR-146a-5p, miR-155-5p, and miR-223-3p, all of which are dysregulated in RA, MS, and SLE.

In PBMCs of patients with the immune-mediated neurodegenerative disease relapsing-remitting MS (RRMS), miR-146a-5p levels are increased compared with healthy controls, and treatment with glatiramer acetate partially restores miR-146a-5p expression to lower levels again (86, 87). Interestingly, the C allele of the single-nucleotide polymorphism (SNP) rs2910164 in the sequence encoding mir-146a is associated with an increased expression of this miRNA and release of proinflammatory cytokines in patients with RRMS, and in females, rs2910164 was proposed to alter MS susceptibility (88). Furthermore, miR-155-5p is expressed more strongly in peripheral blood monocytes and active brain lesions of MS patients than in healthy controls (87, 89). Enhanced miRNA levels have been demonstrated for miR-22-3p in plasma of MS patients (90) and for miR-9-5p in circulating leukocytes of RRMS patients (91). Other miRNAs are decreased in MS patient serum (e.g., miR-23a-3p and miR-223-3p in primary progressive MS) (92). The T allele of SNP rs1044165 located near the sequence encoding mir-223 is associated with a higher risk of developing MS (93). Furthermore, in RRMS, miR-214 levels are downregulated in T cells in patients in the relapsing phase compared with patients in the remitting phase and healthy controls (94). In response to treatment of RRMS patients with natalizumab, the expression of miR-17-5p in T cells is decreased in comparison with untreated patients, and during relapse, miR-17-5p is upregulated (95). In case of miR-7-5p and miR-20b-5p, expression is upregulated in T cells during natalizumab treatment (96).

In RA, miR-22-3p levels are reduced in synovial tissue from RA patients compared with osteoarthritis patients (97), and low expression of miR-22-3p is associated with a good response to adalimumab treatment (98). Furthermore, the expression of miR-9-5p is decreased in plasma of RA patients compared with healthy controls (99). The three miRNAs miR-223 (100), miR-155-5p (101), and miR-146a (102) are upregulated in synovial tissue from RA patients compared with osteoarthritis synovium, and two of them, miR-155-5p and miR-146a-5p, are elevated in PBMCs of RA patients (103, 104). Moreover, miR-223-3p is overexpressed in T cells, and miR-223-3p expression correlates with the titer of rheumatoid factor in patients with RA (105). Of interest, levels of miR-146a-5p are reduced in regulatory T cells of RA patients with active disease and correlate with joint inflammation (106). Furthermore, the rs2910164 genotype GG in the sequence encoding mir-146a is associated with an increased risk of developing RA in females, whereas the CC genotype is related to a decreased disease severity in females and an increased disease severity in males (107).

In patients with SLE, miR-155-5p expression is decreased in PBMCs compared with healthy controls (108). In SLE urine, miR-155-5p levels are significantly higher, treatment with calcitriol reduces miR-155-5p levels, and miR-155-5p levels correlate with proteinuria and disease activity, whereas the amount of miR-146a-5p correlates with the estimated glomerular filtration rate in SLE patients (108). In response to treatment with mycophenolic acid, miR-146a-5p levels are increased in T cells of patients versus untreated patients (109). Additionally, two SNPs upstream of the genome locus encoding mir-146a, rs57095329 and rs2431697, are associated with the miRNA expression level and with SLE development (110, 111). Several inflammasome-regulating miRNAs are downregulated in SLE: miR-223-3p in naive B cells of SLE patients (112), miR-20a-5p in monocytes (113), and miR-17-5p in PBMCs of adult (114) and pediatric SLE patients (115). Moreover, miR-17-5p levels are diminished in untreated patients compared with those on treatment (116). Regarding lupus nephritis, a common manifestation of SLE, higher levels of miR-146a in glomeruli and miR-148a-3p in serum and glomeruli of patients with lupus nephritis have been reported relative to healthy controls (117, 118).

In IBD patients, some inflammasome-modulating miRNAs are also dysregulated. For instance, miR-155-5p is overexpressed in colonic mucosa of IBD patients compared with non-IBD controls (119). The levels of miR-23a-3p are elevated in the blood of patients with Crohn disease in comparison with healthy controls (120). Furthermore, miR-146a expression is increased in colonic mucosa of Crohn disease patients compared with ulcerative colitis and healthy controls, and it is increased in colonic tissue in both Crohn disease and ulcerative colitis (121).

For T1D, miR-146a-5p is so far the only inflammasome-regulating miRNA reported to be downregulated in PBMCs of patients. Low miR-146a-5p expression correlates with ongoing islet autoimmunity in T1D (122).

Additionally, inflammasome-modulating miRNAs have been found to be dysregulated in PD, and although defined as a neurodegenerative disease, PD has been suggested to involve autoimmune (123) and inflammasome-mediated mechanisms as well (124). The expression level of miR-7-5p is decreased in the serum of patients with PD (52). In circulating lymphocytes of treated patients, miR-7-5p and miR-9-5p are upregulated compared with untreated controls (125). Moreover, miR-22-3p is downregulated in whole blood of newly diagnosed and in treated PD patients compared with healthy controls (126), and miR-34c-3p is downregulated in brain areas of PD patients with neuropathological lesions (127).

