Melatonin Prevents Osteoarthritis-Induced Cartilage Degradation via Targeting MicroRNA-140

Osteoarthritis (OA) is characterized by the progressive destruction of articular cartilage, which is involved in the imbalance between extracellular matrix (ECM) synthesis and degradation. MicroRNA-140-5p (miR-140) is specifically expressed in cartilage and plays an important role in OA-induced matrix degradation. The aim of this study was to investigate (1) whether intra-articular injection of melatonin could ameliorate surgically induced OA in mice and (2) whether melatonin could regulate matrix-degrading enzymes at the posttranscriptional level by targeting miR-140. In an in vitro OA environment induced by interleukin-1 beta (IL-1β), melatonin treatment improved cell proliferation of human chondrocytes, promoted the expression of cartilage ECM proteins (e.g., type II collagen and aggrecan), and inhibited the levels of IL-1β-induced proteinases, such as matrix metalloproteinase 9 (MMP9), MMP13, ADAMTS4 (a disintegrin and metalloproteinase with thrombospondin motifs 4), and ADAMTS5. Both the microarray and polymerase chain reaction (PCR) experiments revealed that miR-140 was a melatonin-responsive microRNA and melatonin upregulated miR-140 expression, which was suppressed by IL-1β stimulation. In vivo experiments demonstrated that intra-articular injection of melatonin prevented disruptions of cartilage matrix homeostasis and successfully alleviated the progression of surgery-induced OA in mice. Transfection of miR-140 antagomir completely counteracted the antiarthritic effects of melatonin by promoting matrix destruction. Our findings demonstrate that melatonin protects the articular cartilage from OA-induced degradation by targeting miR-140, and intra-articular administration of melatonin may benefit patients suffering from OA.


Introduction
Osteoarthritis (OA) is a chronic degenerative joint disease, which is primarily characterized by progressive destruction of articular cartilage. In the late stages of this pathology, OA-induced joint dysfunction is considered to be a leading cause of disability in elderly people, which creates a huge economic burden on society. Currently, no therapy has been shown to effectively halt the progression of OA, and the only clinical treatment for patients with late-stage OA is pros-thetic implants [1]. Although the pathogenesis of OA is not yet fully understood, a key factor is the imbalance between extracellular matrix (ECM) synthesis and degradation in cartilage [2]. Chondrocytes, the only cell type in articular cartilage, produce the structural components of the cartilage ECM, specifically, type II collagen (Collagen II) and the proteoglycan aggrecan. However, during OA pathogenesis, chondrocytes become metabolically active to produce matrix-degrading enzymes; these enzymes include members of the matrix metalloproteinase (MMP) and ADAMTS (a disintegrin metalloproteinase with thrombospondin motifs) families, which are responsible for the degradation of Collagen II and aggrecan, respectively [3]. Among these enzymes, MMP13 (collagenase-3) can cleave collagen, aggrecan, and fibronectin and has the highest activity toward Collagen II [4]. ADAMTS4 (aggrecanase-1) and ADAMTS5 (aggrecanase-2) have also been shown to play important roles in OA development [5].
Several recent studies have demonstrated that the gene expression of matrix-degrading enzymes is tightly regulated by microRNAs (miRNAs) at the posttranscriptional level [6]. miRNAs are a class of small (19)(20)(21)(22)(23)(24) nucleotides in length), noncoding RNAs that can regulate gene expression by binding to specific sequences in messenger RNAs (mRNAs), resulting in either degradation of the target mRNAs or repression of their translation [7]. The differential expression of miRNAs between normal and OA cartilage has been identified recently using miRNA microarrays [8]. Among them, microRNA-140-5p (miR-140), which is expressed specifically in cartilage, has been considered a key factor in chondrocyte differentiation and OA-induced matrix degradation. Deficiency of miR-140 in mice resulted in age-related OA-like changes in cartilage, such as the loss of proteoglycan and fibrillation of articular cartilage [9]. Additionally, miR-140 is dysregulated by increased levels of proinflammatory cytokines in OA cartilage, such as interleukin-1 beta (IL-1β) and tumor necrosis factor alpha (TNF-α). The expression of miR-140 rapidly decreased in IL-1β-stimulated chondrocytes, suggesting that miR-140 might be implicated in OA cartilage dyshomeostasis [10]. Therefore, miRNA therapeutics, especially regarding strategies targeting miR-140, is a promising treatment option for OA patients.
Melatonin, a hormone secreted mainly from the pineal gland, has been shown to exert beneficial effects on bone diseases, such as osteoporosis, osteopenia, and periodontal disease [11]. Previous studies from our laboratory and other groups have demonstrated that melatonin can enhance the chondrogenic differentiation of mesenchymal stem cells (MSCs), even in a proinflammatory cytokine-induced environment [12,13]. Recently, a study suggested that the underlying mechanisms involved in melatonin-improved chondrogenesis was through the upregulation of miRNAs, such as miR-526b-3p and miR-590-5p, which activated the Smad signaling pathway by targeting SMAD7, a negative regulator [14]. However, it is unknown whether melatonin can protect articular cartilage from OA-induced matrix degradation, and few investigations have revealed the role of melatonin in regulating miR-140 during OA pathogenesis.
In this study, we investigated the protective effects of melatonin on OA-impaired articular cartilage. To establish an in vitro OA environment, IL-1β was supplemented during the cell culture of human articular chondrocytes, and the effects of melatonin on ECM synthesis and degradation were evaluated. In in vivo experiments, destabilization of the medial meniscus (DMM) surgery was performed to establish an OA mouse model, followed by intra-articular injection of melatonin for up to four weeks. To further investigate the underlying mechanisms, globe miRNA expression analysis was performed and the role of miR-140 in melatoninmediated antiosteoarthritic effects was investigated.

