1 Introduction

Since the first microRNA (miRNA), lin-4, was discovered in Caenorhabditis elegans by the laboratories of Victor Ambros and Gary Ruvkun in 1993 (Lee et al. 1993; Wightman et al. 1993), these small non-coding RNA species have been extensively studied for their roles in post-transcriptional regulation of gene expression. Like other protein encoding genes, primary miRNA transcripts are first transcribed in the nucleus but are sequentially processed by a microprocessor complex, which contains the RNase III Drosha and the double-stranded RNA binding protein DGCR8 in mammals and Pasha in other species, into a characteristic ~60 nucleotide stem-loop structure, named precursor-miRNA (pre-miRNA) (Lee et al. 2003; Han et al. 2006; Denli et al. 2004; Gregory et al. 2004). The pre-miRNAs are then exported via exportin 5 (Lund et al. 2004; Yi et al. 2003) into the cytoplasm, where the hairpin of pre-miRNA is further processed by the evolutionarily conserved RNase III enzyme Dicer to generate mature miRNAs (Bernstein et al. 2001; Hutvagner et al. 2001). These mature miRNAs are capable of being incorporated into the RNA-induced silencing complex (RISC) where they interact with the core component Argonaute protein (Hutvagner and Zamore 2002; Lingel et al. 2003; Liu et al. 2004). Once assembling with mature miRNA and engaging the targets, active RISCs recognize complementary messenger RNA (mRNA) transcripts for degradation or translational silencing (Jing et al. 2005; Song et al. 2004; Liu et al. 2004). By modulating the expression level of their target genes, miRNAs have been found to regulate almost all biological processes including cell growth, differentiation, and apoptosis in both plants and animals (Ameres and Zamore 2013). As is the case for non-immune cells, the role of miRNA-mediated gene regulation in the immune system in recent years has become a focus of intense investigation.

2 miRNAs in the Immune System

Even though our understanding of miRNA biogenesis and its overall cellular function was initially established based on discoveries in nematodes and plants (Lee et al. 1993; Baulcombe 2004), accumulating evidence has demonstrated that miRNAs also have a crucial role in controlling all aspects of immune responses (Mehta and Baltimore 2016). The importance of miRNA in immune cells was first demonstrated in studies where key components in miRNA biogenesis were disrupted. To this end, conditional deletion of Dicer during early B cell development leads to a dramatic block at the pro- to pre-B transition (Koralov et al. 2008). Subsequent studies using mice with Dicer ablated specifically in peripheral activated B cells further demonstrated that miRNAs are also required for germinal center B cell formation and the generation of the antibody diversity (Xu et al. 2012). Interestingly, unlike what has been reported in B cell differentiation, Dicer ablation in early T cell progenitors did not exhibit any substantial alterations in thymic T cell development with the exception of reduced thymic cellularity (Cobb et al. 2005). Nevertheless, the specificity of the role of miRNA in peripheral T cells is much more apparent, as T cells lacking Dicer failed to differentiate into multiple helper T cell lineages and exhibited aberrant effector T cell function (Muljo et al. 2005). Moreover, when Dicer or Drosha ablation was restricted to regulatory T (Treg) cell lineage, mice developed highly aggressive autoimmunity comparable to those devoid of a functional Foxp3 gene (Zhou et al. 2008; Liston et al. 2008; Chong et al. 2008), suggesting an indispensable role of miRNA in controlling Treg cell biology.

To date, more than hundreds of miRNAs have been reported to be differentially expressed in immune cells. Distinct miRNA signatures not only were found in individual immune cell lineages but could also be detected in the same cell subsets that are in different developmental stages. For example, whereas miR-139 is highly expressed in Pro- and Pre-B cells, elevated levels of miR-28, miR-320 and miR-148a are detected in germinal center (GC) B cells and plasma cells, respectively (Kuchen et al. 2010). Moreover, expression of miRNAs in immune cells can also be dynamically regulated in response to a variety of stimuli, such as antigens recognized by T or B cell receptors, proinflammatory cytokines, and microbial components that trigger Toll-like receptors (O’Connell et al. 2007; Taganov et al. 2006; Cobb et al. 2006). To this end, a recent study reported that T cell activation induces proteasome-mediated degradation of Argonaute, and subsequently causes a global down-regulation of mature miRNAs (Bronevetsky et al. 2013). It was suggested that activation-induced miRNA down-regulation confers effector functions to helper T cells via relaxing the repression of genes that direct T cell differentiation. Finally, hierarchical clustering analysis of miRNA profiling clearly separated cells of the immune system from other tissues. Taken together, these results implied certain miRNAs might play a specific role in controlling development and effector functions of the immune system (Kuchen et al. 2010), and that miRNAs need to be tightly regulated as aberrant expression of miRNAs often leads to dysregulated innate and adaptive immunity.

