Mesenchymal stromal cell-based therapy for cartilage regeneration in knee osteoarthritis

Osteoarthritis, as a degenerative disease, is a common problem and results in high socioeconomic costs and rates of disability. The most commonly affected joint is the knee and characterized by progressive destruction of articular cartilage, loss of extracellular matrix, and progressive inflammation. Mesenchymal stromal cell (MSC)-based therapy has been explored as a new regenerative treatment for knee osteoarthritis in recent years. However, the detailed functions of MSC-based therapy and related mechanism, especially of cartilage regeneration, have not been explained. Hence, this review summarized how to choose, authenticate, and culture different origins of MSCs and derived exosomes. Moreover, clinical application and the latest mechanistical findings of MSC-based therapy in cartilage regeneration were also demonstrated.


Background
Osteoarthritis is a common and disabling condition that represents substantial health and socioeconomic costs with notable implications for the individuals affected and healthcare systems [1,2]. Global percentage change of years lived with disability in counts between 2006 and 2016 was 31.5% [3]. Knee osteoarthritis (KOA) is the most prevalent subtype of osteoarthritis [4] that shows symptoms with pain, swell, stiffness, and loss of mobility mainly in the aging and obese populations [5]. The pathogenesis of KOA is complicated, not only associated with the "wear and tear, " which is called mechanical stress [6]. Microenvironmental and genetic factors interact during deterioration that ultimately leads to degeneration of articular cartilage, intraarticular inflammation with synovitis, and changes in subchondral bone [7,8]. Extracellular matrix (ECM) produced and secreted by chondrocytes and synovial fluid secreted by fibroblast-like synoviocytes are the two main important substances to maintain the microenvironment [9,10]. Nonetheless, chondrocytes constitute only 2% of cartilage volume [11,12] and may the first be activated by inflammatory signals originating from synovium or subchondral bone [13], which alter the balance between synthesis and degradation of ECM and cause the limited potential for self-regeneration of articular cartilage. Recommended treatment options range from non-care currently limited to pain control and merely improve the regeneration of articular cartilage in KOA.
Cell-based therapy and novel approaches using mesenchymal stromal cells (MSCs) or exosomes secreted by MSCs are presented as alternative cell-based sources to chondrocytes, which show potential for cartilage regeneration in KOA [14,15]. The International Society for Cell & Gene Therapy refers MSCs as a bulk population with notable secretory, immunomodulatory, and homing properties. The minimal criteria include being plastic adherent, expressing specific surface markers, and capable of in vitro differentiation into adipocyte, chondrocyte, and osteoblast lineages [16]. MSCs and secreted exosomes (Exos) maintain therapeutical potentials for cartilage regeneration, including balancing metabolic
Besides CD44, CD73, CD90, CD105, some protein markers have the potentials to be new and specific markers [73]. Stro-1 and CD271 are cell membrane single-pass type I proteins that translocate from the endoplasmic reticulum to the cell membrane in response to the depletion of intracellular calcium. However, it is unclear whether Stro-1 expression correlates with multipotency [46]. SSEA-4 is an embryonic stem cell marker, and CD146 is detected on perivascular cells around venules [69]. Erdogan et al. reported that AT-MSCs in New Zealand rabbits did not express CD73 and CD90 [58], while Chen et al. detected the expression of CD90 [59]. Some markers appeared already at the optic vesicle stage but did not remain highly expressed in the later differentiation stage [74].

Culture
MSCs are spindle-shaped and adherent cells, capable of proliferation, self-renewal, and differentiating into cells of multi-lineage. One of the characteristic features of MSCs is adhering to tissue culture plastic and generating colonies when plated at low densities [78]. MSCs growing from individual foci, or colonies from the microscopic view, and these colonies generated from progenitor cells have been called the colony-forming unit fibroblast [79]. The ability of MSCs to undergo chondrogenic, osteogenic, and adipogenic differentiation has been reported in vitro and in vivo.
Chondrogenic differentiation largely depends on the culture conditions. Mediators capable of promoting chondrogenesis, such as transforming growth factor-beta (TGF-β), have been elucidated using simplified in vitro models [80]. Recently, Yin et al. indicated that MSCs differentiated into mature chondrocytes after 21 days of co-culture with ECM-derived particles in a microgravity environment without exogenous TGF-β3 [81]. Chondrogenesis can be achieved either in 2-dimensional or 3-dimensional culture systems in vitro. The 3-dimensional culture system facilitates greater cell contacts and interactions of cells with the ECM, allowing cells to adapt to their native morphology [82]. Moreover, the efficiency of chondrogenesis tends to be lower in the 2-dimensional culture system. The scaffold-free 3-dimensional cultures provide a high-density cell culture environment and are commonly classified as pellet or micromass culture systems [83]. In general, the induced cartilage was more similar to hyaline cartilage in the micromass culture technique, while the pellet culture is more useful for clinical applications [84]. Moreover, platelet-rich plasma (PRP), MSCs, and chondrocytes co-culture would favor chondrogenesis without hypertrophic and pathologic responses [85]. Chondrocytes cultured with MSC-Exos enhanced proliferation and chondrogenesis [86]. The flow diagram of applying MSC-based therapy is presented in Fig. 1.

