Macro view of microRNA’s function in Osteoarthritis

Abstract

Osteoarthritis (OA) is the most common musculoskeletal disorder. It is a complex and multifaceted disease, characterized by the degradation of articular cartilage and joint inflammation. Although few pathogenesis pathways have been characterized, current knowledge is incomplete and has not led to effective approaches for prevention or treatment. These limitations can be overcome by advances in the understanding of molecular mechanisms that are involved in the maintenance and destruction of articular cartilage. Understanding extracellular regulators and intracellular signaling mechanisms in joint cells that control cartilage homeostasis has the potential to lead to the identification of new therapeutic targets for OA. Recently, non-coding RNAs, miRNAs, was noted to act as novel regulatory molecules that regulated the expression of several target genes. The major role of miRNAs is to control development and tissue homeostasis through the ‘fine-tuning’ of the gene expression. Several miRNAs exhibit a tissue- or developmental stage-specific expression pattern and have been associated with diseases such as cancer and cardiovascular disorders. This review is based on our observations that miRNAs play an important role in cartilage homeostasis, and is to summarize the information on miRNA involved in OA pathogenesis, and its clinical approach.

Chondrocytes and cartilage; role in joint homeostasis and OA

Cartilage with chondrocytes is replaced with bone by the process of endochondral ossification; however, chondrocytes in articular cartilage survive and continue to maintain joint tissue. Joint diseases, such as osteoarthritis (OA), remarkably reduce the quality of life for patients. OA is a chronic and highly prevalent joint disease. Approximately 40 million Americans are currently affected and this number is predicted to increase to 60 million within the next 20 years as a result of population aging and an increase in life expectancy [1, 2]. OA has been characterized by the degradation of articular cartilage, and it is associated with age-related loss of the homeostatic balance between cartilage degradation and repair mechanisms primarily involved in articular cartilage [3–5]. It means that the gene expression network in chondrocytes changes from an “anabolic” to a “catabolic” phase. Catabolic factors, cartilage-degrading enzymes such as ADAMTS-5 and MMP-13 is a critical enzyme for OA pathogenesis [3–5]. However, available treatments for OA are limited to pain management. In the late phase of the disease process, joint replacement surgery is often indicated. Pharmacologic interventions that alter the progressive loss of articular cartilage are not available. Various risk factors for OA have been identified and these include aging, joint injury, excessive chronic mechanical stress, genetic factors, and metabolic disorders [3–5]. Although some pathogenesis pathways have been characterized, current knowledge is incomplete and has not led to effective approaches for prevention or treatment. These limitations can be overcome by advances towards understanding of the molecular mechanisms that are involved in the maintenance and destruction of articular cartilage.

miRNAs, general introduction

MicroRNAs (miRNAs) are a class of non-coding small RNAs that plays a role in biological processes as novel regulators of gene expression. The regulation mechanism is due to promoting mRNA degradation and/or repressing translation through sequence-specific interactions with the 3′ untranslated regions (UTRs) of specific mRNA targets [6–8]. Current studies have identified miRNA-binding sites within the 5′-UTR and within the coding region in targets gene. miRNAs are mainly transcribed by RNA polymerase II as the long primary transcripts (pri-miRNAs) and processed by the nuclear ribonucrease Drosha in a DGCR8 complex in the nucleus, releasing a ~60 bp hairpin precursor miRNA (pre-miRNAs). Pre-miRNAs are processed by the ribonucrease Dicer, to ~22 nt nucleotide (functionally mature miRNAs), that are complexed together with Agonature (Ago), the core unit of the RNA-induced silencing complex (RISC) [6–8]. The miRNA-RISC complex binds target mRNAs and mediates the translational repression or degradation of mRNAs [9–12] (Figure 1).

Figure 1. miRNA function.