3 miRNAs as Negative Regulators of Immune Responses

3.1 miRNA in Adaptive Immunity

While studies of mice with B or T cell-specific deletion of the entire miRNA pathway seemed to suggest that miRNAs generally play a positive role in promoting adaptive immunity as discussed in the previous section, many miRNAs have been identified as important negative regulators in restricting B and T cell responses (Fig. 1). For example, miR-150, a miRNA that is predominantly expressed in mature B cells was shown to control multiple B cell populations through regulating the expression level of transcription factor c-Myb (Xiao et al. 2007). Genetic ablation of miR-150 resulted in the expansion of B1 cells, one of the subsets of mature B cells, in spleen and peritoneal cavity. While the numbers of follicular B cells were not significantly altered, upon immunization elevated antibody responses could be easily detected (Xiao et al. 2007). Compared to miR-150, miR-210, a miRNA that is highly induced upon B cell activation, appears to play an even larger role in functioning as a negative feedback regulator to restrain B cell responses; deletion of miR-210 leads to the development of age-associated autoantibodies (Mok et al. 2013). In contrast, while miR-142 was shown to target B cell-activating factor receptor (BAFF-R), a molecule that is critical for B cell proliferation and survival, hypogammaglobulinemia phenotype was detected in mice devoid of miR-142 despite having increased follicular B cell numbers (Kramer et al. 2015). Similarly, despite the fact that activation-induced cytidine deaminase (AID), a potent enzyme critical for somatic hypermutation and class-switch recombination (CSR), has been shown to be a bona fide miR-155 target where lack of AID regulation by miR-155 led to defective affinity maturation (Teng et al. 2008; Dorsett et al. 2008), miR-155-deficient mice actually exhibited reduced germinal center function and failed to generate high-affinity IgG1 antibodies (Rodriguez et al. 2007). Together, these results suggested that other miR-142 or miR-155 targets are likely responsible for their respective effects on humoral immunity and further demonstrate the complex nature of miRNA-mediated immune regulation.

Fig. 1
figure 1

miRNAs negatively regulate adaptive immune cells. Specific miRNAs expressed by B or T cells repress key target genes that are involved in adaptive immune responses. Tfh cell, T follicular helper cell; Th cell, T helper cell; Treg cell, regulatory T cell

Full size image

Like B cell, activated T cells also express increased level of miR-155; mice devoid of miR-155 displayed increased lung airway remodeling, and miR-155-deficient CD4+ T cell cells are intrinsically biased toward Th2 differentiation in vitro. Mechanistically, it was shown that miR-155 can inhibit Th2 responses through modulating the level of c-Maf, a transcription factor known to promote Th2 immunity (Rodriguez et al. 2007). In addition to miR-155, we and others have recently demonstrated that miR-24 and miR-27, two members of the mi-R23 cluster family, collaboratively limit Th2 responses and associated immune pathology through targeting IL-4, GATA3 as well as other Th2-related genes in both direct and indirect manners (Cho et al. 2016; Pua et al. 2016). While miR-23 does not seem to play any role in Th2 regulation, it is indispensable for restraining activation-induced necrosis of CD4+ T cells by enforcing intracellular reactive oxygen species (ROS) equilibrium through targeting cyclophilin D, a regulator of ROS escape from mitochondria (Zhang et al. 2016). In addition to Th2 regulation, miRNAs have also been implicated in regulating other Th lineages. To this end, miR-29 was shown to control Th1 responses by repressing multiple genes associated with Th1 differentiation and function including both T-bet and Eomes, two transcription factors known to induce IFN-γ production and IFN-γ itself (Ma et al. 2011; Steiner et al. 2011). On the other hand, hypoxia-induced miR-210 was reported to negatively regulate Th17 responses through restricting the expression of HIF-1α, a key transcription factor that promotes Th17 polarization under limited oxygen (Wang et al. 2014). Similarly, Th17 differentiation and the resultant pathogenesis of experimental autoimmune encephalomyelitis (EAE), a mouse model of multiple sclerosis, could be suppressed by miR-20b via targeting RORγt and STAT3, two key Th17 transcription factors (Zhu et al. 2014). Finally, miR-146a has highly induced in T follicular helper (Tfh) cells, a specialized Th cell subset required for humoral immunity, and can act as a post-transcriptional brake to control Tfh cell and corresponding GC B cell responses by regulating ICOS-ICOSL axis (Pratama et al. 2015).

