Articular cartilage is the connective tissue located on the surface of the synovial joint and plays a role in lubrication and weight-bearing during joint activities. Due to the lack of a vascular system and limited oxygen and nutrient supply, articular cartilage has low regenerative potential (Sophia Fox et al., 2009; Chen et al., 2017; Singh et al., 2019). In addition to the chondrocytes, which were long considered to be the only cell type within articular cartilage, cartilage stem/progenitor cells (CSPC) have been recently identified in OA cartilage (Williams et al., 2010), representing approximately 10% of the total cells (Alsalameh et al., 2004; Bosserhoff et al., 2014; Riegger et al., 2018). Moreover, single-cell RNA-seq analysis revealed various chondrocyte populations in advanced OA, including homeostatic chondrocytes, proliferative chondrocytes, effector chondrocytes, regulatory chondrocytes, pre-hypertrophic chondrocytes, hypertrophic chondrocytes, and fibrocartilage chondrocytes (Ji et al., 2019).

As the resident cell type in tissues, chondrocytes can form an extracellular matrix mainly composed of aggrecan and type II collagen. However, chondrocytes only account for 1–5% of the total volume of cartilage tissue (Bhosale and Richardson, 2008). Owing to the limited number of cells and their pyknotic nature, damage caused by various risk factors, such as abnormal mechanical load, trauma, and inflammation, can lead to changes in the structure and function of cartilage. Joint degenerative diseases are prone to occur without timely and adequate treatment (Palazzo et al., 2016; Vina and Kwoh, 2018; Li et al., 2019; Singh et al., 2019). Although the composition of articular cartilage is simple, its horizontal-layered structure containing chondrocytes with various morphologies and different distribution and secretion characteristics remains well-organized.

Changes in the metabolic state of chondrocytes lead to the imbalance of collagen synthesis and degradation, cartilage degeneration, chondrosenescence, and an intra-articular inflammatory environment, ultimately leading to OA (Martin and Buckwalter, 2002). The metabolic changes in chondrocytes may exhibit distinct characteristics and have been described in four clinical OA phenotypes. First, the inflammation-associated OA phenotype is characterized by a low degree of inflammation (Scanzello, 2017). The continuous accumulation of pro-inflammatory mediators leads to the excessive production of reactive oxygen species and mtDNA damage and drives the catabolic reaction in the chondrocytes, subsequently disrupting the balance between cartilage repair and damage. The second phenotype is the mechanical overload-associated OA phenotype. Normal physiological load is important to maintain chondrocyte function and ECM metabolic balance. Mechanical overloading is thus harmful to chondrocytes and results in a weakened anabolic response, decreased extracellular matrix (ECM) synthesis, and enhanced catabolic response, thereby stimulating the synthesis of matrix metalloproteinases (MMPs) (Loeser et al., 2012). Metabolic syndrome-associated OA phenotype is characterized by increased fasting plasma glucose concentration, increased triglyceride level, decreased high-density lipoprotein level, and/or hypertension (Dickson et al., 2019). The occurrence and development of OA can be positively affected by increasing the production of fat-derived pro-inflammatory mediators, such as advanced glycation end-products (Oren et al., 2011). The last is the aging-associated OA phenotype. The catabolism of aging chondrocytes is active, and disruption of the interplay between autophagy and the inflammasome is observed in an inflammatory environment (Salminen et al., 2012). In addition, the decline in chondrocyte mitochondrial function is accompanied by decreased chondrocyte autophagy and increased apoptosis (Blanco et al., 2018). The proliferation and synthesis ability of chondrocytes in aging articular cartilage also decreases; however, their ability to produce pro-inflammatory mediators and MMPs remains unchanged (Loeser, 2009). Overall, the initiation and progression of OA are closely related to the phenotypic changes in chondrocytes.

Increasing evidence has shown that the mutual communication between different tissues in the joint is essential for maintaining joint homeostasis. The communication between synovium and cartilage not only contributes to OA symptoms but is also a key factor in disease pathogenesis. The morphology and cell composition of the synovium are often used as biomarkers for the development of OA. Although the synovial tissue may not be affected in the early-stage of OA, many patients with advanced OA suffer from severe synovitis (Wenham and Conaghan, 2010).

