Collagen type X expression and chondrocyte hypertrophic differentiation during OA and OS development

Abstract

Chondrocyte hypertrophy and the expression of its specific marker, the collagen type X gene (COL10A1), constitute key terminal differentiation stages during endochondral ossification in long bone development. Mutations in the COL10A1 gene are known to cause schmid type metaphyseal chondrodysplasia (SMCD) and spondyloepiphyseal dyschondrodysplasia (SMD). Moreover, abnormal COL10A1 expression and aberrant chondrocyte hypertrophy are strongly correlated with skeletal diseases, notably osteoarthritis (OA) and osteosarcoma (OS). Throughout the progression of OA, articular chondrocytes undergo substantial changes in gene expression and phenotype, including a transition to a hypertrophic-like state characterized by the expression of collagen type X, matrix metalloproteinase-13, and alkaline phosphatase. This state is similar to the process of endochondral ossification during cartilage development. OS, the most common pediatric bone cancer, exhibits characteristics of abnormal bone formation alongside the presence of tumor tissue containing cartilaginous components. This observation suggests a potential role for chondrogenesis in the development of OS. A deeper understanding of the shifts in collagen X expression and chondrocyte hypertrophy phenotypes in OA or OS may offer novel insights into their pathogenesis, thereby paving the way for potential therapeutic interventions. This review systematically summarizes the findings from multiple OA models (e.g., transgenic, surgically-induced, mechanically-loaded, and chemically-induced OA models), with a particular focus on their chondrogenic and/or hypertrophic phenotypes and possible signaling pathways. The OS phenotypes and pathogenesis in relation to chondrogenesis, collagen X expression, chondrocyte (hypertrophic) differentiation, and their regulatory mechanisms were also discussed. Together, this review provides novel insights into OA and OS therapeutics, possibly by intervening the process of abnormal endochondral-like pathway with altered collagen type X expression.

Introduction

Osteoarthritis (OA), the most prevalent degenerative joint disease among the elderly, is responsible for mobility impairment and pain [1]. The incidence of OA keeps increasing year by year simply because of the aging population and the sharply increased number of obese people. OA progresses slowly and is often characterized by degenerative changes in the articular cartilage, including cartilage loss, osteophyte formation, subchondral osteosclerosis, and synovial inflammation [2]. Due to the limited self-healing ability of articular cartilage, untreated articular cartilage defects are further deformed, leading to joint pain and dysfunction, and eventually typical OA [3]. The pathogenesis of OA is complex, involving aging, mechanical, inflammation and metabolic factors, which eventually lead to the destruction and failure of the entire joint tissue structure. Its incidence continues to rise annually due to the increasing prevalence of obesity and an aging population. Obesity induces joint overloading, altering joint biomechanics, often leading to pronounced wear and tear, particularly in the medial cartilage [4]. Additionally, postmenopausal women experience a substantial decrease in estrogen levels, disrupting the balance between subchondral bone formation and resorption, thereby increasing susceptibility to joint trauma and elevating the risk of OA [5]. Normally, chondrocytes remain quiescent but actively synthesize an array of extracellular matrix (ECM) proteins, including collagen, hyaluronic acid, glycoproteins, and proteoglycans, to support and cushion stresses generated by movement. Conversely, metabolic abnormalities in chondrocytes and the ECM in articular cartilage injury predominantly manifest increased catabolism and decreased synthesis [6]. The ECM synthesized and secreted by chondrocytes is to support and cushion the stress generated during exercise. Progression of OA involves abnormal secretion of matrix degradation factors by chondrocytes, such as MMP13, MMP1, and ADAMTS5, resulting in ECM degradation and a marked reduction in type II collagen and proteoglycans, ultimately leading to defects in cartilage tissue [7].