Human Cartilage Sampling and Chondrocyte
Preparation. The study protocol for using discarded human cartilage samples was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Soochow University. Cartilage samples were obtained from six OA patients (3 males and 3 females, age 60:4 ± 11:6) who underwent total joint replacement.

In Vitro Culture of Human Chondrocytes and
Treatments with IL-1β and Melatonin 2.2.1. Chondrocyte Treatment. Cultured chondrocytes were maintained as a monolayer in DMEM/F12 with 10% FBS at 37°C. To establish an in vitro arthritic microenvironment, recombinant IL-1β at 5 ng/mL (Peprotech, Rocky Hill, NJ, USA) was supplemented in the medium for the indicated periods of time. Melatonin (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in absolute ethanol (EtOH) at a stock concentration of 250 mM and then diluted in a complete medium at a concentration of 1 μM or 100 μM. Cells in the vehicle group were treated with an equal volume of EtOH (0.4 μL per mL medium).

Transfection of miR-140
Antagomir. miRNA-140 antagomir (antago miR-140) and negative control (antago miR-NC) oligonucleotides were obtained from GenePharma Co., Ltd. (Shanghai, China). The sequence of antago miR-140 is 5′-CUCCCUUCUCUUCUCCCGUCUU-3′ and NC is 5′-CUCCCUUCUCUUCUCCCGUCUU-3′. Chondrocytes were seeded in 6-well plates and transfected with miR-140 antagomir (100 nM) or NC miRNA using Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer's protocol. Forty-eight hours after transfection, the cells were treated with 5 ng/mL of IL-1β in the presence or absence of melatonin. To establish the mouse model of OA, DMM surgery was performed to induce mild instability of the knee according to a previous study [15]. Briefly, mice were anesthetized using an inhalation anesthesia system (chamber filled with 2.0% isoflurane plus 30% oxygen, RWD Life Science, Shenzhen, China). The medial meniscotibial ligaments (MML) in the right knees were transected using microsurgical scissors to cause the instability of the medial meniscus. As controls, sham surgery was performed in the left knees, in which the capsular incision was made but the MML was left intact. All surgeries were performed by the same person (ZY).

