To elucidate the role of microRNA (miRNA) in the pathogenesis of rheumatoid arthritis (RA), we analyzed synoviocytes from RA patients for their miRNA expression.
To elucidate the role of microRNA (miRNA) in the pathogenesis of rheumatoid arthritis (RA), we analyzed synoviocytes from RA patients for their miRNA expression.
Synoviocytes derived from surgical specimens obtained from RA patients were compared with those obtained from osteoarthritis (OA) patients for their expression of a panel of 156 miRNA with quantitative stem-loop reverse transcription–polymerase chain reaction. The miRNA whose expression decreased or increased in RA synoviocytes as compared with OA synoviocytes were identified, and their target genes were predicted by computer analysis. We used an in vitro system of enhancing the expression of specific miRNA by transfection of precursors into synoviocytes, and then we performed proliferation, cell cycle, and apoptosis assays, as well as enzyme-linked immunosorbent assays for cytokine production. The effects of transfection on predicted target protein and messenger RNA (mRNA) were then examined by Western blot analysis and luciferase reporter assay.
We found that miR-124a levels significantly decreased in RA synoviocytes as compared with OA synoviocytes. Transfection of precursor miR-124a into RA synoviocytes significantly suppressed their proliferation and arrested the cell cycle at the G1 phase. We identified a putative consensus site for miR-124a binding in the 3′-untranslated region of cyclin-dependent kinase 2 (CDK-2) and monocyte chemoattractant protein 1 (MCP-1) mRNA. Induction of miR-124a in RA synoviocytes significantly suppressed the production of the CDK-2 and MCP-1 proteins. Luciferase reporter assay demonstrated that miR-124a specifically suppressed the reporter activity driven by the 3′-untranslated regions of CDK-2 and MCP-1 mRNA.
The results of this study suggest that miR-124a is a key miRNA in the posttranscriptional regulatory mechanisms of RA synoviocytes.
MicroRNA (miRNA) are a well-established class of small (∼22 nucleotides) endogenous noncoding RNAs that influence the stability and translation of messenger RNA (mRNA) (1). Using various computational and experimental approaches, hundreds of miRNA have been identified in numerous animal species. The miRNA genes are transcribed by RNA polymerase II as primary miRNA (pri-miRNA) (2, 3). The RNase III enzyme Drosha then processes the nuclear pri-miRNA, yielding a ∼70-nucleotide molecule known as precursor miRNA (pre-miRNA) (4), which is exported from the nucleus. Maturation of the pre-miRNA into miRNA is then mediated by the cytoplasmic enzyme Dicer (5), after which the mature miRNA is loaded into the RNA-induced silencing complex (RISC) (6). Once loaded, the miRNA guides the RISC complex to the 3′-untranslated region (3′-UTR) of target mRNA. The so-called “seed region” (nucleotides 2–8) of miRNA is most important for target recognition and silencing (7, 8). MicroRNA usually bind with imperfect complementarity to their target, which is called the “seed sequence” (7). Association of miRNA with their target mRNA silence expression via at least 3 mechanisms: inhibition of translation, inhibition of the initiation of translation, and destabilization of target mRNA (1).
Recent advances have shown that miRNA expression during development is highly tissue-specific (9–12), which suggests that miRNA may be involved in specifying and maintaining tissue identity. For example, the expression of miR-124a is restricted to the brain and spinal cord in the fish and the mouse, and to the ventral nerve cord in the fly (13). In those tissues, it contributes to the differentiation of neural progenitors into mature neurons through degradation of non-neuronal transcripts (14). In non-neuronal cells, miR-124a is targeted by the repressor element 1–silencing transcription factor (REST). Its conserved sequence and expression across species suggest that miR-124a is an ancient molecule that acts in both muscle and brain development. In addition to tissue development, miRNA also appear to be involved in metabolism, in cell differentiation, growth, and death, and in carcinogenesis (1).
Rheumatoid arthritis (RA) is a chronic disease of unknown cause that presents a characteristic constellation of features, including synoviocyte hyperplasia, which results in pannus formation and joint destruction (15–17). The rheumatoid synovium consists of epithelial cells, which include 2 types of synovial lining cells, fibroblast-like synoviocytes (FLS) and macrophage-like synoviocytes, as well as infiltrating leukocytes, which include T cells, B cells, and dendritic cells, among others (18). The local production of cytokines and chemokines by these cells accounts for many of the pathologic and clinical manifestations of RA (18). In culture, RA FLS proliferate and secrete a variety of cytokines/chemokines/angiogenic factors, including fibroblast growth factor, granulocyte–macrophage colony-stimulating factor (GM-CSF), interleukin-6 (IL-6), IL-8, monocyte chemoattractant protein 1 (MCP-1), and macrophage inflammatory protein 1α (MIP-1α), and they present adhesion molecules, such as selectins, vascular cell adhesion molecules, and intercellular adhesion molecules, on their surfaces (19).