Hundreds of miRNAs have been found in various organisms, and many miRNAs are evolutionarily conserved. Moreover, one third of all mammalian mRNAs seem to be under miRNA regulation, suggesting it has an essential role in regulating gene expression [13]. The importance of miRNAs for development and differentiation has been shown by loss-of-function analysis of Dicer, Argonature and DGCR8, enzymes required for miRNA-mediated processing. Loss of Dicer and Argonature in mice results in embryonic lethal or severe developmental defects as a consequence of cell cycle and differentiation [14–16]. Furthermore, the importance of miRNAs in the musculoskeletal system was highlighted by limb- and cartilage-specific deletion of Dicer [17, 18]. These mice had a much smaller limb or body as a consequence of cell death and chondrocyte proliferation, which accelerated their hypertrophic differentiation. Osteoclasts-specific deletion of Dicer in mice increased the levels of bone mass by regulating the activity of bone resorption [19]. These function analysis of Dicer show that gene regulation by miRNA plays an important role in essential cellular functions, and in cartilage and bone development. Indeed, highly specific patterns of miRNA expression correlate with development and several diseases such as cardiovascular disorders, and gain- and loss-of-function of tissue specific-miRNA studies in mice have revealed pathogenic and protective functions of miRNAs in the cardiovascular system [20].

miR-140 and cartilage

Several miRNAs exhibit a tissue- or developmental stage-specific expression pattern and have been associated with several diseases such as cancer and heart disease. However, currently there is only very limited information on miRNA expression and function in the musculoskeletal system. Limb or cartilage specific Dicer knock out mice show a severe phenotype with reduced limb size but normal patterning [17, 18]. As Dicer is indispensable for producing a functional, mature type of miRNA, this finding suggests that the presence of specific miRNA(s) plays a critical role in bone-cartilage development [18]. Systematic whole mount in situ hybridization analysis for miRNAs using zebrafish revealed that many miRNAs showed a tissue specific expression pattern including cartilage [21]. The study in zebrafish embryos showed that miR-140 play a role in palatogenesis by repressing Pdgf signaling [22]. miR-140 is expressed in the cartilage of mouse embryos during the long and flat bone development, and it directly regulates HDAC4 [23]. To identify miRNAs specifically expressed in chondrocytes, we performed gene expression profiling, using miRNA microarrays comparing primary chondrocytes from human articular cartilage to human bone marrow-derived mesenchymal stem cells (MSCs). In primary human articular chondrocytes, several miRNAs were significantly more abundant as compared with undifferentiated MSCs. The largest difference was observed for miR-140 [24]. Our whole mount in situ hybridization data using a probe for pri-miR-140 is consistent with this study and further indicates that tissue specificity of miR-140 expression is regulated at the transcription level (unpublished data). The miR-140 has a chondrocyte differentiation-related expression pattern. The expression increased in parallel with the expression of chondrogenic markers such as Sox9 and Col2a1 during chondrogenesis of MSCs in pellet cultures, while it was reduced in dedifferentiated chondrocytes with each passage. miR-675 indicates a chondrocyte specific expression pattern, and the reduction in its expression depends on dedifferentiation as well as miR-140 [25]. The normal articular cartilage expresses miR-140, and its expression is significantly reduced in OA cartilage, and the in vitro treatment of chondrocytes with interleukin-1 (IL-1β), a cytokine classically involved in the OA pathogenesis, suppresses miR-140 expression [24]. The reduction in miR-140 expression in OA cartilage and the response to IL-1β may contribute to the abnormal gene expression pattern characteristic of OA. Illipolis et al. also showed 16 miRNAs differentially expressed in OA compared to normal cartilage, and they detected miR-140 as one of seven down-regulated miRNAs in OA [26]. These data suggest that miR-140 is associated with OA pathogenesis. Therefore, we generated miR-140-targeted mice and transgenic mice to identify the functions of miR-140 [27]. miR-140-deficient mice were born normally and were fertile. Skeletal development during embryogenesis appeared grossly normal. However, postnatally, miR-140-deficient mice manifested a mild skeletal phenotype of a short stature and craniofacial deformities as a result of impaired proliferation. The skeletal phenotype was almost same with recent reported separately generated miR-140-deficient mice [28]. These results demonstrate that miR-140 is essential for normal endochondral bone development and suggest that the reduced BMP signaling caused by Dnpep upregulation plays a causal role in the skeletal defects of miR-140-deficient mice [28]. Knockdown of miR-140 in limb bud micromass cultures resulted in arrest of chondrogenic proliferation by regulating SP1, acting downstream of BMP signaling [29].