In addition to its role in Tfh cells, miR-146a has also been shown to function as a key molecular regulator to confer suppressor function to Treg cells. In the absence of miR-146a-mediated regulation of STAT1, mice succumbed to spontaneous IFN-γ-dependent Th1-mediated immunopathology (Lu et al. 2010). On the other hand, albeit dispensable for Treg cell suppressor function, Foxp3-dependent miR-155 ensures Treg cell competitive fitness through targeting SOCS1 (Lu et al. 2009, 2015). Moreover, our recent work also demonstrated that miR-27 controls multiple aspects of Treg cell biology and suggests that excessive expression of miR-27 in Treg cells resulted in a breakdown of Treg cell-mediated immunological tolerance (Cruz et al. 2017). Together with the aforementioned studies in which the entire miRNA pathway was ablated in Treg cells, these results suggested that miRNAs are able to mediate their regulatory effects on the immune system indirectly through maintaining optimal Treg cell function and homeostasis (Zhou et al. 2008; Liston et al. 2008; Chong et al. 2008).

3.2 miRNA in Innate Immunity

Similar to what was described in the adaptive immune system, significant progress has been made over the past decade in characterizing individual miRNAs that control the function of innate immune cells (Fig. 2). Among them, miRNA-mediated regulation of myeloid cell function is best characterized. Both miR-146 and miR-155 were identified in macrophages in response to LPS activation (Taganov et al. 2006; O’Connell et al. 2007). Between these two miRNA, miR-146a serves as a negative regulator to limit inflammatory responses by targeting TRAF6 and IRAK1 (Taganov et al. 2006; Boldin et al. 2011). During virus infection, miR-146a expression is also upregulated in macrophages in RIG-I-dependent manner, and its function negatively regulates type I interferon (IFN) production through repressing IRAK2 (Hou et al. 2009). While miR-155 is generally considered as a positive player in mediating inflammatory responses as it can directly repress SOCS1 and SHIP1 (Androulidaki et al. 2009; O’Connell et al. 2009), two known inhibitory molecules affecting multiple signaling pathways, miR-155 can also down-modulate TLR/IL-1 inflammatory pathway via targeting TAB2 (Ceppi et al. 2009). Moreover, protein kinase Akt1, which is activated by LPS in macrophages, positively regulates let-7e, a miRNA that can inhibit TLR4 expression to control endotoxin sensitivity and tolerance (Androulidaki et al. 2009). In addition to direct TLR4 targeting, miR-21, another miRNA induced by LPS can also act as a negative feedback regulator of TLR4 signaling by targeting PDCD4, a proinflammatory protein required for LPS-induced death (Sheedy et al. 2010). Finally, miRNAs can also restrict DC function by directly suppressing the production of proinflammatory cytokines. To this end, NOD2-induced miR-29 expression in human dendritic cells (DCs) was shown to inhibit IL-23 expression by repressing IL-12p40 directly and IL-23p19 indirectly via targeting ATF2 (Brain et al. 2013).