Impact on chondrocyte
Increased chondrogenesis, improved proliferation, reduced apoptosis, and maintenance of autophagy of  [87]. Moreover, human AT-MSCs showed a high chondrogenic potential supported by the increased expression of sex-determining region of Y chromosome-box transcription factor 9 (SOX9), as a chondrocyte precursor marker [88]. One study reported a reduced matrix synthesis and low proliferation rate of chondrocytes in the injured cartilage, while the restored proliferation ki67 enhanced after injecting human umbilical cord-derived MSCs [89]. Additionally, the proliferation rate of chondrocytes was increased while co-culturing with BM-MSCs [90]. In addition, synovial MSC-Exos have the potential to improve proliferation and migration of chondrocytes in vitro and in vivo [91]. TGF-β1-stimulated MSC-Exos increased the cell viability of the chondrocyte-restricted rat cell line [92]. Rat BM-MSC-Exos under hypoxic pretreatment promoted the proliferation and migration of chondrocytes [93]. Apoptosis of chondrocytes is another characteristic of the progression of OA. Chen et al. reported a notable decrease of the proportion of dUTP nick end labeling positive cells that indicated apoptosis in cartilage after injection with MSCs compared with the phosphate-buffered saline [94]. Apoptosis could be a potential therapeutic pathway for KOA [95]. Moreover, the apoptosis of chondrocytes was remarkably inhibited by coculturing with BM-MSCs under hypoxia [90]. Exos also have a positive effect on preventing cartilage damage. Interleukin (IL)-1β treatment inhibited cell viability and DNA synthesis activity, and enhanced apoptosis of chondrocytes. However, BM-MSC-Exos treatment promoted cell viability and DNA synthesis activity with a lower apoptosis rate [96]. Wu et al. also reported AT-MSC-Exos inhibited cell apoptosis and enhanced autophagy [97]. Autophagy can be detected combined with apoptosis in KOA. Autophagy serves as an adaptive response under environmental changes that maintain the survival of chondrocytes by preserving energy metabolism in cells. The levels of Beclin-1 and microtubule-associated protein light chain 3, as two vital proteins in the autophagic process, were detected to manifest a reduced autophagic activity in KOA animal model that treated with phosphate-buffered saline, whereas MSCs therapy maintained almost the same level as that of the normal group [94].

Impact on the ECM
Regulating the balance of synthesis and catabolism of the ECM is essential to treat degenerative diseases, such as KOA. Matrix metalloproteinases (MMPs) refer to a family of zinc-dependent ECM remodeling endopeptidases that degrade the ECM. On the other hand, the tissue inhibitors of MMPs (TIMPs) are important regulators of ECM turnover, tissue remodeling, and cellular behavior that inhibit the proteolytic activity of MMPs within the ECM. BM-MSCs balanced the ratio of MMP-13 to TIMP-1 in cartilage and reduced the expression of cartilage hypertrophic markers such as collagen (Col)-X, fibroblast growth factor receptors 1-3, parathyroid hormone-related protein, and MMP-13 [98]. Several studies reported the higher expression of gene Col 2α1 in KOA cartilage, which encoded the α-1 chain of Col II, after MSC-based therapy [24,99]. Moreover, the expressions of a disintegrin and metalloproteinase with thrombospondin motifs-5 and MMP13 in cartilage were significantly downregulated after treating with human umbilical cord-derived MSCs [89]. Hyaluronan synthase-1 mRNA expression was upregulated in BM-MSCs after co-culture with chondrocytes from the KOA model, whereas hyaluronidase-1 was downregulated [100]. Chen et al. revealed that BM-MSC-Exos could promote the expression of Col II, SOX9, and aggrecan while negatively regulating the expression of chondrocyte hypertrophy markers MMP-13 in mouse models of post-traumatic KOA [101].