Risk factors for OA progression have been considered to be aging, joint injury, excessive chronic mechanical stress and inflammation. To determine the potential role of miR-140 in cartilage homeostasis, first, we observed whether the loss of miR-140 affected the age-related onset of OA changes. Furthermore, we utilized different animal models: a surgical model leading to abnormal mechanical loading, and an antigen-induced arthritis model. The structure of knee joints including articular cartilage, appeared to be normal at birth and 1 month of age. However, these mice spontaneously developed an age-related OA-like pathology. Next, we utilized the surgical OA model, in which the articular cartilage was exposed to an excessive mechanical load, due to joint instability caused by surgical resection of the medial meniscotibial ligament (MMTL). Consistent with observations in the aging OA model, the surgical OA model also demonstrated that miR-140 deficient mice exhibited accelerated proteoglycan loss and fibrillation of articular cartilage in knee joints compared with the wild type mice at eight weeks after surgery. We further examined the role of miR-140 in articular cartilage by utilizing the antigen-induced arthritis model (AIA). In this model, transient joint inflammation causes cartilage damage by the induction of cartilage-degrading enzymes in chondrocytes, synovial cells and infiltrating leukocytes. The AIA model is advantageous in that it leads to cartilage damage in a short time in wild type mice, which enables us to monitor the potential protective effects of miR-140 against cartilage degradation through a gain of function approach. To examine whether miR-140 levels in articular chondrocytes affect cartilage sensitivity to experimental challenges, we assessed AIA in knee joints of miR-140 TG mice, miR-140 deficient mice, and wild-type mice. We observed similar levels of synovial hyperplasia among miR-140 TG mice, miR-140 deficient mice, and wild-type mice; however, miR-140 deficient mice showed reduced Safranin O staining. Importantly, miR-140 TG mice were resistant to proteoglycan and type II collagen loss compared with wild-type mice. MiR-140-deficient mice showed age-related, OA-like changes characterized by proteoglycan loss and fibrillation of articular cartilage. Conversely, transgenic mice over-expressing miR-140 in cartilage were resistant to antigen-induced arthritis. These findings are consistent with the idea that miR-140 protects against OA progression. To reveal this mechanism in transgenic mice of miR-140, we showed the involvement of Adamts-5, a direct target of miR-140. Although the critical role of miR-140 in cartilage maintenance may largely be explained by identifying at least in part, Adamts-5, miRNAs are believed to regulate multiple target mRNAs. Some studies reported that HDAC4 and IGFBP5 were down-regulated by miR-140 [23, 30]. These results suggest that miR-140 may suppress pathways other than Adamts-5 expression, thus regulating the overall balance of cartilage matrix synthesis and degradation. In deed, gene expression analysis revealed an up-regulation of cartilage-related catabolic factors such as matrix degradation enzymes and a down-regulation of cartilage matrix genes in chondrocytes of miR-140 deficient mice.

The miR-140 is located on an intron of the E3 ubiquitin protein ligase Wwp2 gene. The intronic miRNAs and their host genes regulate independently or are processed from spliced intronic miRNA from the host gene. Recently, two groups reported that the host gene of miR-140, Wwp2 is expressed in the cartilage which is regulated by Sox9, a master gene for chondrogenesis [31, 32]. Furthermore, Wwp2 interacts physically with Sox9 and is associated with the Sox9 transcriptional activity via its nuclear translocation, and then induce expression of Col2a1 [32]. Its deficient mice develop malformations of the craniofacial region such as a shortened snout and craniofacial deformities as does the miR-140 deficient mice [31]. miR-140 might be processed from its specific transcript, however, we cannot exclude the possibility that miR-140 could also be processed from spliced intronic RNA from Wwp2. Although The expression profile of many intronic miRNAs are correlated with their host gene [33], intronic miRNAs and their relationship with their host gene remain unknown. In conducting further research on miRNA, we should examine more carefully the relationship of miRNA and its host gene.

Taken together, miR-140 has a critical and nonredundant role in cartilage development and homeostasis and protects against OA development through Adamts-5 expression, regulated directly by miR-140 [27, 34]. Therefore, miR-140 is a novel regulator of cartilage homeostasis and changes in its expression and function play an important role in OA development (Figure 2). These findings suggest not only new mechanisms of OA pathogenesis, but the potential for preventing and treating OA based on miRNA biology.

