The present study was designed to explore whether exosomal lncRNA-KLF3-AS1 derived from human mesenchymal stem cells (hMSCs) can serve as a positive treatment for osteoarthritis (OA). hMSCs and MSC-derived exosomes (MSC-exo) were prepared for morphological observation and identification by transmission electron microscopy and flow cytometry. IL-1β-induced OA chondrocytes and collagenase-induced rat model of OA were established for the further experiments. Lentivirus-mediated siRNA targeting KLF3-AS1 was transfected into MSCs for silencing KLF3-AS1. The real-time quantitative PCR and western blotting analysis were performed to examine the mRNA and protein levels of type II collagen alpha 1 (Col2a1), aggrecan, matrix metalloproteinase 13 and runt-related transcription factor 2. Cell proliferation, apoptosis and migration were evaluated by CCK-8 assay, flow cytometry and transwell assay. HE (hematoxylin and eosin) staining and immunohistochemistry were used for histopathological studies. MSC-exo ameliorated IL-1β-induced cartilage injury. Furthermore, lncRNA KLF3-AS1 was markedly enriched in MSC-exo, and exosomal KLF3-AS1 suppressed IL-1β-induced apoptosis of chondrocytes. Further in vivo investigation indicated that exosomal KLF3-AS1 promoted cartilage repair in a rat model of OA. Exosomal KLF3-AS1 promoted cartilage repair and chondrocyte proliferation in a rat model of OA, which might be an underlying therapeutic target for OA.
Osteoarthritis (OA) is a common age-related degenerative joint disease, characterized by progressive destruction of articular cartilage and cartilage loss caused by disequilibrium between anabolism and catabolism of chondrocytes and extracellular matrix (ECM) . Increasing studies regard the importance of cartilage repair in the treatment of OA. However, cartilage tissues have limited capacity for self-repair and remodeling. In recent years, cell-based therapy is considered to be one potential treatment strategy for cartilage repair .
Mesenchymal stem cells (MSCs) are a population of self-renewing and endogenous multipotent precursor cells which have the capacity to differentiate into osteocytes, chondrocytes, adipocytes, neurocytes, muscular cells and so on in specific conditions . MSCs could be routinely obtained from autogenous or allogeneic tissues . In vitro proliferation of bone marrow- or adipose tissues-derived MSCs produces more cells for transplantation. Hence, MSCs have become attractive sources of stem cells, which were used for repairing damaged tissues. MSC-based therapies were reported to restore the structure and function of damaged tissues. Indeed, the efficacy of autogenous or allogeneic MSCs in cartilage repair has been confirmed in animal studies  and human clinical trials .
Several studies have suggested that MSCs secrete a series of biologically active factors to mediate the damaged tissue microenvironment, thereby regulating subsequent cellular processes, including cell migration, proliferation, differentiation and matrix synthesis . Exosomes are defined as 60–140 nm microvesicles originating from multivesicular bodies, which are released from several cell types into the extracellular space, where they play vital roles in intercellular communication via the transfer of proteins, mRNAs, microRNAs and long noncoding RNAs (lncRNAs) [8,9]. Recently, accumulating evidence has strongly implied that MSC-derived exosomes (MSC-exo) promote cartilage repair [10–12]. A few lncRNAs have been found enriched in exosomes and function as novel biomarkers and therapy targets for various diseases. For example, Chen et al.  showed that macrophage-derived exosomal lncRNA growth arrest-specific 5 (GAS5) deteriorated the apoptosis of vascular endothelial cells after oxidized low-density lipoprotein stimulation, suggesting that down-regulation of exosomal lncRNA GAS5 might be an alternative strategy for the treatment of atherosclerosis. However, whether MSC-derived exosomal lncRNAs can serve as a therapeutic target for OA remains unclear.
The MSC-derived exosomal lncRNA KLF3-AS1 (Ensembl: ENST00000440181; Refseq NR 026804.1) has been previously reported in ExoCarta (http://www.exocarta.org), a manuallydatabase of exosomal molecular components. In the present study, we further investigated the biological roles of exosomal KLF3-AS1 in regulating the function of chondrocytes in OA.