Impact on the inflammatory cytokines
Inflammatory response plays an important role in the pathogenesis of KOA. The most important groups controlling the disease seem to be pro-inflammatory cytokines and anti-inflammatory cytokines, which have an antagonistic effect. The former mainly includes IL-1β, tumor necrosis factor-α (TNF-α), IL-6, IL-15, IL-17, and IL-18. Another is formed by TNF-stimulated gene 6, IL-4, IL-10, IL-13, IL-37, and others. IL-37 partly rescued IL-1β and impaired cartilage formation of MSCs. This effect contributed to a lower MMP3 expression and an increased ratio of Col-II/ Col-I without increasing hypertrophy markers [102]. Pro-inflammatory (M1) macrophages are associated with a high production of pro-inflammatory mediators such as TNF-α, IL-6, IL-1β, and IL-12, and are required for T cell activation. These cytokines induce destructive processes in chondrocytes manifesting a lower expression of Col-II and aggrecan synthesis [103].
Some researchers found decreased expressions of inflammatory and catabolic markers including IL-1β, TNF-α, and MMP13, after AT-MSCs injection [104], BM-MSCs injection [105], or induced pluripotent stromal cell-chondrocytes transplant in KOA model. However, no difference in Col-II and Col-I expression was found between transplanted cartilage and other groups [106]. One study determined that both IL-1β and TNF-α immunostaining in chondrocytes in the cartilage were significantly enhanced after human umbilical cordderived MSCs treatment and reserved almost back to normal tissue. Additionally, umbilical cord-derived MSCs therapy also led to increased expression of anti-inflammatory factors, TNF-stimulated gene 6, and IL-1 receptor antagonist, in the articular chondrocytes [89]. Human AT-MSCs seem to adapt and respond better to both inflammatory stimuli and autologous protein solution than BM-MSCs in vitro [107]. Co-culture with AT-MSCs counteracted the IL-1β-induced mRNA upregulation of the MMP-3, MMP-13, TNF-α, and IL-6 in chondrocytes. Importantly, AT-MSCs increased the expression of the anti-inflammatory cytokine IL-10 in chondrocytes [108,109].
As for treatment with Exos, injection of miR-9-5p-contained Exos alleviated the inflammation in KOA, which was evidenced by downregulated levels of inflammatory factors and reduced oxidative stress injury [110]. Lu et al. reported synovial MSC-Exos enhanced IL-1β-induced cell proliferation, whereas inhibited apoptosis and inflammation and the target relationship of miR-26a-5p and phosphatase and tensin homologue were predicted and confirmed [111]. Moreover, Zhe et al. investigated miR-26a-5p in human BM-MSCs exerted an alleviatory effect on the damage of the synovial fibroblasts [112].

Impact on the immunity
Macrophages could play a crucial role in modulating inflammation during the pathogenesis of KOA via various secreted mediators. These cells can polarize to pro-inflammatory and anti-inflammatory (M2) phenotypes. One study exposed AT-MSCs to osteoarthritic synovial fluid for two days for determining the effect of their secretome on differentiation of monocytes into pro-inflammatory M1-like macrophages and mature dendritic cells, and the effect on T cell proliferation and expansion of T regulatory cells. The results suggest that the exposure of AT-MSCs upregulated the immunosuppressive factors that induce monocytes into the M2-like phenotype and inhibit differentiation of monocytes into mature dendritic cells. Only the secretome of exposed AT-MSCs was detected to inhibit proliferation of T cells and promote T regulatory cells expansion [113].
More than 240 micro-RNAs were found in AT-MSCs and accounted for most of the genetic message that protected chondrocytes and M2 macrophage polarizing. Ragni et al. [114] have confirmed an increased M2 phenotype marker CD163 and reduced the chondrocyte inflammation marker vascular cell adhesion molecule-1 on inflamed macrophages and chondrocytes. BM-MSCs-Exos and AT-MSCs-Exos have reported relieving KOA by promoting the phenotypic transformation of synovial macrophages from M1 to M2 [115,116]. Moreover, TGF-β1-stimulated BM-MSC-Exos reduced pro-inflammatory factors by promoting M2 polarization of synovial macrophages [117].