Our aim in the present study was to investigate the extent to which specific miRNA are involved in the pathogenesis of RA by comparing miRNA expression profiles in FLS from RA patients with those in FLS from osteoarthritis (OA) patients. Our findings suggest that miR-124a plays a key role in regulating the proliferation and chemokine production of RA FLS.
Joint tissue specimens from RA and OA patients were obtained at the time of joint surgery. RA and OA were diagnosed according to the criteria of the American College of Rheumatology (20, 21). The clinical characteristics of the patients are shown in Table 1. FLS were also isolated from synovial joint tissues collected from these RA and OA patients. Samples were obtained in accordance with the Declaration of Helsinki Ethical Principles for Medical Research Involving Human Subjects, as approved by the World Medical Association. All patients provided informed consent, and the study was approved by the Ethics Committees of Kobe University Hospital and Hyogo Prefectural Rehabilitation Center Hospital.
|Patient/age/sex||Disease duration, years||Steinbrocker RA stage||Presurgical CRP, mg/dl||Source of synovium||Medications|
|RA1/65/F||2||III||0.17||Hip||Tacrolimus 2 mg/day|
|RA2/66/F||5||IV||0.79||Knee||MTX 6 mg/week|
|RA3/55/F||9||IV||0.50||Knee||Pred. 4 mg/day; MTX 8 mg/week|
|RA5/65/F||20||IV||5.20||Knee||Pred. 7.5 mg/day; MTX 6 mg/week|
|RA6/59/F||6||III||0.54||Knee||Pred. 5 mg/day|
|RA7/63/F||17||IV||0.33||Knee||Pred. 5 mg/day; MTX 6 mg/week|
|RA9/56/F||26||IV||0.42||Knee||Pred. 4 mg/day|
|RA10/61/F||17||IV||1.70||Knee||Pred. 2 mg/day; Pen. 200 mg/day|
|RA11/78/F||32||IV||0.66||Knee||Pred. 10 mg/day; SSZ 1 gm/day|
|RA12/78/F||15||IV||0.37||Hip||Pred. 5 mg/day; SSZ 1 gm/day; MTX 8 mg/week|
|RA13/71/M||30||IV||7.52||Knee||MTX 6 mg/week|
|RA14/75/F||21||III||5.6||Knee||SSZ 1 gm/day|
|RA20/68/M||25||III||0.25||Knee||Pred. 5 mg/day|
|RA21/63/M||14||IV||2.05||Elbow||SSZ 1 gm/day|
|RA26/65/F||3.5||III||1.93||Knee||Pred. 2.5 mg/day; MTX 6 mg/week|
|RA28/54/F||21||III||0.20||Knee||Pred. 7.5 mg/day; MTX 4 mg/week|
Tissue specimens were minced and digested, and dissociated cells were cultured as described previously (22). After 7 days in culture, nonadherent cells were removed, and adherent cells were maintained in RPMI 1640 supplemented with 10% fetal calf serum (FCS) and 4 mML-glutamine. All experiments were conducted using cells from passages 2–5. The E11 line of human rheumatoid synovial cells, which was established by electroporation with the SV40 large T antigen, were kindly provided by Dr. Yoshiya Tanaka (University of Occupational and Environmental Health, Fukuoka, Japan) (23). Morphologically and phenotypically, E11 cells are a typical synovial FLS clone. E11 cells were grown in RPMI 1640 supplemented with 10% FCS and 4 mML-glutamine.
Normal human FLS were purchased from Cell Applications (San Diego, CA). The normal FLS were first cultured in synoviocyte growth medium (Cell Applications) for proliferation. Then, the growth medium was changed to RPMI 1640 containing 10% FCS and 4 mML-glutamine, and the cells were cultivated for >1 week before they were used in the experiments.