Figure 2. Dual roles of miR-140 in endochondral bone development and articular cartilage homeostasis.

miR-140 is required for both endochondral bone development and articular cartilage homeostasis by repressing Dnpep and Adamts-5 expression as direct target.

OA risk factors related to miRNA (Aging, mechanical stress, inflammation)

Various risk factors or stress for OA have been identified and these include aging, mechanical stress, and inflammation [3–5]. The function of miRNAs, especially, is apparent under certain stress factors. Aging is one of the most important risk factors. Even though the relation ship between miRNAs and the aging is not fully understood, several studies have provided evidence showing that miRNAs was associated with aging [35]. These studies showed that many miRNAs are involved in pathways known to modulate the aging process. miRNA expression is showed more predominantly in up-regulated miRNAs than down-reguated miRNAs found in livers of mice during aging. Parallel proteomic profiling and bioinformatic mapping show that the up-regulated miRNA expression corresponds to the down-regulation of genes functionally involved in the control of intermediate metabolism, apoptosis, DNA repair, oxidative defense, particularly mitochondrial oxidative phosphorylation, etc [36]. Accumulation of stress might be a major factor in inducing aging and mechanical stress in OA, and stress-dependent changes of miRNA expression during aging might be play an important in OA development.

The appropriate mechanical stress has been known to be required for cartilage homeostasis, however, excessive mechanical loads due to joint trauma leads to acute posttraumatic arthritis and in the majority of individuals, as a long-term complication to OA [3]. Understanding the mechanotransduction pathway might be one way of approaching cartilage homeostasis and its break mechanisms. miR-222 expression patterns in articular cartilage are higher in the weight-bearing area as compared with the non-weight-bearing area. Thus, miR-222 might be a potential regulator of the articular cartilage mechanotransduction pathway [37].

It is well known that inflammatory cytokines, including TNF-α and IL-1β, are up-regulated in arthritis joint tissue and the acute posttraumatic inflammatory response, and that cell signaling triggered by these effectors disrupt articular cartilage homeostasis [4, 5]. Increased IL-1 expression has been documented after mechanical joint injury or during OA development, and correlates with the severity of cartilage damage [3, 5]. miR-146a in THP-1 cells was induced in response to lipopolysaccharide (LPS) and proinflammatory mediators such as TNF and IL-1, and by NF-kB, and it regulates cytokine signaling in a negative feedback loop involving the down-regulation of the IL-1 receptor associated kinase 1(IRAK1) and TNF receptor associated factor 6 (TRAF6) [38]. miR-146a and miR-155 are expressed in rheumatoid arthritis (RA) synovial tissue, and its expression is induced by stimulation with TNF-α and IL-1β in RA synovial fibroblasts (SFB) [39, 40] and is involved in regulating the inflammatory response. The expression of miR-203 was higher in RA-SFB than in OA-SFB or fibroblasts from healthy donors. Enforced expression of miR-203 led to significantly increased levels of MMP-1 and IL-6 via NF-κB pathway [41]. The key transcription factors, including the NF-κB, play critical roles in inflammatory genes expression, however, little is known about how cartilage-specific genes including miRNAs are down-regulated by inflammatory cytokine stimulation.

In chondocytes, IL-1β-stimulated OA chondrocytes, 42 miRNAs were down-regulated, miR-146 and miR-491 were up-regulated, and the expression of 308 miRNAs remained unchanged. Down-regulated miR-27b directly targets and regulates Mmp-13 expression [42]. miR-146a was highly expressed in early OA cartilage [43]. However, it was down-regulated in sever OA cartilage, and its overexpression in chondrocytes reduced IL-1β induced TNF-alpha production [44]. In contrast, miR-146a is expressed at reduced levels in DRGs and at the dorsal horn of the spinal cords isolated from rats experiencing OA-induced pain [45]. These indicate that miR-146a controls knee joint homeostasis and OA-associated algesia by balancing inflammatory responses in cartilage and synovium with pain-related factors in glial cells. Furthermore, although the peripheral blood mononuclear cells (PBMC) from patients with RA expressed miR-146a [46], overexpression of miR-146a into PBMC could inhibit osteoclastogenesis, and administration of miR-146a could prevent joint destruction in arthritic mice [47]. As upregulated miRNA in OA cartilage, miR-22 in cartilage is correlated with the Body Mass Index (BMI). It directly regulates PPARA and BMP-7 expression, and its up-regulation induced inflammatory and catabolic changes [26]. miR-9 was found to inhibit secretion of MMP13 in isolated human chondrocytes [44]. These results demonstrated that miRNA such as miR-140 positively regulates cartilage homeostasis [27], and miR-146 regulates cytokine signaling [38]. These finding suggest that administration of both miR-140 and miR-146 might have the potential to be a novel targets in OA. Thus, it will be required to identify miRNAs sets associated with OA development.