Impact on the mitochondrial function
Aging and exposure to stress would determine the chondrocyte phenotype in osteoarthritis and age-related mitochondrial dysfunction and associated oxidative stress might induce senescence in chondrocytes [118]. The mitochondrial transfer was found from BM-MSCs to osteoarthritis chondrocytes. One study showed an increased mitochondrial membrane potential when cocultured with mitochondria from MSCs compared with chondrocytes without mitochondria transfer. The activity of mitochondrial respiratory chain enzymes and the content of adenosine-triphosphates were significantly improved [119].

Impact on the paracrine effect
Some researchers thought the paracrine effect of MSCs was mediated or performed by MSC-derived extracellular vesicles, while others support the induction of paracrine effect was independent of extracellular vesicles [120]. In general, the paracrine effect and Exos both represent cell-to-cell contact and biological information delivery. Extracellular vesicles have been traditionally classified into four subtypes, mainly based on their origins and sizes. MSC-Exos, as the smallest extracellular vesicles, have recently been suggested as a mechanism for their therapeutic potentials [121]. Figure 2 shows the functions of MSC-based therapy.

Pathways or axis
Regulations of inflammation, immunoregulatory, and metabolic pathways and the expression of key molecules by the MSCs-based therapy have been revealed in recent research. The nuclear factor-kappaB (NF-κB) family plays crucial roles in various biological processes including mechanical processes, immunity, inflammation, and oxidative stress response, which could be activated by chemokines, pro-inflammatory cytokines, and degradation factors. NF-κB comprises five members: RelA (p65), c-Rel, RelB, NF-κB1 (p50), and NF-κB2 (p52) [122]. Yu et al. reported that the deletion of RelA in culturing BM-MSCs could increase the chondrogenic differentiation [123]. Treatment with human AT-MSCs significantly reduced the content of signal transducer and activator of transcription 3, which is a DNA-binding molecule that regulates the levels of many cytokines. Activation of signal transducer and activator of transcription 3 leads to increased pro-inflammatory cytokine production and immune responses [124]. BM-MSC-Exos inhibited mitochondrial-induced apoptosis in response to IL-1β, involving p38 mitogen-activated protein kinase (MAPK), extracellular regulated protein kinases (ERK), and protein kinase B pathways (AKT) [125].
Chronic inflammation may contribute to stress-induced senescence of chondrocytes and cartilage degeneration during the progression. Treatment of chondrocytes with MSCs down-regulated senescence markers induced by IL-1β including senescence-associated β-galactosidase activity, accumulation of γH2AX foci, and morphological changes with enhanced formation of actin stress fibers. Additionally, the treatment reduced the activation of MAPK, ERK1/2, and p38 and to a lower extent the phosphorylation of c-jun N-terminal kinase 1/2 [126], which represents a classical inflammatory pathway. Besides this, inflammation is a prerequisite for the protective effect of AT-MSCs. Van Dalen et al. proved that local application of AT-MSCs in KOA joints led to rapid clustering of polymorphonuclear cells around AT-MSCs, while IL-1β stimulated this clustering which reduces the pro-inflammatory activity of the polymorphonuclear cells in vitro [127].
In human chondrocytes, oxidative stress may lead to DNA damage and senescence. In addition, reactive oxygen species (ROS) production is important in signaling pathways activated by IL-1β in chondrocytes. As In chondrocytes from the KOA model, IL-1β quickly induced the production of ROS and enhanced levels of 4-hydroxy-2-nominal-modified proteins, whereas AT-MSCs reduced the level of them [126]. In addition, human AT-MEC-Exos reduced the production of oxidative stress in OA chondrocytes stimulated with IL-1β and resulted in an upregulation of peroxiredoxin 6 [128]. Cell proliferation activated by AT-MSCs in KOA rabbits may be specifically regulating the glycogen synthase kinase-3β (GSK-3β)/cyclin D1/cyclin-dependent kinase (CDK)4/CDK6 pathway. It reduced the elevated serum level of cartilage oligomeric matrix protein, blocked increases in the mRNA, and protein expression of GSK3β while decreasing the mRNA and protein expression of cyclin D1/CDK4 and cyclin D1/CDK6 in cartilage [129]. Wu et al. demonstrated BM-MSCs increased the levels of the ECM proteins Col-II and SOX9 and decreased chondrocyte apoptosis and inflammation by upregulating the mediators of the autophagy phosphatidylinositol 3 kinase/AKT/mammalian target of rapamycin (mTOR) pathway [130].