Fig. 2
figure 2

miRNAs negatively regulate innate immune cells. For miRNAs expressed in different innate immune cell populations, their key target genes are shown. NK, natural killer

Full size image

Besids myeloid cell subsets, much experimental evidence has also pointed to miRNAs as important negative regulators for other innate immune cells. For example, miR-223 has been shown to function as a cell intrinsic negative modulator of neutrophil activation and killing. Mice lack of miR-223 developed spontaneous inflammatory lung pathology and exhibited exaggerated tissue damage upon LPS treatment (Johnnidis et al. 2008). Moreover, miR-27a*, miR-378, and miR-30e have all been shown to act as negative regulators of NK cell cytotoxicity by silencing perforin and granzyme (GzmB) expression (Kim et al. 2011; Wang et al. 2012). Finally, miRNAs can also regulate immune responses by targeting in non-innate immune cells. To this end, miR-146a expression was shown to be induced by TLR2 stimulation in human keratinocytes. miR-146a can then serve as a potent negative feedback regulator to prevent further TLR2-induced inflammatory responses (Meisgen et al. 2014). Collectively, the aforementioned studies not only provide important insights into miRNA-mediated immune regulation but also offer the molecular basis to understand the precise role of miRNAs in the pathogenesis of human immunological diseases.

4 miRNAs in Human Immunological Diseases

As summarized in the previous sections, studies employing genetically manipulated mice with in vitro experimental approaches and in vivo disease models have helped us to gain valuable knowledge of miRNA function in regulating immune responses. However, it is also important to study miRNA in the context of human immunological diseases. Indeed, numerous studies have shown the correlation between the expression of several miRNAs and immune related disorders, such as autoimmunity, hypersensitive diseases, and hematopoietic malignancies.

4.1 miRNA in Autoimmunity and Hypersensitivity Diseases

Abnormal expression of miRNAs has been associated with many autoimmune diseases (Table 1). One of the best examples is systemic lupus erythematosus (SLE), a multifaceted autoimmune disease with a strong genetic predisposition, characterized by enhanced type I interferon signaling. To this end, it has been reported that peripheral blood mononuclear cells (PBMCs) isolated from SLE patients express reduced miR-146a. The amount of miR-146a was shown to negatively correlate with the clinical disease activity and type I interferon levels in patients. It was suggested that lack of miR-146a-mediated regulation of STAT1 and IRF5 led to the excessive production of type I interferon (Tang et al. 2009). Moreover, sequencing analysis of single-nucleotide polymorphisms (SNPs) in SLE patients identified a genetic variant in the miR-146a promoter region that is functionally significant in downregulating the expression of miR-146a by altering its binding affinity for Ets-1 (Luo et al. 2011). In addition to miR-146a, diminished expressions of both miR-125a and miR-155 were also reported in patients with SLE (Zhao et al. 2010; Lashine et al. 2015). While elevated level of RANTES in the absence of optimal miR-125a-mediated regulation was considered to promote the disease, increased expression of protein phosphatase 2A (PP2A) in juvenile SLE patients with reduced miR-155 was thought to be responsible for enhanced pathogenesis. On the other hand, elevated levels of miR-21 and miR-148a were detected in circulating CD4+ T cells from SLE patients. Both of these two miRNAs are able to upregulate autoimmune-associated methylation-sensitive genes such as CD70 and LFA-1 through promoting DNA hypomethylation and by repressing the expression of RASGRP1 and DNA methyltransferase 1 (DNMT1), respectively (Pan et al. 2010).

Table 1 miRNAs involved in human immunological diseases
Full size table

Like SLE, multiple sclerosis (MS) has also been linked to the aberrant expression of many miRNAs. For example, the expression level of miR-326 has been shown to be highly correlated with disease severity in MS patients (Du et al. 2009). Mechanistically, miR-326 promotes Th17 cell induction through targeting Ets-1, a known negative regulator of Th17 differentiation. Moreover, increased miR-27, miR-128, as well as miR-340, were detected in CD4+ T cells of patients with MS. It was shown that through repressing BMI1, a molecule that stabilizes GATA3, miR-27b inhibits Th2 differentiation and promotes proinflammatory Th1 autoimmune responses (Guerau-de-Arellano et al. 2011). Following up this study, miR-27 was further shown to dampen TGFβ signaling, leading to impaired Treg cells and enhanced susceptibility to developing multiple sclerosis (Severin et al. 2016). Besides these two autoimmune diseases, abnormal miRNA expression has also been associated with other autoimmune inflammation including rheumatoid arthritis (RA), psoriasis and type I diabetes (Li et al. 2010; Stanczyk et al. 2008; Lu et al. 2014; Sonkoly et al. 2007; Sebastiani et al. 2011).