In healthy joints, the synovium is mainly composed of two types of synovial cells. Type A macrophage-like synovial cells, which are relatively small, mainly have a phagocytic function and produce pro-inflammatory cytokines. Meanwhile, type B fibroblast-like synovial cells provide structure, nutrition, and lubrication and represent approximately 75% of the cells in the synovium (Firestein and McInnes, 2017). In addition, fibroblasts-like synovial cells can migrate to the site of tissue remodeling and interact with ECM molecules via specific surface receptors (Pap and Bertrand, 2012). They can perceive and respond to the changes in the composition and structure of the surrounding synovial tissue by adjusting their interaction with and the production of ECM components. In the initiation of OA, the intimal lining layer becomes hypertrophic and is infiltrated by macrophages, fibroblasts, mast cells, T cells, B cells, dendritic cells, and neutrophils, leading to a 5- to 10-fold increase in cell density (Culemann et al., 2019). These infiltrating cells promote the production of pro-inflammatory cytokines and catabolites, thereby changing the composition of synovial fluid, an important source of these pro-inflammatory mediators in OA (Falconer et al., 2018). In the synovial fluid of OA patients, the proportion of macrophages is relatively low, whereas that of mast cells is relatively high (de Lange-Brokaar et al., 2012; Robinson et al., 2016; Xie et al., 2019).

The degeneration and degradation of articular cartilage had long been considered the leading cause of OA, and many treatment strategies have been developed to protect the cartilage. Although the relationship between cartilage degeneration and subchondral bone destruction is close (Thielen et al., 2019), not all patients with OA exhibit abnormalities in the articular cartilage bone. In addition, in the aging OA phenotype, the imbalance in chondrocyte metabolism occurs before abnormal subchondral bone remodeling (Peffers et al., 2020). In contrast, in the trauma-induced OA phenotype, the early micro-injury in subchondral bone is detected first (Barton et al., 2017).

However, increasing evidence demonstrates that maintaining the integrity and remodeling balance of the articular subchondral bone can combat cartilage degeneration to restore homeostasis in joint tissues (Castañeda et al., 2012; Hoshi et al., 2017). Thus, exploring the mechanism underlying subchondral bone remodeling in OA can provide new insights for developing treatments for early-stage OA. The microstructural changes in articular subchondral bone in OA include the formation of subchondral bone cysts, bone marrow edema-like lesions, and osteophytes caused by early bone loss, late bone sclerosis, and histopathological changes (Li et al., 2013). These changes are caused by chondrocytes, osteoblasts, osteoclasts, endothelial cells, and the subchondral bone microenvironment (Henrotin et al., 2012).

Osteoblasts differentiate from mesenchymal cells and undergo four stages of maturation, namely, preosteoblasts, osteoblasts, bone-lining cells, and bone cells (Clarke, 2008; Katsimbri, 2017). The phenotype and activity of OA osteoblasts in subchondral bone are altered. For example, the levels of OCN, RANKL, (Kwan Tat et al., 2008) insulin-like growth factor 1, (Hilal et al., 1998) transforming growth factor β1, (Abed et al., 2017) vascular endothelial growth factor (Corrado et al., 2013), and alkaline phosphatase activity are elevated in OA, which subsequently lead to osteoclastogenesis, sclerosis (Wang et al., 2013) and angiogenesis. Osteoclasts are multinucleated cells derived from bone marrow myeloid progenitor cells and are mainly responsible for bone resorption and formation (Teitelbaum, 2000; Katsimbri, 2017). During osteoclast formation, progenitor cells are recruited to specific parts of the bone surface to differentiate into osteoclasts (monocytes) and fuse, thereby forming multinucleated mature osteoclasts. Mature osteoclasts adhere to the old bone area and release hydrogen ions and catalytic enzymes to dissolve the bone.

Mesenchymal stem cells (MSCs) are multipotent progenitor cells that originate from the mesoderm. MSCs by default can’t regenerate bona fide articular hyaline cartilage (it’s primarily fibrous cartilage and hypertrophy) (Buhrmann et al., 2010). However, MSCs can contribute to cartilage and bone repair, and their function in immune regulation and organ regeneration has been extensively studied (Levy et al., 2020). MSCs can be roughly divided into three types: embryonic stem cells, pluripotent stem cells, and adult stem cells (Vizoso et al., 2017).