Key transcription factors
Key factors included TGFBI/BIGH3 (TGF-β-induced gene product-h3), bone morphogenetic proteins (BMPs), Nanog, and Oct4, which have important functions in cell adhesion, migration, proliferation, and apoptosis. TGFBI is a chondroprotective factor, released by MSCs and an anabolic regulator of cartilage homeostasis. Priming with TGFβ3 upregulated TGFBI transcription in murine MSCs and human MSCs and increased TGFBI secretion in human MSCs. Moreover, incubation of osteoarthritislike mouse chondrocytes with TGFβ3-primed murine MSC-conditioned media significantly upregulated the expression of chondrocyte anabolic markers but did not change the expression of catabolic and inflammatory factors [131]. BMPs are a subfamily of the TGF-β superfamily that participate in the induction of bone and cartilage formation. BMP6 enhanced chondrogenesis of MSCs [132]. As key transcription factors for pluripotency and self-renewal, the overexpression of Nanog and Oct4 also enhanced chondrogenic reported that cell therapy by using MSCs after neurogenic differentiation and maintained MSCs properties in various culture conditions [133]. Stromal cell-derived factor-1α has been detected to promote stem cell migration and homing [134].

The physical situation on MSC-based cells
The external physical situation could affect the cartilage and MSCs phenotype, such as conditions with hypoxia, hydrostatic pressure, compression, or magnetic fields. Nonetheless, the standard culture systems of the external physical situation have not been well-established yet and mechanisms are unclear.
Several studies have determined the application of low oxygen tension or hypoxia in MSC chondrogenesis and culture. In the presence of IL-1β, a significant in glycosaminoglycan, as a measure of proteoglycan levels, collagen, and water content, was observed under hypoxic condition (2% O 2 , 5% CO 2 , 93% N 2 ) [150]. Moreover, the mRNA expression of Col-II and aggrecan was upregulated in chondrocytes co-cultured with BM-MSCs under hypoxia (5% CO 2 and 95% N 2 ), and DNA methylation of the SOX9 promoter was significantly decreased under hypoxia [90]. Low oxygen tension (5%) was observed to promote ECM production by chondrocytes and enhanced the chondrogenesis of AT-MSCs compared to that cultured in normal condition [151]. Grayson and colleagues showed that under a 2% O 2 hypoxia, the expression of stromal cell genes Oct-4 and Rex-1 was upregulated [152].
Magnetic fields have been reported to enhance the chondrogenic differentiation of MSCs. Pulsed electromagnetic fields drastically promoted chondrogenesis by a specific hydrogel with high expressions of Col-II, aggrecan, and SOX9 genes [153]. Besides pulsed electromagnetic fields, a static magnetic field with 0.4 T was demonstrated to produce a strong chondrogenic differentiation response after 14 days of culturing through the TGF-β pathway [154]. Further, the presence of electromagnetic fields could partly replace the addition of TGF-β3, while the efficacy of chondrogenesis was statistically increased in the culture system. In TGF-β3-treated pellets, a further significant increase of 72.7% in aggrecan gene expression was induced by electromagnetic fields at 5 weeks [155].
Ultrasound-targeted microbubble destruction has been confirmed to increase the homing of transplanted MSCs to targeted organs. Stromal cell-derived factor-1α, as an important role in BM-MSCs migration, was loaded in microbubble. The number of migrated cells was higher when loaded microbubble under the guidance of ultrasound that the ultrasonic irradiation conditions of duty ratio 10%, intensity 1 W/cm 2 , time 30 s [134].

Dosage of MSCs
The single injection dosage ranged from 1.56 × 10 4 to 1 × 10 8 cells, and the most wildly proved dosage was 5 × 10 7 cells. Repeated injections or high dosage showed superiority than single injection or low dosage. Matas et al. clarified that repeated umbilical cord-derived MSCs (2 × 10 7 cells, every half year) treatment showed better improvements in pain and function than receiving injection only once at 1-year follow-up for individuals with KOA [15]. Moreover, Lamo-Espinosa et al. reported a high dosage with 1 × 10 8 BM-MSCs together with hyaluronic acid resulted in a larger clinical and functional improvement [160]. In addition, Chahal et al. found lower cartilage catabolic biomarkers and MRI synovitis in participants with higher doses [166] and the effects were maintained until 2-year follow-up [177].