Small RNA was extracted from cultured FLS using a mirVana miRNA isolation kit (Ambion, Austin, TX). Levels of 156 miRNA were determined in 10-ng samples of small RNA by quantitative stem-loop reverse transcription–polymerase chain reaction (RT-PCR), using a TaqMan microRNA Assays Human Panel, Early Access kit (Applied Biosystems, Foster City, CA). Because the stem-loop RT-PCR method is insensitive to precursors and genomic DNA, it has better specificity and sensitivity than conventional RT-PCR (24). The fold change in miRNA expression between RA and OA FLS was calculated as ΔCt, indicating a change in the Ct value from RA FLS to OA FLS.
For transfection of pre-miRNA into primary FLS or E11 cells, 300 pmoles of Pre-miRNA or Pre-miR Negative Control #1 (both from Ambion) was mixed with 10 μl of RNAiMAX in 1 ml of Opti-MEM (both from Invitrogen, Carlsbad, CA). The mixture was seeded into 60-mm culture plates and incubated for 20 minutes at room temperature. Thereafter, 5 ml of FLS (1 × 105 cells/ml in RPMI 1640) was added to the plates and incubated in a CO2 incubator. To confirm the successful transfection, mature miRNA-specific stem-loop RT-PCR was performed. After incubation for 24–96 hours, the cells were used in the experiments.
After transfection of pre-miR-124a into FLS, the cells were counted, and XTT assays (XTT Cell Proliferation Kit II; Roche, Branchburg, NJ) and carboxyfluorescein succinimidyl ester (CFSE) assays (CellTrace CFSE Cell Proliferation kit; Invitrogen) were performed according to the manufacturer's instructions. For cell cycle analysis, cells were pretreated with propidium iodide (PI)/RNase staining buffer (Becton Dickinson, San Diego, CA), after which the cell cycle was analyzed using ModFit software (Becton Dickinson). To detect apoptotic cells, annexin V/PI staining was performed using a MEBCyto apoptosis kit (MBL, Nagoya, Japan), and TUNEL assays were performed using an ApopTag fluorescein in situ apoptosis detection kit (Millipore, Billerica, MA).
Western blotting was performed according to standard procedures. Mouse anti-human cyclin-dependent kinase 2 (CDK-2; D-12) and mouse anti-human actin IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). After treatment with anti–CDK-2 or anti-actin monoclonal antibody, specific bands were detected by chemiluminescence using ECL Plus detection reagents (Amersham Japan, Tokyo, Japan). Densitometric analysis was performed with a LightCapture system and Atto Image Analysis Software (Atto, Amherst, NY). Actin was used as an internal control. The levels of CDK-2 were quantified and normalized against those of actin.
A BD Cytometric Bead Array system (Becton Dickinson) and flow cytometric analysis were used to screen for changes in cytokine/chemokine secretion from FLS after transfection of pre-miR-124a. The cytokines/chemokines analyzed included angiogenin, granulocyte colony-stimulating factor, GM-CSF, IL-6, IL-8, MCP-1, MIP-1α, MIP-1β, RANTES, tumor necrosis factor α (TNFα), and vascular endothelial growth factor (VEGF). Conventional ELISAs were used to confirm the results obtained with the Cytometric Bead Array system. ELISAs for human MCP-1, angiogenin, IL-8, and VEGF were performed using commercially available kits (R&D Systems, Minneapolis, MN).
The 3′-UTR of human CDK-2 mRNA (GenBank accession no. AF512553; bases 5281–6439) was PCR-amplified using the primers CDK-2 Xho I (CTCGAGCCCTTTCTTCCAGGATGTGA [forward]; bases 5232–5251) and Not I (GCGGCCGCTGTTCACCGTCAGCACTAGC [reverse]; bases 6592–6573), which respectively, contain Not I and Xho I restriction site overhangs. The PCR product, which contains putative miR-124a binding sites, was cloned into TOPO TA Cloning vector (Invitrogen), after which the Xho I–Not I fragment was excised and cloned into the psiCHECK-2 vector (Promega, Madison, WI) immediately downstream of the Renilla luciferase reporter gene using T4 DNA ligase (Invitrogen), thereby generating psiCHECK-CDK-2 3′-UTR.