Materials and methods
Human MSCs (hMSCs) were purchased from American Type Culture Collection (ATCC, Manassas, VA, U.S.A.) and cultured in Dulbecco's modified Eagles medium (DMEM)/F12 (Gibco, Carlsbad, CA, U.S.A.) supplemented with 10% fetal bovine serum (FBS, Gibco) in a humidified atmosphere containing 95% air and 5% CO2 at 37°C and observed under transmission electron microscopy (TEM; Hitachi, Tokyo, Japan).
MSCs of the third generation were washed twice with PBS for cell sorting, diluted to 1.0–2.0 × 107 cells/ml with PBS and then incubated with phycoerythrin-conjugated anti-CD106, anti-CD29, anti-CD34, fluorescein isothiocyanate-conjugated antibody CD44 and rat anti-mouse allophycocyanin-conjugated antibody CD45 (1 : 100; BD Biosciences, San Jose, CA, U.S.A.) away from light at 4°C for 40 min. Subsequently, cells were washed with cold PBS, centrifuged at 1500g for 5 min, resuspended in PBS and ultimately performed using a FACScan flow cytometer (BD Biosciences).
Extraction and identification of MSC-derived exosomes
MSC-conditioned medium was collected after 48 h culture, followed by centrifugation at 3000g for 15 min. The medium was incubated with 0.5 volume of total exosome isolation reagent (Thermo Scientific, Waltham, MA, U.S.A.) overnight at 4°C. The mixture was centrifuged at 12 000g for 1 h and the supernatant was discarded. Exosome pellets were resuspended in PBS and prepared for identification.
The expressions of surface markers TSG101 and CD63 were tested by flow cytometry. Briefly, after washed with 0.1% BSA for three times, exosome-coated beads were incubated with anti-TSG101 (1 : 500, Abcam, Cambridge, U.K.) and anti-CD63 (1 : 200, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, U.S.A.) for 30 min, followed by incubation with anti-IgG secondary antibody (1 : 1000; Cell Signaling Technology, Boston, MA, U.S.A.) and analyzed on a FACSCalibur flow cytometer (BD Biosciences).
Culture and identification of chondrocytes
Male Sprague–Dawley (SD) rats maintained under specific pathogen-free conditions, purchased from the Animal Experiment Center of Nanjing University and were used in the present study. All protocols were approved by the Ethic Committee of Luhe People's Hospital of Nanjing. SD rats were killed by cervical dislocation. Bilateral knee articular cartilages were obtained from SD rats, rinsed with PBS and cut into 1 mm3 pieces. Cartilage tissues were then digested in 0.2% collagenase II for 4 h. Subsequently, 10% FBS was added to terminate digestion. Chondrocytes were collected and resuspended in a DMEM/F12 medium containing 20% FBS at 37°C with 5% CO2after filtration and centrifugation at 800g for 10 min. The medium was replaced every three days.
The articular chondrocytes were identified by toluidine blue staining. In brief, second-generation chondrocytes were digested by 0.25% trypsin-EDTA and transferred on the glass cover slides of six-well plates. After fixing with 4% paraformaldehyde for 30 min, chondrocytes were stained with 1% toluidine blue for additional 30 min, washed with 95% ethanol and observed under a microscope.
Co-culture of chondrocytes and MSC-exo
After phenotypic characterization, rat chondrocytes were treated with a normal medium as a negative control, IL-1β (10 mg/ml), PBS and MSC-exo (1, 5 and 10 µg/ml) for 24 h to investigate the effect of MSC-exo on IL-1β-induced chondrocyte injury. The real-time quantitative PCR (qRT-PCR) and western blotting analysis were performed to examine the mRNA and protein levels of chondrogenic and ECM gene type II collagen alpha 1 (Col2a1), aggrecan, matrix metalloproteinase 13 (MMP13) and runt-related transcription factor 2 (RUNX2). We also observed the proliferation, apoptosis and migration of chondrocytes.