In addition to the enhanced production of matrix-degrading enzymes, a prominent feature of OA is the presence of a hypertrophic chondrocyte phenotype. Chondrocyte hypertrophy constitutes a transient physiological process essential for bone formation within cartilage. In the context of OA, chondrocytes react to the accumulation of injurious biochemical and biomechanical stimuli by transitioning into a hypertrophic state [8]. Articular cartilage is a layer of hyaline cartilage, mainly composed of ECM and resting chondrocytes, including collagen (Collagen II, VI, IX), hyaluronic acid and proteoglycan [9]. With the progression of OA, articular chondrocytes are activated, and the expression of articular cartilage related markers such as proteoglycan, type II collagen, are significantly down-regulated [10], while hypertrophic markers, including type X collagen, MMP13, and ADAMTS5, may become highly expressed in chondrocytes undergoing hypertrophic changes [11]. The abnormal hypertrophy of articular chondrocytes results in cartilage matrix degradation, vascular invasion, and osteoid formation, mirroring the terminal differentiation of chondrocytes observed during endochondral bone formation [12]. This process is not limited to human patients, but is also observed in most animal models of OA [13,14]. Indeed, numerous studies concerning OA have focused on hypertrophic chondrocyte-specific marker genes and signaling factors, such as Col10a1, Mmp13, and Runx2 [15,16]. However, a direct causal relationship between the development of a hypertrophic chondrocyte phenotype and the progression of OA remains to be firmly established. Whether the occurrence of OA is triggered by chondrocytes and the underlying mechanism of abnormal hypertrophic differentiation of articular chondrocytes in the development of OA need further study.

Osteosarcoma (OS) is the most common malignant bone tumor and often affects children and adolescents [17]. The exact pathophysiology of OS is unknown, but the osteoid and immature bone in OS might be produced by malignant mesenchymal cells [18]. It is worth noting that cells of OS are often derived from osteoblasts, including pre-osteoblasts and osteoblasts. Chondrocytes and fibroblasts are also present in the tumor stroma of many osteosarcomas. In addition, in the induced mouse OS model, cartilage matrix was found present in the tumor tissue and ossification occurred through similar process of endochondral ossification [19,20]. As a transcription factor, RUNX2 is known to play a crucial role in osteoblast differentiation and chondrocyte hypertrophy or maturation. Studies have shown that activation of Wnt/β-catenin in OS upregulates RUNX2 expression, and thus promoting the expression of metastasis-related genes [21]. p53 gene is involved in the normal development and physiological mechanism of bone, and is considered an important factor in tumorigenesis. Deletion of p53 and Rb-1 leads to early OS [22]. Moreover, an interaction relationship between P53/RB-1 and RUNX2 has been proven [23]. We also notice that there are transcription factors, or regulators specifically secreted by chondrocytes are known to play a significant role in OS apoptosis and tumor drug resistance or sensitivity [24]. These findings together with the chondroid tissue found in osteochondrosarcoma may indicate a chondrogenic element during OS development.

Here we systematically reviewed the altered collagen X expression and chondrocyte (hypertrophic) differentiation in multiple OA models, as well as the potential chondrogenic element in OS pathogenesis, which may help with future OA and OS therapeutics.

Animal model of osteoarthritis

OA is the most common chronic joint disease and a leading cause of mobility impairment and pain in the elderly [25,26]. When OA occurs, it always affects the joints under stress, usually the knees, hands, and spine [27]. OA is characterized by joint destruction, cartilage loss, and osteosclerosis and osteophyte formation (Figure 1) [28,29]. In order to further understand the pathogenesis of osteoarthritis as well as newly developed treatment options, multiple OA animal models have been established, as OA is a heterogeneous disease, a single animal model won’t reflect all aspects of disease [30]. Studies have also found that the underlying mechanism may be different in animal models of OA induced by different approaches [31]. Here, we reviewed following transgenic, surgical induction, chemically and mechanical load induced OA models, aiming to identify the different etiologies on the occurrence and development of OA.

Figure 1.

Healthy joint was compared with OA joint. Clinical data show that majority of OA patients have a variety of pathologic phenotypes, including vascular invasion, osteophytes formation, synovial inflammation, articular cartilage erosion, and subchondral bone sclerosis. Left side shows the normal joint structure. Right side shows the possible joint structural symptoms and changes of osteoarthritis.