Intra-Articular
Injection of Melatonin and the miRNA-140 Antagomir. Melatonin was dissolved in absolute ethanol and diluted in saline (0.9% NaCl) to yield a final concentration of 10 mg/mL. After surgery, the sham-op and DMM-op mice were treated with equal amounts (10 μL) of melatonin or saline via intra-articular injection through the patellar tendon. The mice were injected twice a week for four weeks, then euthanized to collect samples of their medial femoral condyle cartilage. For miR-140 inhibition experiments, DMM-op mice were treated with equal amounts (10 μL) of miRNA control (NC) or miR-140 antagomir (250 nM) via intra-articular injection on days 3 and 7 postsurgery along with melatonin administration.
2.3.3. Histology and Immunohistochemistry. Dissected mouse knees were fixed in 10% formalin and decalcified in 10% EDTA (pH = 7:4, Sigma-Aldrich) for 2 weeks. After decalci-fication, each specimen was embedded in paraffin and sagittally sectioned at a thickness of 6 μm. For histological analysis, the sections were stained with hematoxylin and eosin (H&E) and Safranin O (S.O.)/Fast Green (Sigma-Aldrich). Histological images were taken with a bright-field microscope (Zeiss Axiovert 200, Oberkochen, Germany). The slides stained with Safranin O/Fast Green were scored by three independent investigators (ZY, LT, and HF) blinded to group assignment, using the Osteoarthritis Research Society International (OARSI) scoring system [16].
For immunohistochemistry, the paraffin-embedded sections were dewaxed using xylene and hydrated in decreasing graded ethanol solutions. The slides were incubated with 1% hydrogen peroxide (H 2 O 2 ; Sigma-Aldrich) for 30 min and then treated with 2 mg/mL testicular hyaluronidase (Sigma-Aldrich) for 30 min at 37°C. The slides were blocked in 1.5% goat serum, followed by incubation with specific anti-Collagen II (COL II; ab34712) or anti-Collagen I (COL I; ab34710, Abcam) primary antibodies overnight at 4°C. A secondary antibody of biotinylated goat anti-rabbit (Vector Laboratories, Burlingame, CA, USA) was applied for 30 min, after which avidin-biotin complex amplification (Vectastain ABC kit, Vector Laboratories) was used. Finally, immunohistochemistry was detected using 3,3 ′ -diaminobenzidine (DAB; Vector Laboratories) as a substrate, and counterstaining was performed with hematoxylin. The percentage of COL II-or COL I-positive cells was counted (per 100 μm 2 area in each section) for quantitative evaluation.  2.6. Statistical Analysis. All statistical analyses were performed using the SPSS 13.0 statistical software (SPSS Inc., Chicago, IL, USA). Data points are presented as means ± standard error of mean (S.E.M.) for at least three independent experiments. Statistical significance was determined using the two-tailed Student's t-test for comparisons between two groups and one-way Analysis of Variance (ANOVA) with Tukey's post hoc test for multiple group comparisons. p < 0:05 ( * ) was considered statistically significant and p < 0:01 ( * * ) was considered highly significant.

The Effect of Melatonin on Human Chondrocytes under
Normal Culture Conditions. Human articular chondrocytes were treated with melatonin at 1 μM and 100 μM concentrations. The CCK-8 results showed that treatment with 100 μM melatonin improved cell proliferation by 11.4% on day 5 and 24.5% on day 7 (Figure 1(a)). Furthermore, melatonin treatments significantly upregulated transcript levels of COL2A1 by 24.0% at 1 μM and 34.2% at 100 μM ( Figure 1   proliferation by 15.4% compared with the IL-1β group (Figure 2(a)). Real-time PCR data suggested that, in IL-1βstimulated chondrocytes, transcript levels of COL2A1 were downregulated by 56.6%. Treatment with 100 μM melatonin significantly upregulated the gene expression of COL2A1 by 36.3% compared with the IL-1β group, but the level was still 40.8% lower than that of the control group (Figure 2(b)). Consistently, the high concentration of melatonin increased the transcript levels of ACAN by 51.4% (Figure 2(c)) and SOX9 by 52.1% (Figure 2(d)) compared to the IL-1β-treated chondrocytes. The immunofluorescence staining showed that the synthesis of Collagen II in chondrocytes was suppressed by IL-1β treatment but restored by melatonin supplementation (Figure 2(e)). Western blot data confirmed that protein levels of cartilage ECM, such as Collagen II and aggrecan, were significantly increased by 100 μM of melatonin (Figure 2(f) and Supplementary Figures 1A and 1B).

Inhibition of IL-1β-Induced Matrix Degradation by
Melatonin in Chondrocytes. We further investigated the effect of melatonin on IL-1β-induced matrix-degrading enzymes. MMPs and ADAMTSs are important proteases that can degrade a wide range of matrix components. Real-time PCR data showed that IL-1β stimulation resulted in an 87.2% increase in the gene expression of ADAMTS4, whereas melatonin treatment downregulated its expression in a dosedependent manner (by 24.9% at 1 μM and by 47.8% at 100 μM; Figure 3(a)). Similarly, treatment with 100 μM melatonin significantly decreased the transcript levels of ADAMTS5 by 40.6% (Figure 3(b)), MMP9 by 50.5% (Figure 3(c)), and MMP13 by 50.7% (Figure 3(d)). The inhibitory effects of melatonin on these matrix-degrading enzymes at the protein levels were confirmed by Western blot assays (Figure 3(e) and Supplementary Figures 1C-1F).