Lentivirus-mediated siRNA targeting KLF3-AS1
The stable knockdown of lncRNA KLF3-AS1 in MSCs was generated by transducting a lentiviral-mediated expression of siRNA-specific target of lncRNA KLF3-AS1 that was synthesized by GenePharma (Shanghai, China). The siRNA sequence was as follows: 5′-GCCUCUUCUUAACUUUGAUTT-3′. The virus-transfected MSCs were harvested after 48 h post-transfection using an X-tremeGENE siRNA transfection reagent (Roche, Indianapolis, IN, U.S.A.), and the knockdown efficiency was tested by qRT-PCR. To further explore the underlying roles of exosomal KLF3-AS1 in chondrocyte injury, rat chondrocytes were, respectively, treated with the normal medium, IL-1β, PBS, MSCscramble-Exos and MSCsi-KLF3-AS1-Exos for 24 h. The expressions of Col2a1, aggrecan, MMP13 and RUNX2 at mRNA and protein levels along with cell proliferation, apoptosis and migration were evaluated.
In vivo experiments
Consistent with the in vitro study, SD rats were randomly divided into five groups (n = 6 per group): normal, OA, PBS, MSCscramble-Exos and MSCsi-KLF3-AS1-Exos. For the establishment of experimental OA model, collagenase II was dissolved in sterile PBS (pH 7.4), filtered with a 0.22-µm membrane and then injected into the knee joint cavity of SD rats under anesthesia, as described previously . After 21 days of operation, rats were respectively injected with PBS, MSCscramble-Exos and MSCsi-KLF3-AS1-Exos. All rats were killed 8 weeks after injection and knee articular cartilage was then isolated for subsequent experiments.
RNA extraction and qRT-PCR analysis
Total RNA was extracted from chondrocytes and cartilage tissues using Trizol reagent (Invitrogen, Carlsbad, CA, U.S.A.) and reverse-transcribed to cDNA using the first-strand cDNA synthesis kit (Tiangen Biotech, Beijing, China; Cat#KR105) according to the manufacturer's protocol. Relative mRNA expression levels of Col2a1, aggrecan, MMP13, RUNX2 and KLF3-AS1 were detected using the SYBR® Premix Dimer Eraser kit (Takara Shiga, Japan; Cat#DRR420A) performing on an ABI 7500 Real-Time PCR system (Applied Biosystems, Carlsbad, CA, U.S.A.) and calculated by the method. β-Actin was used as an internal control.
Total protein was extracted by lysis buffer and centrifugated at 12 000g for 20 min. Aliquots of 20 µg protein from each sample were blended with loading buffer, denatured in boiling water for 5 min, separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) and electrotransferred to polyvinylidene fluoride membranes which were blocked with 5% skim milk at room temperature for 2 h followed by elution with Tris-buffered saline Tween (TBST). The membranes were probed with primary antibodies against Col2a1 (1 : 500, Santa Cruz Biotechnology, Inc.), aggrecan (1 : 100, Santa Cruz Biotechnology, Inc.), MMP13 (1 : 1000, Santa Cruz Biotechnology, Inc.), RUNX2 (1 : 200, Santa Cruz Biotechnology, Inc.), β-actin (1 : 1000; Santa Cruz Biotechnology, Inc.) overnight at 4°C, and then blots were incubated with horseradish peroxidase-conjugated secondary antibodies (1 : 2000; Santa Cruz Biotechnology Inc.) at room temperature for 2 h after washed with TBST. The protein levels were quantified using enhanced chemiluminescence (Thermo Scientific).
Cell proliferation, apoptosis and migration assay
For cell viability assay, rat chondrocytes were seeded into 96-well plates at a density of 2 × 103 cells/well and cultured for 48 h. Cell proliferation was determined using CCK-8 kit (Beyotime, Shanghai, China; Cat#C0037) according to the manufacturer's protocol. Absorbance was detected at the wavelength of 450 nm.
The apoptosis of rat chondrocytes was analyzed using flow cytometry. After different treatments, chondrocytes were collected by centrifugation, then washed once with PBS followed by removing PBS by centrifugation again. After fixed with ethyl alcohol at 4°C for 2 h, these cells were suspended in PBS. Afterwards, cells were stained with propidine iodide at 4°C under darkness for 30 min after filtration and centrifugation. Finally, cells were recorded using flow cytometry (BD Biosciences).