Pluripotent stem cells are found in bone-related tissues, such as the bone marrow, synovium, infrapatellar fat pad, and adipose tissues (Pers et al., 2015). Bone marrow-derived MSCs (BMSCs) are the most well-characterized pluripotent stem cells. As early as 2002, stem cell therapy based on the in vitro expansion of autologous BMSCs has been used to treat OA. Although there was no significant difference in clinical results, improvement in symptoms was observed from arthroscopic and histological findings (Wakitani et al., 2002). In vitro, synovial MSCs (SMSCs) exhibit a particularly high capacity for cartilage differentiation (Shirasawa et al., 2006; Kurth et al., 2007). Studies have shown that SMSCs from OA patients can repair cartilage through allogenic tissue-engineered constructs in both in vitro and in vivo models. In experimental animal models, the injection of SMSCs into the joint cavity achieved a similar effect (Kondo et al., 2019; Enomoto et al., 2020; Dragoo et al., 2012). Infrapatellar fat pad (IPFP), a column of fat tissue located behind the patella, and synovium are involved in the occurrence and development of intra-articular diseases, such as OA (Favero et al., 2017). As the MSCs/stromal cells derived from IPFP (Kouroupis et al., 2019) are similar to SMSCs, (To et al., 2019) IPFP MSCs are speculated to have the ability for tissue repair (Galipeau et al., 2016), indicating IPFP as a potential target for joint diseases (Attur et al., 2010). Its potential in chondrogenic, osteogenic, and adipogenic lineages has been reported (Sun et al., 2018). In vitro, the chondrogenic differentiation ability of IPFP MSCs was greater than that of BMSCs, adipose tissue MSC (AMSCs), and UC-MSCs (Ding et al., 2015). AMSCs can promote cartilage regeneration and regulate inflammation. Because they are versatile and readily available, AMSCs are an excellent source of cells for OA treatment (Kim and Koh, 2018; Lee et al., 2019). However, the mechanism by which AMSCs induce cartilage regeneration remains unclear. Current evidence indicates that AMSCs regulate the local microenvironment through paracrine nutritional factors, thereby making it more favorable for regeneration and repair and subsequently delaying cartilage degradation and improving joint function (Damia et al., 2018). Previously, a small number of MSC-like progenitor cells-chondrogenic stem cells/progenitor cells (CSPCs) were detected in cartilage tissues (Alsalameh et al., 2004; Fickert et al., 2004; Koelling et al., 2009). Because they share similar properties with BMSCs, CSPCs are speculated to be involved in cartilage regeneration. CSPC migration occurs upon cartilage damage, and their proliferation and immune regulation capabilities are enhanced (Seol et al., 2012; Riegger et al., 2018).

Induced pluripotent stem cells (iPSCs) can provide an unlimited cell source for tissue engineering and are an attractive substitute for primary cells. iPSCs are characterized by a high degree of plasticity and promising differentiation potential and have promising potential in cell therapy. Patient-specific iPSCs can be engineered to minimize the autoimmune response, making them an almost ideal cell source for cell-based therapy. Studies on cartilage tissue engineering of iPSCs have demonstrated their utility for functional cartilage repair and as models for studying cartilage pathology (Diekman et al., 2012; Willard et al., 2014). iPSCs provide a platform for identifying candidate, patient-specific OA therapeutic agents (Lietman, 2016). For example, iPSCs reprogrammed from somatic cells (Takahashi et al., 2007; Yu et al., 2007) could generate endless OA patient-specific stem cells for drug research. Moreover, iPSCs isolated from the tissues of OA patients could differentiate into the cartilage, which provides opportunities for cartilage tissue research (Wei et al., 2012; Lach et al., 2014; Nam et al., 2018). However, there are no clinical studies published about cartilage cell therapy using iPSCs. iPSCs not only have excellent proliferation and differentiation capabilities similar to other stem cells but also do not cause immune rejection and ethical issues (Moradi et al., 2019). Therefore, more studies are required to improve the future applications of iPSC-derived chondrocytes in OA replacement therapy.

Extracellular vesicles (EVs) are nano-sized communication messengers secreted by cells that transmit biological signals between cells and are mainly divided into three categories, namely, exosomes, microvesicles, and apoptotic bodies. EVs have received increasing research attention in the field of regenerative medicine. Increasing studies have demonstrated potential value for alleviating inflammation and promoting tissue repair and regeneration (Bjørge et al., 2017).