Prevention of DMM-Induced OA in Mice by Intra-Articular Injection of Melatonin.
To investigate the antiarthritic effects of melatonin, sham-op and DMM-op mice were treated with melatonin via intra-articular injection. Histological and immunohistochemical analyses confirmed the loss of proteoglycan and Collagen II in the DMM group, whereas melatonin treatment prevented DMM-induced cartilage matrix destruction (Figure 4(a)). In particular, the positive staining of Collagen I was observed in the DMM-treated mice, which is a typical marker for OA development. In contrast, melatonin treatment suppressed the expression of Collagen I. In the melatonin-treated OA mice, the OARSI score was significantly reduced compared to that of the DMM group (Figure 4(b)). Consistently, the percentage of chondrocytes that stained positively for Collagen II was increased in the DMM+melatonin group (Figure 4(c)), while the percentage of Collagen I-positive cells was decreased (Figure 4(d)).
These results indicate that intra-articular injection of melatonin successfully attenuated surgically induced OA progression in mice by modulating cartilage matrix homeostasis.

Melatonin-Mediated Differential miRNA Expression
Profiles in Human Chondrocytes. To identify the differentially expressed miRNAs that were regulated by melatonin, micro-array analysis was performed. As shown in Figure 5(a), after 72 h of stimulation of chondrocytes with 100 μM melatonin, the heat map of miRNA expression patterns revealed that a total of 50 differentially expressed miRNAs (1.5-fold up-or downregulated) were identified (Supplementary Table 2). Among them, 29 miRNAs were upregulated by melatonintreated cells, including miR-193b-5p (3.00-fold), miR-26b-3p (2.86-fold), miR-140-5p (2.37-fold), and miR-95-3p (2.27fold). On the other hand, 21 miRNAs were downregulated by melatonin treatment, including miR-9-5p (5.87-fold), miR-204-5p (4.05-fold), miR-145-5p (3.23-fold), and miR-181a-3p (2.29-fold). Considering the important role of miR-140 in the pathogenesis of OA, we performed realtime PCR assays to validate the changes of miR-140 expression. IL-1β stimulation resulted in a marked reduction in the level of miR-140 in human chondrocytes by 49.2%. Melatonin treatment upregulated the expression of miR-140 by 1.2-fold and 1.5-fold, respectively, compared with the CTRL and IL-1β groups ( Figure 5(b)). To investigate specific miRNA-regulated gene pathways involved in the pathology of OA, pathway enrichment analysis was performed using DAVID. As shown in Figure 5(c), 27 pathways were significantly enriched, including protein processing in the endoplasmic reticulum, PPAR signaling pathway, p53 signaling pathway, Notch signaling pathway, and Hedgehog signaling pathway. We further investigated the effect of intra-articular injection of miR-140 antagomir on OA progression. After the DMM surgery, OA mice were injected with miR-140 antagomir along with melatonin. Histology results showed reduced levels of proteoglycan and Collagen II in the antagomir-treated mice (Figure 7(a)). Importantly, immunohistochemical analysis revealed strong positive staining for Collagen I in the antagomir-treated mice. In the miR-140 antagomir group, the OARSI score was significantly increased (Figure 7(b)) and the percentage of Collagen II-positive chondrocytes decreased (Figure 7(c)), in contrast to the increased percentage of Collagen Ipositive cells (Figure 7(d)). These results suggest that inhibition of miR-140 completely abrogated the antiarthritic effect of melatonin.