Cell migration was analyzed with Transwell cell culture chambers. Chondrocytes were resuspended in a serum-free medium, and 200 µl of the cell suspension was added to the upper chamber, while 600 µl of DMEM containing 10% FBS was added to the bottom wells of the 24-well chamber. After 24 h incubation, chondrocytes were washed with PBS, fixed with 4% formaldehyde for 15 min at room temperature and stained with 0.5% crystal violet for 20 min. Finally, migratory cells were counted from 10 different fields of each filter.
Hematoxylin and eosin staining
Rat knee articular cartilages were fixed with 10% formalin, decalcified with 20% EDTA and embedded in paraffin. Serial 5 µm microsections were prepared and stained with hematoxylin and eosin. The stained specimens were observed under a light microscope. The pathological changes were used to assess the severity of cartilage degradation based on the Mankin scoring system, as previously described .
Immunohistochemical detection of anti-proliferating cell nuclear antigen
Paraformaldehyde-fixed tissue sections were permeabilized with 0.2% Triton X-100, blocked with 5% BSA and incubated with anti-proliferating cell nuclear antigen (PCNA) antibody (1 : 1000; Santa Cruz Biotechnology, Inc.) overnight at 4°C. After washed with PBS, the slides were reacted with biotinylated secondary antibody (Thermo Scientific) for 1 h at room temperature. Finally, the sections were counterstained with diaminobenzidine (DAB) and mounted on gelatin-coated slides. Images were taken at five randomly selected fields, and the average number of PCNA-positive cells was calculated.
All results were expressed as statistically mean ± standard deviation (s.d.). Measurement data were conducted using SPSS version 18.0 (SPSS, Inc., Chicago, IL, U.S.A.). Analysis of variance and Student's t-test were used to evaluate the difference among groups and values of P < 0.05 were statistically significant. The flowchart of the experiment is shown in Supplementary Figure S1.
Morphological observation and identification of MSCs and MSC-exo
The ultrastructure of hMSCs was observed under TEM. TEM revealed that after 72 h of culture, hMSCs showed uniform morphology and spindle-shaped appearance and were arranged in whorls (Figure 1A). Furthermore, flow cytometry displayed that almost all cells were positive for classical MSC surface markers CD106, CD29 and CD44, and negative for hematopoietic lineage and endothelial markers CD34 and CD45 (Figure 1B), suggesting that these cells belonged to high-purity MSCs. MSC-exo identification was also done by TEM and flow cytometry. Flow cytometry confirmed enrichment of the exosomal surface markers TSG101 and CD63 in purified exosomal fractions (Figure 1C). The flow cytometry results of isotype controls for CD106, CD29, CD44, CD34, CD45, TSG101 and CD63 are shown in Supplementary Figure S2.
Morphological observation and identification of MSCs and MSC-exo.
MSC-exo suppressed IL-1β-induced rat cartilage injury
The third-generation rat chondrocytes presented as triangle, polygon or irregular shape. After 3 days of culture, the cells were gradually integrated in the form of ‘paving stone’ shape (Figure 2A). Toluidine blue staining showed blue nuclei with purple or purple-red granules in the cytoplasm (Figure 2B).
Effect of MSC-exo on the regulation of rat cartilage injury.
To determine whether MSC-exo exerted therapeutical effect on IL-1β-induced OA chondrocytes, we detected the effects of MSC-exo on proliferation, migration and apoptosis of OA chondrocytes. Rat chondrocytes were treated with either normal medium, IL-1β, PBS or MSC-exo with different concentrations. As revealed in Figure 2C,D, chondrogenic genes col2A1 and aggrecan were both decreased at mRNA and protein levels by IL-1β stimulation but subsequently increased by MSC-exo treatment in a concentration-dependent manner. In contrast, expression levels of chondrocyte hypertrophy markers RUNX2 and MMP13, which were increased by IL-1β induction, were both significantly decreased by MSC-exo treatment. Moreover, IL-1β significantly reduced the proliferation and migration and promoted apoptosis, while observably increased proliferation and migration, and decreased apoptosis was detected in OA chondrocytes after treatment with MSC-exo (Figure 2E–G). These data suggested that MSC-exo alleviated IL-1β-induced OA chondrocytes by promoting chondrocyte proliferation, and migration, and decreasing apoptosis.