Extracellular vesicles are small double lipid membrane vesicles with a diameter of approximately 30–2000 nm that can carry various biologically active molecules, such as RNA subtypes (mRNA, microRNA, and lncRNA), DNA fragments, lipids, proteins, and enzymes (Lamichhane et al., 2015). Most cells, such as those in connective tissues and MSCs, can produce and secrete EVs into various biological fluids, such as the blood and synovial fluid (Rani et al., 2015). Once released, EVs can take effect immediately or be transported to a distant place. EVs communicate with the recipient cell by producing the same effect as their donor cell.

Exosomes are small vesicles with a diameter of approximately 30–150 nm derived from the endosomal compartment. They are the most widely studied type of EVs (Théry et al., 2002). Endosomal membrane invaginates to form multivesicular bodies containing intraluminal vesicles. When multivesicular bodies fuse with the plasma membrane, intraluminal vesicles are secreted as exosomes into the extracellular space. Microvesicles are slightly larger than exosomes, with a diameter ranging between 50 and 1000 nm (De Jong et al., 2014). Because microvesicles are directly shed from the plasma membrane, their markers on the membrane surface are also the same as the donor cells. Similar to exosomes, microvesicles can transport biologically active molecules to recipient cells. Apoptotic bodies are the largest type of EVs, with a diameter of ≥1000 nm, and are formed in the late-stage of apoptosis (De Jong et al., 2014). Apoptotic bodies may function as biomarkers. However, at present, the connection between regenerative medicine and cells remains unclear.

Extracellular vesicles serve as mediators of intercellular communication (Raposo and Stoorvogel, 2013; De Jong et al., 2014). EVs interact with the surface receptors of recipient cells through their transmembrane proteins to activate downstream intracellular signaling pathways. EVs can also be directly endocytosed by recipient cells to release their contents (Jaiswal et al., 2014). The type of donor cells and the environment in which they are located, such as being in a state of stress, can affect EVs’ function and contents (de Jong et al., 2012). The biological functions of EVs and their biogenesis require more studies. Increasing studies in regenerative medicine have focused on the production of protective and pro-regenerative EVs, particularly in cartilage repair.

Studies have revealed the role of EVs in OA pathogenesis, inflammation, and cartilage regeneration and have demonstrated their potential implications for joint disease therapy (Zhang S. et al., 2018). EVs derived from different types of joint cells participate in maintaining joint homeostasis and can initiate and promote the progression of OA (Murphy et al., 2018). Macrophages and leukocytes infiltrating the synovium could interact with fibroblast-like synovial cells through EVs (Malda et al., 2016). The activated fibroblast-like synovial cells then transmit inflammatory signals, such as cytokines and enzymes, to the macrophages and leukocytes, thus forming a feedback loop that further aggravates OA. These EVs could also cause the degradation of the extracellular matrix and result in changes in the subchondral bone. In an in vitro OA model, EVs enhanced cartilage anabolism and relieved inflammation (Wang et al., 2017), thereby delaying cartilage degradation and the progression of OA (Zhu et al., 2017). In addition, EVs protected chondrocytes (Headland et al., 2015) and regulated the physiological activities of various types of immune cells (Lo Sicco et al., 2017), indicating their anti-inflammatory effects. The induction of a regenerative immune phenotype and enhanced metabolic level of chondrocytes promote the formation of type II collagen-rich cartilage that can repair cartilage defects in rat and rabbit OA models (Zhang et al., 2016; Zhang S. et al., 2018). Hydrogel encapsulation delivers EVs at more accurate positions and higher doses, thereby significantly enhancing repair ability (Liu et al., 2017).

Recently, the application of stem cell-derived EVs in the treatment of joint damage and OA has received increasing attention. They have been derived from the bone marrow (Cosenza et al., 2018; Vonk et al., 2018), adipose tissue (Lo Sicco et al., 2017), synovium (Zhu et al., 2017), or pluripotent cells, including embryonic stem cells (Zhang et al., 2016; Zhang S. et al., 2018) and iPSCs (Liu et al., 2017; Zhu et al., 2017). Exosomes are the main research target in EVs. The difference between exosomes and microvesicles derived from the same cell has been investigated (Cosenza et al., 2017, 2018). Moreover, EVs containing both exosomes and microvesicles have been analyzed (Headland et al., 2015; Lo Sicco et al., 2017). In addition to sharing similar biological functions with stem cells, stem cell-derived EVs offer significant advantages due to their small size and low immunogenicity. Issues associated with direct cell injection can also be avoided. EVs do not pose the risk of antigen presentation due to differentiation into specific cell types such as MSCs, which allows them to be used in allogeneic therapy. In addition, the biologically active cargoes inside MSC-derived EVs are more stable, and problems such as senescence following expansion or cartilage calcification following induction are eliminated.