In addition to autoimmunity, dysregulated miRNAs also contribute to the development or pathogenesis of hypersensitivity diseases like asthma. miRNA profiling analysis of human airway-infiltrating T cells revealed that miR-19a, a member of the miR-17-92 cluster, was greatly upregulated in CD4+ T cells isolated from asthmatic airways compared with cells from healthy subjects. Mechanistically, miR-19a was shown to repress multiple genes including PTEN, SOCS1, and TNFAIP3, leading to specific augmentation of Th2 responses and associated immune pathology (Simpson et al. 2014). Together, more and more studies have revealed a causative role of miRNA in the development of many human immunological diseases.

4.2 miRNA in Blood Cancer

As one of the major functions of miRNAs is to regulate cell differentiation and proliferation, it is not surprising that when dysregulated miRNAs can drive the development of malignancies of the immune system (Table 2). miRNAs can either act as tumor suppressors or function as oncomirs to promote or prevent tumorigenesis. As tumor suppressors, miR-15a and miR-16 were shown to inhibit the development of B cell chronic lymphocytic leukemias (B-CLL) through targeting Bcl-2, and that in more than 50% of B-CLL patients a region encoding miR-15a and miR-16 was found to be deleted (Cimmino et al. 2005; Calin et al. 2002). miR-28 was also identified as a tumor suppressor and is significantly downregulated in Burkitt lymphoma. Oncogene Myc was shown to negatively regulate miR-28 expression leading to the uncontrolled proliferation of certain B cell subsets (Schneider et al. 2014). Another example is miR-29b whose expression is deregulated in primary acute myelogeneous leukemia (AML). Restoration of miR-29b in AML cell lines was able to induce apoptosis and dramatically reduce tumorigenicity, pointing to a clear tumor suppressor role (Garzon et al. 2009).

Table 2 miRNAs involved in human blood cancer
Full size table

In contrast to the aforementioned roles of miRNA in preventing tumorigenesis, many miRNAs also exhibit oncogenic activity in hematologic malignancies. For example, the miR-17-92 cluster, which is located in a region of DNA that is frequently amplified in human B cell lymphomas, has been shown to promote malignancy of immune cells (He et al. 2005; Tagawa and Seto 2005; Inomata et al. 2009; Lu et al. 2010). In addition to miR-17-92 cluster, miR-155, another well-characterized oncomir, has also been shown to be expressed at higher levels in many different types of B cell lymphomas including Hodgkin’s lymphoma, DLBCL, Burkitt’s lymphoma as well as NK cell lymphoma (Eis et al. 2005; Kluiver et al. 2005; van den Berg et al. 2003; Metzler et al. 2004; Yamanaka et al. 2009). On the other hand, miR-223 was shown to be upregulated in human T cell acute lymphoblastic leukemia (T-ALL) in a TAL1-dependent manner (Mansour et al. 2013). It is thought that miR-223 promotes T-ALL through repressing a tumor suppressor, FBXW7. Finally, elevated levels of the oncomir, miR-125b, has been reported in a variety of human neoplastic blood disorders and could potentially induce myeloid- and B cell leukemia by inhibiting IRF4 (So et al. 2014). The miRNA signatures identified from these clinical studies not only provide great value as prognostic parameters for cancer progression but might also serve as potential novel therapeutic targets to treat human hematopoietic malignancies.

5 Other Non-coding RNAs

Like miRNAs, many other non-coding RNA (ncRNA) species including long non-coding RNAs (lncRNAs) have also been identified as important gene regulators in the immune system (Fig. 3). Since the first lncRNA H19 was reported in 1990 (Brannan et al. 1990), extensive investigation in lncRNA-mediated gene regulation has demonstrated that lncRNAs can regulate gene expression in various biological processes, including immune responses (Ponting et al. 2009). LncRNAs are transcribed by RNA polymerase II, 5’-capped, polyadenylated, and undergo splicing similar to that for mRNAs (Guttman et al. 2009). They can function both in cis to regulate the gene expression in close genomic proximity at the site of transcription, or in trans to target distant transcriptional activators or repressors (Ponting et al. 2009). Moreover, since lncRNAs usually contain multiple modular domains that can either interact with proteins or form complementary pairs with nucleotides, these molecules could connect DNA, RNA, and proteins and be involved in nearly all stages of gene regulation (Guttman and Rinn 2012).