In the same way, the 3′-UTR of human angiogenin mRNA (GenBank accession no. M11567.1; bases 2253–2425) was PCR-amplified using the primers ANG Xho I (CTCGAGAATTTTCCGTCGTCCGTAAC [forward]; bases 2234–2253) and Not I (GCGGCCGCTGGGGGAAAGATCAATATGC [reverse]; bases 2474–2453) and processed to generate psiCHECK-ANG 3′-UTR. Part of the human MCP-1 3′-UTR stretch (GenBank accession no. AF519531; bases 5596–5981), which also contains putative miR-124a binding sites, was PCR-amplified with the primers MCP-1 Xho I (CTCGAGTCCCCAGACACCCTGTTTTA [forward]; bases 5648–5667) and Not I (GCGGCCGCCAAAACATCCCAGGGGTAGA [reverse]; bases 5846–5827) and processed to generate psiCHECK-MCP-1 3′-UTR.
Site-directed mutagenesis was performed to substitute 2 nucleotides in the seed sequences of CDK-2 and MCP-1. The primers used for the CDK-2 3′-UTR mutation (3′-UTRmut) were TGAACTTCGCTTAAACACTCACCTTCT (forward) and TTTAAGCGAAGTTCAGAGGGCCCACC (reverse), and those used for the MCP-1 3′-UTRmut were ACATTATCGCTTAAGTAATGTTAATTC (forward) and CTTAAGCGATAATGTTTCACATCAAC (reverse). PrimeStar Max DNA Polymerase (Takara, Kyoto, Japan) was used for PCR amplification.
On day 1, pre-miRNA were transfected into E11 cells as described above. On day 2, the supernatants were replaced with RPMI 1640 supplemented with 10% FCS, after which the cells were transfected with psiCHECK-CDK-2 3′-UTR, psiCHECK-MCP-1 3′-UTR, psiCHECK-ANG 3′-UTR, or one of the mutants (psi-CHECK-CDK-2 3′-UTRmut or psiCHECK-MCP-1 3′-UTRmut) using Lipofectamine 2000 and Opti-MEM (both from Invitrogen) according to the manufacturer's instructions. After 4 hours of incubation, the medium was refreshed with RPMI 1640 plus 10% FCS. On day 4, the cells were lysed, and luciferase activity was measured using a Dual-Luciferase Reporter Assay system (Promega) according to the manufacturer's instructions. All experiments were performed in triplicate and normalized to the activity of the firefly luciferase gene, which is contained within the psiCHECK-2 vector as an internal control.
RA FLS cultured for >2 months, which proliferate less than fresh RA FLS, and OA FLS were incubated for 24 hours with 100 ng/ml of either TNFα, IL-1β, IL-6, IL-12, IL-17, IL-18, or interferon-γ (IFNγ; R&D Systems) or lipopolysaccharide (LPS; Sigma, St. Louis, MO). After incubation, small RNAs were prepared, and the levels of miR-124a and RNU6B (control miRNA) expression were determined using TaqMan MicroRNA assays (Applied Biosystems). The results are reported as ΔCt, which was deduced by subtracting the Ct value for RNU6B from that for miR-124a.
Statistical comparisons between groups were made using Student's unpaired t-test. We also applied Welch's unpaired t-test to confirm the significance of differences between 2 groups. The Mann-Whitney U test was used to identify significant differences in miRNA expression between RA and OA FLS. P values less than 0.05 were considered significant. We applied the Bonferroni correction for experiments that were performed in parallel.
To compare the expression of miRNA in RA and OA FLS, we used quantitative mature miRNA-specific RT-PCR to analyze FLS samples from 14 RA patients and 7 OA patients. The detected miRNA were then sorted based on their expression levels (cycles) in the 2 groups (Figure 1). Of the 156 miRNA assayed, miR-124a was the only one that was more strongly expressed in OA than in RA FLS. Five other miRNA (miR-146a, miR-223, miR-142-3p, miR-142-5p, and miR-133a) were expressed more strongly in RA than in OA FLS (P < 0.001) (Figure 1).
To investigate the function of miR-124a in RA, we transfected pre-miR-124a, the precursor of miR-124a, into RA, OA, and normal FLS. Subsequent overexpression of mature miR-124a in transfected cells was confirmed by quantitative mature miRNA–specific RT-PCR. We confirmed the transformation efficiencies of RA FLS (Figure 2A), as well as the OA and normal FLS. At 48 hours after transfection with pre-miR-124a, the Ct values for miR-124a in all cell types was reduced by >10, indicating at least a 210-fold increase in miR-124a.