Exosomes maintain homeostasis (Gao et al., 2018) and facilitate cell-to-cell communication in diseases (Li Z. et al., 2018). The source and content of exosomes determine their functions and biological characteristics. Exosomes secreted by therapeutic cells help treat diseases, while those released by cells in the pathological microenvironment accelerate the disease process. In this section, we discuss the various kinds of exosomes derived from joint tissues and their biological effects in OA.

Recent attempts to differentiate iPSCs derived from OA patients into chondrocytes have been conducted. Generally, the addition of growth factors, such as TGF-β, FGF-2, BMP, and WNT3A, and paracrine factors, such as Ihh and Runx, to the culture medium is necessary to drive iPSCs to the chondrogenic lineage (Yamashita et al., 2018; Zhao Y. et al., 2020). Currently, four main methods are available: (1) transformation of iPSCs into MSC-like cells and their differentiation into chondrocytes (Nejadnik et al., 2015); (2) co-culture of MSCs derived from iPSCs with primary chondrocytes (Qu et al., 2013); (3) formation of embryoid bodies (EBs) (Nakagawa et al., 2009); (4) cultivation of iPSCs in a medium that mimics the physiological environment during development (Diekman et al., 2012). Studies aiming to form iPSCs through EB and co-culture them with chondrocytes have been conducted (Wei et al., 2012). First, the chondrocytes of OA patients were reprogrammed into OA-iPSCs by lentivirus induction. After the formation of EB, they were continually cultured in the chondrogenic medium for 14 days. Subsequently, the iPSCs were transfected with a lentivirus carrying TGF-1β and inoculated on alginate matrix-coated dishes. After culturing for another 14 days, TGF-1β/iPSCs were subcutaneously injected into the back of mice. Ectopic cartilage tissue was observed at 6 weeks after transplantation.

Engineered cartilage tissue from chondrocytes, when transiently transfected circuits activate the PTGS2 gene, immunomodulatory IL-4 is produced, thereby representing a new immunomodulatory method (Nims et al., 2021). Autologous articular chondrocytes are an established cell-based tissue engineering strategy for treating knee cartilage or osteochondral defects (Brittberg et al., 1994). Most of the current scaffolds or biomaterials contain MSCs that can undergo chondrogenic differentiation and are used as clinically relevant chondrogenic implants to repair cartilage defects. However, after their implantation in the body, differentiated chondrocytes showed a hypertrophic phenotype (collagen X, MMP13) and induced ectopic bone formation (Pelttari et al., 2006). Therefore, the production of articular chondrocytes with stable, extracellular matrix and phenotype is the main goal of in vitro cartilage tissue engineering.

Macrophages are among the main types of cells that affect joint homeostasis and have been applied for the development of related cell engineering technologies. In addition to using biomaterial scaffolds to regulate macrophage-induced inflammation, macrophages themselves have been utilized for drug delivery or treatment. The intrinsic homing ability of macrophages allows their migration to the site of inflammation or injury in OA. Using this feature, autologous M1 macrophages are used to deliver nanoparticle-encapsulated drugs to induce transient phagosome maturation arrest (Visser et al., 2019). The regular use of clustered interspaced short palindromic repeats (CRISPR)-Cas9 genome editing to create a cell-autonomous system (”SMART” cell), which is derived from mouse induced pluripotent stem cells (miPSCs), has also been attempted. Chondrocytes can automatically regulate inflammation in both in vivo and in vitro OA models. When “SMART” chondrocytes receive specific targeting signals (such as inflammatory cytokines IL-1 or TNF-α), they released corresponding biological drugs, such as IL-1Ra and or soluble TNFR1, to relieve inflammation (Brunger et al., 2017; Pferdehirt et al., 2019). In the same way, the design of self-regulating “SMART” macrophages enables these cells to not only automatically home to the inflammation/injury site but also to have cytokine-activated feedback-controlled capabilities. Thus, targeted drugs can be effectively delivered to treat joint diseases (Adkar et al., 2017). Overall, cell tissue engineering is a powerful tool to develop new modalities for OA treatment.