Clinical application

Several molecular targets in OA development are discovered using gene network approaches through transgenic mice experiments. These molecules support the therapeutic potential of pathways relevant in cartilage homeostasis including proteases, transcriptional factors, proteoglycan and miRNAs [27, 48–54]. However, there are several issues to be resolved before these can be used for clinical applications to human OA treatment. The dysregulation in cartilage catabolism finally leads to cartilage destruction by proteases such as cartilage degrading enzymes (e.g., MMP-13 and ADAMTS-5). The trials applying the protease inhibitors for clinical use as a disease-modifying treatment have been unsuccessful, because of the lack of efficiency or adverse effects [55]. Recently, some reports showed that the activation of ADAMTS-5 is regulated the proteoglycan, syndecan-4, through direct interaction with the protease [48]. Transcriptional factors that turen on the gene expressions are induced by several OA risk factors such as inflammation. HIF-2α is an extensive regulator of the endochondral ossification process during OA development, and it directly induces the expression in chondrocytes of genes encoding catabolic factors and osteogenic factors like RUNX2, IHH [51, 52]. RUNX2 regulates Adamts-5 [56] and hedgehog signaling regulates the expression of Adamts-5 via RUNX2 [57]. However, Saito et al. discussed that RUNX2 does not seem to be essential for the OA induction by HIF-2α, and transcriptional factors in cartilage may be difficult to target with a systemic drug because cartilage lacks vasculature [58]. Although the mechanisms of cartilage degrading enzymes such as ADAMTS-5 expression or activation remain poorly understood, miR-140 directly regulates Adamts-5 expression in cartilage at the post-transcriptional level [27]. Another approach would favor the activity of anabolic processes by using growth factors (e.g., FGF-18 and BMP-7) or regulatory molecules [59]. Further understanding extracellular regulators and intracellular signaling mechanisms that regulated chondrocyte activation has the potential to lead to the identification of new therapeutic targets able to control joint destruction and repair. Consequently, siRNA and miRNA are expected to be part of the next generation of therapeutic molecules. Indeed, Phase I–III clinical trials with siRNA are being conducted [60]. The downregulation of each taeget genes by siRNA is more striking than miRNA. However, miRNAs might be more available and safer as therapeutic molecules than siRNA because the role of miRNAs is to ‘fine-tune’ gene expressions to control homeostasis. For arthritis, Nagata et al. demonstrated the possibility of therapeutic intervention for joint diseases. They showed that an intra-articular injection of synthetic double-stranded miR-15a successfully induces cell apoptosis by inhibiting the target gene BCL-2 in synovium in arthritic mice. Injected labeled-miR-15a into articular spaced was detected in cells of synovium [61]. However, it could not be detected in chondrocytes. It demonstrated that miR-140 regulates cartilage homeostasis [27], and miR-146 regulates cytokine signaling in a negative feedback loop involving down-regulation of IL-1 receptor associated kinase 1(IRAK1) and TNF receptor associated factor 6 (TRAF6) [38]. These findings suggest that administration of miR-140 and miR-146 might have the potential to used as a novel treatment for the early stage of OA, a combination of several miRNAs might be more effective than targeting a single miRNA (Figure 4). However, the key to siRNA and miRNA application in vitro and in vivo is overcoming the difficulty of delivering siRNA or miRNA into the cells such as chondrocytes in cartilage. Owing to their high molecule size and strong anionic charge, siRNAs or double-stranded miRNA cannot pass through the highly regulated and restricted plasma membrane. siRNA and miRNA delivery systems are still required to enhance the siRNA uptake. Eguchi et al reported an efficient siRNA delivery approach that uses a peptide transduction domain-double-stranded RNA-binding domain (PTD-DRBD) fusion protein [62]. PTD-DRBD-delivered siRNA induced rapid RNAi in a large percentage of various primary and transformed cells, including T cells, human umbilical vein endothelial cells and human embryonic stem cells. This system observed no cytotoxicity, minimal off-target transcriptional changes and no induction of innate immune responses.