XTT assays showed that transfection with pre-miR-124a significantly suppressed the proliferation of RA FLS, E11 cells, and HeLa cells (cervical cancer–derived cells) but did not suppress the proliferation of OA and normal FLS (Figure 2B). Analysis of cell counts showed that transfection with pre-miR-124a suppressed the proliferation of RA FLS (Figure 2C). Interestingly, transfection with pre-miR-124a suppressed proliferation, but did not induce cell death, in RA FLS, which is in contrast to the induction of extensive cell death in E11 cells (Figure 2C). CFSE assays using E11 cells confirmed that transfection with pre-miR-124a suppressed cell proliferation, as compared with controls (Figure 2D).
Cell cycle analysis revealed that transfection with pre-miR-124a induced G1 arrest in RA FLS (Figures 3A and B). In addition, annexin V/PI staining showed no significant increase in apoptosis of RA, OA, or normal FLS after transfection with pre-miR-124a (Figures 3C and D), whereas there was increased apoptosis of pre-miR-124a–transfected E11 cells (Figures 3C and E). These results suggest that the low expression of miR-124a in RA FLS might protect them from cell cycle arrest, thereby promoting cell proliferation. We speculate that the apoptosis of E11 cells provoked by miR-124a is an additional property that is related to the transformation process of immortalizing this cell line from RA FLS.
For comparison, we transfected pre-miR-146a, pre-miR-223, pre-miR-142-3p (representing pre-miR-142), and pre-miR-133a, all of which were elevated in RA FLS, into OA FLS. XTT assays showed that transfection of these pre-miRNA did not significantly promote the proliferation of OA FLS (data not shown). In addition, XTT assays showed that transfection with the miR-133a inhibitor did not suppress the growth of E11 cells under the same conditions in which the inhibitor had suppressed the miR-133a levels (to one-twentieth of that in the control sample) (data not shown).
We used miRanda 3.0, a widely used program provided by the Sanger Institute (online at http://microrna.sanger.ac.uk/), to search for putative miR-124a targets. This analysis predicted that the 3′-UTR of human CDK-2 mRNA and the 3′-UTRs of 1,299 other mRNA were targets of miR-124a (data not shown). We focused on CDK-2 because it is known to be an important CDK, acting at the G1 cell phase. The levels of miR-124a expression were negatively correlated with CDK-2 expression, although there was no statistically significant difference in CDK-2 protein levels between RA and OA FLS (Figures 4A and B). When we tested the effect of miR-124a on CDK-2 expression by transfection of pre-miR-124a into primary RA FLS and E11 cells, we found that overexpression of miR-124a substantially suppressed the expression of CDK-2 in both cell types (Figure 4C).
To confirm the direct effect of miR-124a on CDK-2 mRNA, its wild-type 3′-UTR or a mutant form was cloned into a luciferase reporter plasmid and cotransfected into E11 cells along with pre-miR-124a. As shown in Figure 4D, miR-124a selectively suppressed the luciferase activity driven by the wild-type 3′-UTR, but not the mutant form, of CDK-2 mRNA, indicating that CDK-2 mRNA is a direct target of miR-124a.
We next measured the levels of cytokines/chemokines in culture medium conditioned with RA FLS after overexpression of miR-124a, and we detected changes in the levels of 3 of them: MCP-1, angiogenin, and VEGF (Figure 5A). VEGF levels increased by ∼50% compared with those in medium conditioned with control cells. In contrast, MCP-1 and angiogenin levels were significantly decreased in the presence of miR-124a overexpression.
We also transfected pre-miR-133a, pre-miR-142-3p, pre-miR-146a, and pre-miR-223 into OA FLS and analyzed the culture media for cytokines/chemokines. The sera were first screened with the Cytometric Bead Array system for changes in cytokine/chemokine expression, and possible candidates for changes were further analyzed by ELISA. In contrast to the findings with pre-miR-124a, the overexpression of these other pre-miRNA did not stimulate any cytokine/chemokine secretion from OA FLS (data not shown).
We then used miRanda 3.0 to search for the 3′-UTR sequences of the mRNA encoding MCP-1, angiogenin, and VEGF. We found that only MCP-1 mRNA contained a seed sequence for miR-124a, which suggests that miR-124a binds directly to its 3′-UTR. To test that idea, we inserted the 3′-UTR fragment from MCP-1 mRNA, including the putative target site, or its mutant into a luciferase reporter plasmid and cotransfected it along with pre-miR-124a into E11 cells. Subsequent luciferase assays revealed that miR-124a specifically suppressed the luciferase activity driven by the 3′-UTR of MCP-1 mRNA (Figure 5B). In contrast, miR-124a had no appreciable effect on luciferase activity driven by the 3′-UTR fragment of angiogenin mRNA, which has no putative binding site for miR-124a (Figure 5C).