Fig. 3
figure 3

Non-coding RNAs in the immune system. The schematic describes miRNAs (blue font) and lncRNAs (red font) that have key roles in controlling immune responses. CLP, common lymphoid progenitor; CMP, common myeloid progenitor; GMP, granulocyte–monocyte progenitor; MPP, multipotent progenitor

Full size image

A Recent report suggests that up to two third of transcribed genes across all cell types in humans are classified as lncRNAs (Iyer et al. 2015). In T cells, a lncRNA, NRON (non-coding RNA, repressor of NFAT) was shown to form a large cytoplasmic RNA-protein scaffold complex that can repress the transcriptional activation of NFAT-responsive genes via regulating NFAT nuclear trafficking (Willingham et al. 2005; Sharma et al. 2011). Moreover, another lncRNA, lncRNA-CD244, whose expression has been shown to be driven by CD244 signaling upon Tuberculosis (TB) infection, is able to inhibit cytokine production by CD8+ T cells by mediating histone H3K27 trimethylation at promoter regions of IFN-γ and TNF-α (Wang et al. 2015). On the other hand, in Th17 cells, rather than inhibiting their effector function, a lncRNA, lncRNA Rmrp was shown to interact with a complex of RORγt and an RNA helicase, DDX5, to activate RORγt-dependent Th17-relative gene transcription (Huang et al. 2015).

As for the role of lncRNAs in the innate immune cells, a TLR signaling induced intergenic lncRNA, lncRNA-Cox2, was reported to repress the expression of many critical inflammatory genes in macrophages through forming a complex with heterogeneous nuclear ribonucleoprotein (hnRNP) A/B and A2/B1 (Carpenter et al. 2013). In addition to lncRNA-Cox2, lncRNA-EPS was also recently identified as a key regulator in controlling macrophage inflammatory responses. Mice with lncRNA-EPS deficiency exhibited enhanced inflammation and lethality upon LPS challenges. Mechanistically, lncRNA-EPS was shown to limit the expression of immune response genes (IRGs) by controlling nucleosome positioning through interacting with hnRNP L (Atianand et al. 2016).

Finally, the link between lncRNA and human autoimmune diseases has also been demonstrated. Genome-wide association studies (GWAS) of celiac disease patients identified five SNPs in a region encoding a lncRNA, lnc13. Biopsies from celiac disease patients appeared to have substantially lower amounts of lnc13 compared with healthy donors. Further studies have shown that lnc13 is primarily expressed in the nucleus of human macrophages from the lamina propria. Like the aforementioned lncRNAs, lnc13 was also shown to repress many inflammatory genes through interaction with a hnRNP, hnRNP D in particular, as well as Hdac1 and chromatin. It was thus suggested that decreased levels of lnc13 in intestinal tissue from patients with celiac disease likely contributes to the observed inflammation in this autoimmune disorder (Castellanos-Rubio et al. 2016). Despite the great efforts made to study lncRNA biology, unlike miRNAs, it remains a major challenge to functionally evaluate a lncRNA as the sequence of the transcript lends no insight into how it may actually work within a given cell type. Nevertheless, it is evident that lncRNAs exhibit important regulatory functions in controlling immune responses as well as other biological processes.

6 Concluding Remarks

Over a decade of intense scrutiny into the role of miRNAs in the immune system, there is little doubt that miRNAs function as crucial gene modulators that would impact almost all facets of immune responses in both physiological and pathological settings. While miRNAs do not completely turn off (or in some cases, turn on) the expression of their targets, they can act by repressing genes involved in positive-feedback regulatory circuits or by regulating a set of genes that are in a shared pathway or protein complex. As such, even relatively small changes in gene expression introduced by miRNAs could cause major biological consequences. From impairment of immune functions to the pathogenesis of a variety of immunological diseases, the fact that dysregulation of individual miRNAs in the immune system has repeatedly been demonstrated to have profound physiological effects further supports this notion. Moreover, beyond gaining further molecular insights into miRNA-mediated gene regulation in immunological research, recent advances in modulating miRNA function by miRNA mimics or antisense oligonucleotides have shown promise in miRNA-targeted therapeutics. With increasing knowledge of miRNA biology and the development of novel approaches for efficient delivery of miRNA modifying agents to specific immune cell subsets, we are confident that manipulating miRNA pathways will soon become a viable option to treat a wide array of human immunological diseases.