Exosomal KLF3-AS1 derived from MSCs inhibited IL-1β-induced apoptosis of chondrocytes
To verify whether lncRNA KLF3-AS1 existed in MSCs and MSC-exo, qRT-PCR analysis was performed. As expected, lncRNA KLF3-AS1 was found in rat bone marrow MSCs and MSC-exo (Figure 3A). Using the lentivirus-mediated siRNA gene knockdown system, we knocked down lncRNA KLF3-AS1 expression in MSC-exo, since more than 50% of lncRNA KLF3-AS1 expression was decreased in the gene knockdown system (Figure 3B). Then, we treated rat chondrocytes with a normal medium, IL-1β, PBS, MSCscramble-Exos and MSCsi-KLF3-AS1-Exos for 24 h, and measured the expression levels of Col2a1, aggrecan, MMP13 and RUNX2 and cell proliferation, apoptosis and migration. Compared with the control group, IL-1β led to a remarkable decrease in col2A1 and aggrecan and a prominent increase in RUNX2 and MMP13 (Figure 3C,D), reduced chondrocyte viability (Figure 3E), accelerated apoptosis (Figure 3F) and suppressed migration (Figure 3G). Besides, MSC-exo treatment up-regulated Col2a1 and aggrecan, inhibited MMP13 and RUNX2, promoted cell proliferation and migration but suppressed apoptosis, which were reversed by lncRNA KLF3-AS1 silencing in MSC-exo. These findings highlighted the importance of lncRNA KLF3-AS1 in the inhibition of IL-1β-induced apoptosis of chondrocytes.
Effect of exosomal lncRNA KLF3-AS1 on apoptosis of chondrocytes.
Exosomal KLF3-AS1 promoted cartilage repair in a rat model of OA
To validate in vitro the protective effects of exosomal lncRNA KLF3-AS1 on articular cartilage, MSCscramble-Exos and MSCsi-KLF3-AS1-Exos were delivered into rats of OA model following intra-articular injection, as recently described . In a parallel experiment, PBS was injected as a negative control. The chondrocyte numbers and cartilage thickness were significantly decreased, as the Mankin score was significantly elevated (Figure 4A), PCNA positive cells were increased (Figure 4B), col2A1 and aggrecan expressions were down-regulated, and RUNX2 and MMP13 expressions were up-regulated (Figure 4C,D) in the OA model group than in the normal group. However, MSCscramble-Exos group exhibited a decreased Mankin score and RUNX2 and MMP13 expressions, as well as increased PCNA-positive cells and col2A1 and aggrecan expressions as relative to the PBS group, whereas exosomes isolated from the lncRNA KLF3-AS1 knocked down MSCs led to an opposite effect. These results indicated that lncRNA KLF3-AS1 existed in MSC-derived exosomes and exosomal lncRNA KLF3-AS1 promoted cartilage repair.
The role of exosomal lncRNA KLF3-AS1 in cartilage repair in a rat model of OA.
The limited self-regeneration potential of articular cartilage partly results from the inability of adjacent articular chondrocytes to migrate and produce matrix . In the present study, MSC-exo promoted chondrocyte proliferation and migration and inhibited apoptosis. The therapeutical effect of MSC-exo on OA is related to newly discovered exosomal lncRNA KLF3-AS1.