Figure 4. miRNA function as therapeutic molecules.

miR-146 was induced in response to proinflammatory mediators, it regulates cytokine signaling in a negative feedback loop involving the down-regulation of the IL-1 receptor associated kinase 1(IRAK1) and TNF receptor associated factor 6 (TRAF6). While, miR-140 was reduced in response to proinflammatory mediators and it impair the balance of cartilage homeostasis. Therefore, administration of both miR-140 and miR-146 into joint might have the potential to be a novel targets in OA treatment.

Another potential for a clinical application of miRNAs, we can consider miRNAs as a diagnostic marker. The previously described studies focused on the intracellular role of miRNAs in regulating gene expression. Recently, miRNAs were found in the extracellular space including blood and other body fluids [63]. The profiling of circulating tumor miRNAs could potentially be used as surrogate diagnostic markers for biopsy profiling, extending its utility to screening asymptomatic populations. Indeed, many studies already indicate that extracellular miRNA in body fluids correlate with cancer [64, 65]. In arthritis, the PBMC from patients with rheumatoid arthritis exhibited about a 2-fold increase in miR-146a, miR-155, miR-132, and miR-16 expression compared with healthy donors [46]. Murata et al. detected miRNAs in the synovial fluid from patients with RA and OA, and those in plasma. miR-16, miR-146a, miR-155 and miR-223 in synovial fluid of RA were significantly higher than those of OA [66]. Especially, miR-16 and miR-146a in PBMC and plasma correlated with the disease activities of RA. miRNA in body fluids such as synovial fluids could be promising diagnostic biomarkers for joint disease. Furthermore, it has been expected to identify disease-specific profile of extracellular miRNAs to characterize diseases. Success of OA treatments might be more efficient in identifying the extracellular signal that directly activates because this might be easier to target than intracellular signals. The extracellular RNAs are packaged in secretory microparticles such as exosomes, representing genetic material that is transferable from tissue to tissue and from human to human [67]. Therefore, miRNA may be not only a regulatory molecule within the cell, but also, like cytokines, a paracrine regulatory molecules for cell-cell, tissue-tissue communication. Recently, exosomes were used as a carrier for the delivery system, and the therapeutic potential of exosome-mediated siRNA delivery was demonstrated [68]. However, mechanisms involving transport and the functions of exosomes containing extracellular miRNAs are not well understood, and have not yet been investigated in joint cells and joint diseases. The extracellular miRNAs as a novel communication molecule among joint tissue cells might open new insights on joint disease mechanisms such as arthritis. This may lead to the identification of novel therapeutic targets and biomarkers in OA. Furthermore, it would not only pave the way for understanding and treating joint diseases, but it also has the potential to open doors to understanding the general role of extracellular miRNAs as a signaling pathway in a complex network under various physiological and pathophysiological conditions.

Conclusion

The complicated interactions between miRNAs and its multiple target genes are likely to play an important role in gene regulation and the control of pathophysiological pathways. Research on the molecular mechanisms and gene network including new regulators such as miRNA has led to the elucidation of OA pathogenesis pathways. It is not only the positive regulation cascade of gene expression by transcriptional factors but the negative regulation network of gene expression by negative regulators such as miRNAs that may also bring forth new insights of pathophysiological pathways. Furthermore, the findings of the gene expression network in joint tissue cells have a promising therapeutic target in OA, and these may lead to the clinical applications including diagnosis.

Figure 3. The expression of miRNA in the normal- and Arthritis- joint.

The changes of miRNA expression contribute to joint homeostasis or pathological changes. The expression of miRNAs in joint tissue is shown. miR-140, miR-675 and miR-27b play a role maintenance in normal cartilage by regulating cartilage homeostasis and differentiation. miR-146, miR-155 and miR-203 contribute to the regulation of the inflammation system in synovium.