Since their first description in 1993 (128), miRNAs have emerged as important regulators of gene expression. Some miRNAs are key to human development; others can mediate diseases but the function of most miRNAs is still barely understood. A decade after the discovery of inflammasomes, the first miRNAs that control inflammasome components have been identified (47, 48). Since then, several miRNAs that target NLRP3, caspase-1, caspase-8, and upstream signaling cascades via IL-1R, TLRs, IFNAR, TNFR, and mitochondrial stress pathways have been demonstrated to modulate inflammasome activity and to be dysregulated in human immune-mediated diseases.

However, the network of possible miRNA–inflammasome interactions has only been touched upon. The miRNA-mediated posttranscriptional control of other inflammasome components such as ASC, NLRP1, NLRC4, and AIM2, noncanonical inflammasomes, and upstream pathways has not been thoroughly investigated so far. For many miRNAs, bioinformatics-based miRNA–target interaction predictions in the database miRWalk2.0 (41) point at a role in inflammasome regulation; for example, miR-520f-3p is predicted to target NLRP3 by 7 out of 12 algorithms. Valuable insights into autoimmune mechanisms may be gained by investigating such putative interactions experimentally at the RNA and protein level and with luciferase reporter assays that allow for the assessment of direct interactions. Not only human miRNAs may be interesting to study further, but so too could miRNAs from viruses, which have been associated with diseases, as was already shown for ebv-miR-BART15-3p from EBV (47). Moreover, for many miRNAs, which exact mature miRNA isoforms act as posttranscriptional regulators is poorly understood. Finally, several underexplored miRNAs may be expressed at high levels only under certain conditions (e.g., in specific immune cell subtypes or during inflammation) (129–131). Investigating in more detail the expression and function of miRNAs in compartments that are especially relevant for diseases, like microglia for MS, may be valuable for a better understanding of the mechanisms underlying pathologic immune responses.

Several of the inflammasome-targeting miRNAs are associated with and relevant to the development of autoimmune diseases such as MS, RA, and SLE in humans. This suggests that miRNAs may also be useful as biomarkers in immune-mediated disorders. In particular, miR-146a-5p, miR-155-5p, and miR-223-3p are dysregulated in diverse autoimmune diseases, and SNPs close to or within the pri-miRNA–coding sequences are associated with disease risk and severity. Exploring these miRNAs and also other miRNAs and the respective SNP alleles may further increase our understanding of the genetics underlying complex autoimmune disorders. Likewise, SNPs in the target sequences in mRNA 3′ UTRs can alter their regulation by miRNAs and thereby also affect disease risk. Regarding the use of mature miRNA molecules as biomarkers, combinations of several miRNAs in a biomarker panel can yield better predictions, as was already shown for the combination of miR-142-3p, miR-146a-5p, and miR-155-5p for the stratification of RRMS (87, 132). Most of the studies described in this review found dysregulated miRNA levels in blood compartments and some in pathologic tissues of patients with autoimmune diseases. Serum and plasma are easily accessible biological materials, and exploiting miRNAs from these compartments as biomarkers is feasible. It may be interesting to investigate whether miRNAs with biomarker potential that have only been detected in certain cell types and tissues are differentially expressed in serum and plasma as well when employing sensitive methods. Still, for many miRNAs discussed in this review, studies with large patient cohorts are required to evaluate whether they can indeed serve as reliable biomarkers for disease diagnosis, as well as for prognosis and for monitoring the individual response to therapy and whether the accuracy can be improved by combining different miRNAs. Such efforts should also shed light on to the role of these miRNAs in the pathomechanisms underlying many autoimmune diseases but also autoinflammatory diseases such as cryopyrin-associated periodic syndrome and familial Mediterranean fever.

Finally, in animal disease models, the modulation of miRNA expression levels (e.g., miR-7-5p and miR-30e-5p in PD mouse models) (50–52) already showed promising therapeutic effects, but this potential has yet to be translated into the clinical setting. Novel miRNA therapeutics have been developed in the past few years, but some therapeutic miRNA studies have already been terminated during the preclinical stage because of side effects involving liver damage and poor miRNA stability (133). However, miRNA drug delivery, efficacy, and safety issues are currently being addressed, and recently, the first miRNA-based therapeutic, miravirsen, an inhibitor of miR-122-5p, yielded positive results in a phase II clinical trial (134). Further optimization of miRNA therapeutics may yield a wider spectrum of successful miRNA drugs in the years to come, possibly including inflammasome-modulating miRNAs for the treatment of autoimmune disorders.

N.B. received speaking fees and travel funds from Novartis. M.H. received speaking fees and travel funds from Bayer HealthCare, Biogen, Novartis and Teva. U.K.Z. received research support as well as speaking fees and travel funds from Almirall, Bayer HealthCare, Biogen, Merck Serono, Novartis, Sanofi Genzyme, and Teva.