Finally, we examined the effects of inflammatory cytokines and stimulants on the expression of miR-124a. RA or OA FLS were treated with TNFα, IL-1β, IL-6, IL-12, IL-17, IL-18, IFNγ, or LPS (all at a concentration of 100 ng/ml, which has been reported to have optimal biologic effects), and the levels of miR-124a expression were determined by quantitative mature miRNA–specific RT-PCR. We found that none of these mediators elicited a change in miR-124a expression at the concentrations we used (data not shown).
Notable features of miRNA include their redundancy with respect to their target binding sequences in the 3′-UTR of mRNA and their relatively small total number, which is speculated to range from 500 to 1,000. For example, analysis using miRanda 3.0 predicted the 3′-UTR of 1,300 human mRNA as potential targets of miR-124a. These unique properties prompted us to speculate that an miRNA could regulate a number of molecules involved in the pathogenesis of RA. However, miRNA are regarded as negative regulators of the translation of mRNA. Therefore, there are bidirectional possibilities for the involvement of miRNA in the pathogenesis of RA: not only the down-regulation of specific miRNA binding to the 3′-UTR of mRNA, which generates proinflammatory proteins, but also the up-regulation of specific miRNA binding to the 3′-UTR of mRNA, which generates antiinflammatory proteins, might make a contribution. In this regard, both elevated and suppressed levels of miRNA are important in RA FLS.
In trying to detect miRNA whose expression differed in RA and OA FLS, we found that the level of miR-124a in RA FLS was less than one-sixth of that seen in OA FLS. Although the mechanism by which the expression of miR-124a is regulated is not yet clear, our findings suggest that low levels of miR-124a expression during RA pathogenesis could have significant effects on synovial cell proliferation, leukocyte chemotaxis (MCP-1), and angiogenesis (angiogenin and VEGF).
In addition, our reporter assays showed that cotransfection with pre-miR-124a decreased luciferase levels from the reporters that were fused to the wild-type 3′-UTRs, but not those that were fused to the mutant 3′-UTRs, of CDK-2 and MCP-1 mRNA, whereas Western analyses showed that overexpression of miR-124a in RA FLS suppressed the expression of CDK-2 and the secretion of MCP-1 by synoviocytes. These results strongly suggest that CDK-2 and MCP-1 are the direct targets of miR-124a in RA synoviocytes. Using ELISAs, we also found that overexpression of miR-124a in synoviocytes leads to the down-regulation of angiogenin and the up-regulation of VEGF. It is not clear, however, how miR-124a influences the expression of these angiogenic chemokines, since miRanda 3.0 analyses detected no seed sequences for miR-124a in either angiogenin or VEGF mRNA, nor did any other available miRNA database program, and miR-124a did not suppress luciferase activity driven by the 3′-UTR of angiogenin. It may be that miR-124a acts indirectly on various molecules, but that idea remains to be tested.
The contribution of cell cycle–related proteins, such as cyclins, CDKs, and CDK inhibitors, to carcinogenesis has been intensively investigated. Recent reports have shown that in human cancers, the expression of specific miRNA closely related to the regulation of cell growth and apoptosis differs from that in normal tissues (25, 26). RA is characterized by pronounced synovial hyperplasia and by synovial fibroblasts that appear to be transformed (27, 28). In animal models of RA, this transformed appearance of RA synoviocytes could be mitigated by transferring the CDK inhibitor genes p16INK4a and p21Cip1 into inflamed joints (29, 30). CDK-2, which is inhibited by p21Cip1 and p27Kip1, is another key CDK: CDK-2/cyclin E complexes are required for the G1-to-S phase transition and initiation of DNA synthesis, whereas CDK-2/cyclin A complexes function during the progression of cells through S phase (31). Our in vitro finding that overexpression of miR-124a caused G1 phase arrest in RA FLS is indicative of a direct suppression of CDK-2 mRNA by miR-124a.