Transgenic or genetically modified OA models

The application of transgenic mice has greatly improved our understanding of the physiological and molecular mechanism of many diseases [32]. In the past few decades, the number of transgenic or genetically modified OA models has been increasing and has gradually become the best candidate model to explore the molecular mechanism of OA (Table 1).

Table 1.

List of transgenic or genetically modified mouse models

Gene	Mice model	Changes of mice bone	Molecular mechanisms
Type II collagen	Del1(+/-) [34]	Cartilage erosion	↓Col II
		Osteosclerosis
		Meniscus degeneration
		Various joint structures are mineralized
	Disproportionate micromelia (Dmm/+) [36,37]	Abnormal condylay cartilage in TMJs	↑HTRA1, MMP13, DDR2 in tempormandibular joint (TMJs)
		Proteoglycan deficiency	↓Col II
		TMJs tremor at 6 months
		Cartilage erosion
Runx1	Runx1F/FCol2a1-CreERT [38]	Accelerated cartilage destruction	↑MMP13
		Joint deformity	↓COL2A1, SOX9, COL10A1
		Proteoglycan deficiency
Runx2	Col10a1-Runx2 [41]	Chondrocyte maturation is delayed	↑SOX9, COL10A1, RUNX2
		Shortened femur length
		Cortical bone thinning
		Reduced joint destruction in induced OA
Sox9	Col10a1-Sox9 [43]	Chondrocyte columns disorder	↑SOX9, MMP13, ADAMTS4/5,
		Reduced hypertrophic cartilage mineralization	↓COL10A1
		Hypertrophic chondrocytes increased adipogenesis
		Subchondral osteosclerosis
		Osteophyte formation

In heterozygous Del1 mice, harboring a short deletion mutation in type II collagen gene lead to early-onset degenerative changes in the knee joint and progress to end-stage osteoarthritis at 12-15 months [33,34]. Histological examination of Del1 heterozygous mice showed earlier and more severe cartilage erosion. Subchondral osteosclerosis and osteophyte formation occurred in Del1 heterozygous mice with age [34]. At the same time, studies have shown that lifelong running significantly increases the incidence and severity of OA in mice with Col2a1 mutations [35]. Early-onset OA and up-regulation of related markers (Ddr2, Mmp13, HtrA1) in the temporomandibular joint were found in mice with Col2a1 mutations [36,37]. A recent study also showed that markers of chondrocyte hypertrophy were significantly upregulated in mice with Col2a1 mutations, such as Col10a1, Runx2, and accelerated the progression of OA [8]. Mice that knocked out Runx1 with Col2-Cre showed impaired cartilage formation and reduced bone density, and overexpression of Runx1 prevented cartilage destruction in OA [38]. Runx2 and Col10a1 are molecular markers of chondrocyte hypertrophy [39,40]. In view of the important role of abnormal hypertrophy of chondrocytes in OA, transgenic animal models of hypertrophic chondrocyte related genes have also been extensively studied (Figure 2). Chondrocyte maturation was delayed in Col10a1-Runx2 transgenic mice overexpressing Runx2 and was able to protect the mouse joints against OA [41]. SOX9 is an essential regulator of cartilage development [42]. In a new Col10a1-Sox9 transgenic mouse, OA was shown at six months of age. Histological findings showed that growth plate structure of the mouse was abnormal and the mineral content was significantly reduced [43]. A recent study induced OA lesions by using AggrecanCreERT2 transgenic mice to delete the Fermt2 gene encoding the key focal adhesion protein Kindlin-2 in chondrocytes of articular cartilage, highly mimicking human OA pathological changes [44]. In that study, the articular chondrocytes exhibited hypertrophic differentiation, with significant increases in the protein expression of the hypertrophy markers COL10A1, MMP13 and RUNX2. The Col11a1+/- mice developed several features of the typical OA knee at 15 months, including loss of articular cartilage and meniscal malformation, and a significant increase in MMP13 was also observed [45]. Although the transgenic mouse model cannot simulate the corresponding function of human joints, it is the main choice to study the molecular mechanism of OA. The genes that are increased or decreased in the transgenic model can be considered as potential targets for the treatment of OA and may provide an important pathway for the development of new therapeutic options.