Discussion
The imbalance between the anabolism and catabolism of the cartilage matrix is the hallmark of OA, which is induced by the increased number of proinflammatory cytokines. IL-1β is an important mediator of joint inflammation; not only does it blunt the synthesis of cartilage ECM, it also stimulates the expression of matrix-degrading enzymes. It was reported that IL-1β inhibited chondrogenesis of human MSCs through activation of the nuclear factor κ-light-chain enhancer of activated B cell (NF-κB) signaling pathway [17]. In this study, we used IL-1β to mimic the OA inflammatory environment in vitro and showed that IL-1β stimulation significantly suppressed the synthesis of cartilagespecific ECM (e.g., Collagen II and aggrecan). However, treatment with melatonin promoted the expression of cartilage ECM at both the mRNA and protein levels, even in the presence of IL-1β. SOX9 plays an important role in chondrogenesis by regulating matrix protein synthesis during neocartilage formation [18]. Consistent with previous studies, we found that melatonin significantly increased the level of transcriptional factor SOX9 even in IL-1β-treated chondrocytes. In addition, chondrocyte proliferation was improved by melatonin treatments in a dose-dependent manner, in which SOX9 may contribute to an increased number of proliferating chondrocytes [19].
On the other hand, melatonin was demonstrated to have an inhibitory effect on cartilage ECM degradation. In this study, we demonstrated that, in IL-1β-stimulated chondrocytes, melatonin treatments led to a significant reduction in the expression of matrix-degrading enzymes such as ADAMTS4 and ADAMTS5. These two proteinases are responsible for aggrecan degradation during OA development, which are found at increased mRNA expression and protein levels in OA cartilage [20]. Knockout of ADAMTS4 and ADAMTS5 in mice protected the hyaline cartilage against proteoglycan degradation and decreased the severity of murine OA [21]. Inhibition of ADAMTS4 and ADAMTS5 by melatonin preserved the intact proteoglycan network in cartilage ECM, which may further protect the cartilage collagen network (especially Collagen II) from degradation induced by MMPs during OA progression [22]. In addition, we showed that melatonin significantly decreased MMP13 levels in IL-1β-induced chondrocytes, which may contribute to the attenuation of matrix loss in OA cartilage. However, the regulatory effect of melatonin on MMP13 is still disputable. Hong et al. suggested that melatonin promotes ECM remodeling in OA cartilage through the upregulation of MMP13 [23]. The different effects of melatonin on MMP13 may be due to the low dosage of melatonin and the complex interventions involved, which combined treadmill exercise with melatonin treatments of only up to 1 μM.
In addition to the transcriptional regulations, the present study is the first to demonstrate that melatonin prevents OAinduced cartilage destruction by regulating the expression of matrix-degrading enzymes at the posttranscriptional level. Previous studies have reported that several miRNAs are involved in the pathogenesis of OA, such as miR-9, miR-27b, miR-34a, miR-140, and miR-146 [24]; however, the effects of melatonin on modulating miRNAs in OA chondrocytes have not been reported. In this study, we first identified miR-140 as a melatonin-responsive miRNA in chondrocytes, characterized by a significant upregulation by melatonin with or without IL-1β stimulation. miR-140, a cartilage-specific miRNA, has been shown to inhibit the NF-κB signaling pathway in IL-1β-treated articular chondrocytes and to reduce cartilage ECM breakdown via downregulation of ADAMTS5 [25]. The negative effect of miR-140 on MMP9 has been reported to inhibit breast cancer invasion [26], while estrogen has been shown to suppress MMP13 expression through the upregulation of miR-140 in the development of menopausal arthritis [27]. In addition, the microarray results in this study demonstrated that melatonin modulated the expression of several miRNAs involved in cartilage development and OA progression. For example, we observed that melatonin treatment upregulated the expression of miR-95 and downregulated the expression of miR-9 and miR-204. miR-95 has been reported to promote cartilage matrix expression by directly inhibiting histone deacetylase 2/8 (HDAC2/8) [28]. Previous studies have demonstrated that miR-9 [29] and miR-204 [30] markedly increase in OA cartilage and contribute to the disrupted matrix homeostasis via upregulation of IL-6 expression and inhibition of the sulfated proteoglycan biosynthesis pathway, respectively. Therefore, further studies are necessary to investigate the effects of melatonin on other OA-related miRNAs and to unravel the regulation of melatonin-responsive miRNAs on cartilage matrix homeostasis.
One limitation of this study is that we investigated the therapeutic effects of melatonin on the early stage of OA; it is unknown whether intra-articular injection of melatonin will show the same protective effects on late-stage or other types of OA. Lim et al. showed that intra-articular injection of melatonin protected chondrocytes from oxidative stressinduced cartilage degradation through activation of the silent information regulator type 1 (SIRT1) signaling pathway [41]. Our future studies will seek to investigate the effects of melatonin-mediated miR-140 expression on late-stage OA.

Conclusions
In summary, we demonstrated that intra-articular injection of melatonin ameliorated the progression of OA by protecting cartilage from matrix destruction in surgically induced OA mice. In an in vitro IL-1β-induced arthritic environment, melatonin treatment enhanced the synthesis of matrix proteins and suppressed the expression of matrix-degrading enzymes in human articular chondrocytes. Further molecular experiments identified miR-140 as a melatonin-responsive miRNA and revealed that melatonin regulated the expression of proteinases at the posttranscriptional level by upregulating miR-140; it was evidenced by the fact that inhibition of miR-140 counteracts the antiarthritic effects of melatonin on OAinduced cartilage loss. Future studies are necessary to unveil the specific mechanisms by which melatonin modulates miR-140 expression and to investigate the therapeutic effects of melatonin on late-stage OA.

Data Availability
The data used to support the findings of this study are available from the corresponding authors upon request.

Conflicts of Interest
The authors declare that they have no conflict of interest.