The microRNA miR-124a was initially identified as a crucial regulator involved in neurogenesis (13). Our findings suggest that miR-124a takes on other functions at different stages of human development. Recently, Pierson et al (32) reported that the level of miR-124a expression, which is enriched in brain tissue, is low in medulloblastomas. They also showed by luciferase assay that the 3′-UTR of CDK-6 mRNA is a direct target of miR-124 and that CDK-6 expression is suppressed by miR-124 overexpression in medulloblastoma cell lines. We confirmed that the expression of CDK-6 protein was higher in RA FLS than in OA FLS and that CDK-6 expression was suppressed when pre-miR-124a was transfected into E11 and RA FLS–like CDK-2 (data not shown). Since CDK-6 is also known as a G1/S phase regulator, as is CDK-2, we think that miR-124a is an important regulator of the G1/S transition in synovial tissue as well as in tumors.
We also showed that MCP-1 is down-regulated by miR-124a. The pivotal role played by MCP-1 in RA in humans is highlighted by the findings of enhanced production of MCP-1 in serum and/or synovial fluid from patients with RA (33–35). Moreover, data from studies of animal models suggest that MCP-1 is involved in the pathogenesis of RA (36–38), and MCP-1 was recently reported to be a sensitive marker of disease activity in patients with juvenile RA (39). MCP-1 attracts memory T lymphocytes and natural killer cells, which are major contributors to the pathogenesis of RA (40). In addition, MCP-1 mediates angiogenesis via VEGF (41). We identified a putative binding site for miR-124a in the 3′-UTR of MCP-1 mRNA by database analysis and demonstrated miR-124a–specific suppression of MCP-1 secretion from RA FLS, which suggests that down-regulating miR-124a in RA FLS would facilitate MCP-1 secretion, thereby enhancing its chemotactic effects.
There have been studies showing that the expression of specific miRNA is altered by extracellular signals, such as cytokines (TGFβ, IFNγ, TNFα, and IL-1β) and Toll-like receptor ligands (LPS and poly[I-C]) (42–44). Those studies demonstrated that the miRNA induced by extracellular signals regulate the mRNA of proteins that are closely linked to cell type–specific functions. Considering our findings as well as the inflammatory environment of RA synovium, it is therefore possible that some extracellular signals may regulate the expression of miR-124a in synoviocytes. However, in our screening tests, we identified no cytokine/chemokine that affected miRNA-124a expression.
Investigating the constituents of the culture supernatants may be helpful, since RA fibroblasts continuously produce epidermal growth factor and platelet-derived growth factor in an autocrine manner over several months of culture (45). This suggests that the reduction in miR-124a expression may be caused by an epigenetic event that may not be simulated by short-term exposure to cytokines. Consistent with this idea, Lujambio et al (46) recently reported that miR-124a is a proliferation-associated miRNA and that it is silenced by the hypermethylation of the miR-124a gene in a variety of cancer cells (46). Since our data showed that the proliferation of RA FLS was suppressed by the introduction of pre-miR-124a, it would be of interest to investigate the methylation status of the miR-124a gene in RA FLS.
Transcription factors that regulate the activity of miRNA promoters have recently been described (47, 48). Moreover, several miRNA have been shown to regulate the 3′-UTR of mRNA that encode transcription factors (49), and a circuit that sequentially involves miRNA and transcription factors in a mutual negative feedback loop has been described (48, 49). As for the transcription factors, it has been reported that REST inhibits miR-124a expression in non-neuronal cells. However, the role of REST has not been investigated aside from neuronal development, let alone its role in synovial cell biology or RA pathogenesis. Many genes that encode miRNA have configurations similar to those of standard gene loci that generate mRNA and proteins, and interestingly, one such group of gene loci is the miR-124a (50). By taking a different approach, such as searching for transcription factors that bind to the promoter region of miR-124a, it may be possible to identify the transcription factor(s) that regulates miR-124a expression in RA FLS.
Taken together, our findings suggest that the pathogenesis of RA will be better understood when miRNA are added to the big picture that illustrates the molecular kinetics of RA. We anticipate that miRNA will be considered in future strategies aimed at diagnosing and treating RA.
Dr. Kawano had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Nakamachi, Kawano, Kumagai.
Acquisition of data. Nakamachi, Kawano, Sakai, Chin, Saura, Kurosaka.
Analysis and interpretation of data. Nakamachi, Kawano, Takenokuchi, Kumagai.
Manuscript preparation. Nakamachi, Kawano.
Statistical analysis. Nakamachi, Nishimura.
We thank Ms Kyoko Tanaka for technical assistance and Dr. William Goldman for editing the English version of the manuscript.