Safety of MSC-based therapy
In general, MSC-based therapy is safe with mild adverse events. The most common adverse events were transient arthralgia, swelling of joints after local injection [171], and low back pain [158], which were mild to moderate and were usually spontaneously relieved within 7 days without special treatment or controlled with ibuprofen [159]. Song et al. reported one patient experienced mild edema and cramps of bilateral lower extremities that were relieved in 21 days without treatment [170]. Pers et al. demonstrated one severe adverse event that one patient with hypertension and hyperlipidemia experienced unstable angina pectoris without creased cardiac markers 3 months after AT-MSCs injection [171]. Adverse events were predominant in the high doses (> 5 × 10 7 cells) [163].

Effects on structure, pain, activities, and quality of living
Intra-articular injections of MSCs improved structure, pain, the function of the knee joint, rendering them a promising novel treatment for KOA. Besides these, this therapy had potential in activities, such as climbing the stairs and walking, and inflammatory factors. These benefits may last for several years, even be more apparent after months. Orozco et al. demonstrated an average 27% decrease of poor cartilage areas in severe KOA [159]. For MSCs-treated patients, Vega et al. [157] and Dilogo et al. [175] reported cartilage quality improvements that were quantified by T 2 relaxation measurements. Although Lee et al. thought there was no improvement of cartilage at 6 months in the MSCs group, whereas the defect in the control group was increased [19]. Radiological, arthroscopic, and histological measures consistently demonstrated decreased deterioration by regeneration of hyaline-like articular cartilage [173]. For pain and function, Soler et al. [158] revealed a relevant pain relief since day 8 and maintained after one year. Khalifeh et al. clarified improvement in range of motion of the knee joint after MSCs injection was significant between the 2-week and 24-week follow-up [176]. The walking time without pain improved [165]. Although knees deteriorate gradually, they were still better than at baseline at 5 years [178]. Pro-inflammatory monocytes/macrophages [166] and IL-10/12 levels [168] decreased in the synovial fluid after MSCs injection. Compared with ongoing conventional conservative management, AT-MSCs showed clinically significant pain and functional improvement after one year [18]. Moreover, greater improvements in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total score, WOMAC pain, and physical function were recorded than saline [164]. There were greater improvements in the pain, function, daily living activities, and sports and recreational activities subscales in people receiving MSCs therapy than PRP, although less than the combination of two therapies [167].

Future directions
Based on the current application states, different cell sources have been clarified in the clinic by phase I/ II studies, while no MSC-Exos are used. Therefore, a standard for MSCs therapy in KOA is required, which includes cell selection, authentication (phenotypic analysis and multipotent differentiation potential, especially distinguish with progenitor cells), culture or expansion methods, dosages, and rehabilitation program after injection. Second, Exos are good cargo and have potential in clinic application. The contained specific non-coding RNAs are important and may have an essential influence on the therapeutic effects. LncRNAs and circR-NAs in Exos and the safety and doses of Exos need large research, which will be the aim of future clinical trials. Third, previous researchers have focused on the influence of physical situations on MSCs, while the mechanisms are still unclear. In addition, no study reported the changes of MSCs-Exos, RNA, or DNA sequence after physical exposure, which might be an unexpected trend. Although we merely discuss biological materials in MSCbased therapy, scaffold-assisted grafts or complex 3D hybrid tissues of MSCs or Exos with or without electromagnetic fields are interesting trends for severe KOA. At last, besides cartilage regeneration, functional subchondral bone regeneration also has a significant impact on KOA treatment.

Conclusions
MSCs and MSC-exosomes, as new therapeutic methods for KOA, showed unique advantages. The selection of different origins of MSCs may be inconsistent based on the research goal and the phenotype exhibits various characteristics. MSCs and the derived exosomes carried out various functions in the treatment of KOA which include of increase of chondrogenesis, proliferation of chondrocyte, reduction of apoptosis, maintenance of autophagy, regulation synthesis and catabolism of the ECM, regulation of immune response, inhibition of inflammation, monitoring the mitochondrial dysfunction, and the paracrine effect. These functions were partly demonstrated through several biological pathways or axis, such as NF-κB, MAPK, ROS, and mTOR pathways. Exosomes are primary mediators of intercellular communications especially by transferring non-coding RNAs to adjacent cells or remote cells. Different physical conditions (hypoxia, magnetic fields, and ultrasound) have been studied to enhance the functions in MSC-based therapy experiments. This review has presented the evidence for MSC-based therapy as a new approach to the cell-free treatment of KOA. However, a standard for MSC-based therapy in KOA is required.