OA is a chronic degenerative joint disease featured with the degeneration and destruction of the articular cartilage and osteophyte formation . LncRNAs, which were defined as a novel class of RNA transcripts more than 200 nucleotides with narrow protein coding functions, were involved in regulating various cell biological processes containing cell growth, metabolism, differentiation and apoptosis . Several lncRNAs play protective roles in OA progression. For example, Pan et al.  reported that MALAT1 alleviated LPS-induced chondrogenic cell inflammatory injury by up-regulating miR-19b through Wnt/β-catenin and NF-κB pathways, suggesting the critical roles of MALAT1 in OA. Xu et al.  found that MEG3 exerted its anti-proliferation and pro-apoptosis in OA progression by regulating miR-16 and SMAD7. Moreover, Li et al.  showed that lncRNA PVT1 promoted the apoptosis of chondrocytes by acting as a sponge for miR-488-3p in OA. Current pharmacological treatments and physical therapy could not restore articular cartilage regeneration in spite of pain relief and joint mobility improvement . Although surgical arthroplasty can solve these limitations, it is easy to cause infection . Recently, stem cell therapy has been rapidly advancing in regenerative medicine for OA. MSCs can be isolated from various adult tissues for regenerative therapy due to its ability to differentiate into chondrocytes, together with the immunosuppressive and anti-inflammatory properties.
More importantly, the ability of MSCs to secrete bioactive molecules contributes to providing a regenerative microenvironment for injured tissues . Among them, exosomes, described as a kind of extracellular vesicle secreted by MSCs under resting condition or certain types of stress such as hypoxia, irradiation or oxidative injury, can function as messengers between MSCs and differentiated cells to induce physiological changes . So far, abundant studies have emphasized the importance of MSC-exo in alleviating myocardial ischemia–reperfusion injury, kidney injury and retinal damage and improving graft versus host disease, wound healing and liver regeneration [26,27]. In our study, we observed that exosomes derived from hMSCs overturned IL-1β-induced down-regulation of chondrogenic genes col2A1 and aggrecan, up-regulation of hypertrophy markers RUNX2 and MMP13, inhibition of chondrocyte proliferation and migration, and promotion of apoptosis in a dose-dependent manner. These findings confirmed our conjecture that MSC-exo exerted therapeutical effect on OA.
In recent years, abundant studies provided strong evidence that MSC-exo participated in the damage repair process by delivering functional protein and RNAs. For example, Chen et al.  reported that increased exosomal miR-223 was associated with the onset, severity and short-term outcomes of acute ischemic stroke, suggesting exosomal miR-223 as a novel biomarker for ischemic stroke. Wang et al.  found that exosomal miR-21 promoted cardioprotective effect by inhibiting myocardial apoptosis via the PTEN/Akt pathway. Tao et al.  showed that exosomes derived from miR-140-5p-overexpressing synovial MSCs promoted the proliferation and migration of rat chondrocytes, which in turn ameliorated OA. Searching the differently expressed RNAs in MSC-exo, we noted a novel lncRNA KLF3-AS1, which localizes at chromosome 4p14 and is enriched in MSC-exo according to the exocarta database. The further qRT-PCR analysis confirmed the existence of lncRNA KLF3-AS1 in hMSCs and MSC-exo. Silencing lncRNA KLF3-AS1 in MSC-exo could inhibit proliferation and migration and promote apoptosis of chondrocytes, thereby in vitro aggravate IL-1β-induced OA. In vivo, we used MSC-exo following intra-articular injection in the rat model of OA. The knee joint cartilage damage, which was caused by OA, was deteriorated by MSCsi-KLF3-AS1-Exos.
However, one of the limitations of our study is that we have not provided any radiographic images of the joint structures showing the reduced joint space and osteophyte formation induced by an OA model .
Here, we concluded that MSCs derived exosomal lncRNA KLF3-AS1 transplantation could facilitate cartilage repair by promoting chondrocyte proliferation and migration and inhibiting apoptosis. Therefore, our findings highlighted the possible mechanism for OA therapy by cellular delivery of exosomal lncRNA KLF3-AS1.
Dulbecco's modified Eagles medium
fetal bovine serum
growth arrest-specific 5
matrix metalloproteinase 13
proliferating cell nuclear antigen
runt-related transcription factor 2
Tris-buffered saline Tween
transmission electron microscopy
Y.L., C.W., F.Z. and F.L. designed the experiments. Y.L., R.Z. and Z.W. performed the experiments. F.Z. and F.L. wrote the manuscript. All the authors approved the final manuscript.
This work was supported by the Science and Technology Development Planning Project of Nanjing  and the Medical Science and Technology Development Project of Nanjing [YKK17244 and YKK17243].
The Authors declare that there are no competing interests associated with the manuscript.