Figure 2.

Regulatory mechanisms in hypertrophic chondrocytes of OA. Activation of Wnt/β-Catenin signaling pathway in OA induces Runx2 gene expression [144,145]. Activation of the inflammatory NF-κB pathway, upregulation of mTORC1 and hypoxia induced expression of HIFs protein in articular cartilage chondrocytes, and subsequently HIF-2α promoted cartilage destruction [146-148]. TGFβ/BMP regulates the function of Runx2 through Smad pathway, affecting chondrocyte differentiation and possible development of OA [149].

Surgical induction OA models

Surgical induction model is the most common OA model in experimental animals (Table 2). The surgical instability of medial meniscus (DMM) is the most widely used OA model. After DMM, the medial meniscus was displaced, but the cruciate ligament remained intact, there was no ectopic chondrogenesis, and OA progression gradually progressed from mild to moderate [46]. Significant synovial hyperplasia and osteophyte formation were both observed at 6 and 12 weeks after DMM surgery [47]. At the same time, the results of immunohistochemical staining showed that the protein expression of COL10A1 and MMP13 were significantly up-regulated in the mice undergoing DMM surgery [47]. The results of a recent study also showed significantly increased COL10A1 expression in articular cartilage and subchondral bone in DMM-induced animal models [48]. Platelet-rich plasma (PRP) protects chondrocytes against destabilization of the DMM-induced OA and suppresses the dysregulation of chondrocyte matrix-related factors (Sox9, Col10a1, Col2a1) [49]. In one study, the protective effect of the EHFR pathway on OA was investigated by performing DMM surgery in Col2-Cre HBEGF overexpressed mice [50]. The DMM model has been used in many studies of OA disease progression and treatment options [51-53]. Anterior cruciate ligament rupture can lead to biomechanical changes and joint instability, eventually leading to OA [54,55]. Anterior cruciate ligament transection (ACLT) is a surgical method alike to DMM. Severe OA symptoms were observed within a short time in a mouse model after ACLT surgery, with destruction of subchondral bone and growth plate, and ectopic bone formation at 8 weeks [46]. 30 days after ACLT was performed in rat and was also enough to induce the typical symptoms of knee OA, such as pain, dysfunction, synovial inflammation [56]. The model could also be used in follow-up studies within a short time after induction. Interestingly, one study used an ACLT rat model to investigate the correlation between subchondral bone and early cartilage changes in OA, and the early progression of OA could be clearly appreciated [57]. And histological evaluation of the ACLT model found that the surgery group showed more hypertrophic differentiation of chondrocytes [57]. In addition to mice and rats, rabbits and pigs are also widely used in vivo preclinical studies [58,59]. ACLT surgery in rabbit is more suitable for macroscopic evaluation because of its large joint compared to mice [58]. Histopathological analysis of rabbit ACLT at 12 weeks postoperatively revealed decreased staining of Alican blue and type II collagen on the surface of femur and tibia cartilage, suggesting loss of cartilage matrix [58]. In addition, meniscus injury after ACLT surgical induction will further accelerate the development of OA [60]. In a study of Micro-RNA, the effect of miR-26b-5p on chondrocyte hypertrophy was validated using an ACLT-induced animal model. The results show that miR-26b-5p can protect the articular cartilage of OA mice and reduce its synovial inflammation by targeting Col10a1 [61]. The loss of estrogen will reduce the bone mass of subchondral bone and promote the progression of OA [62]. Estrogen physiologically decreased in postmenopausal women, and ovariectomy (OVX) rats have been shown to be useful as an in vivo study model for OA in postmenopausal women [63]. Researchers have induced experimental OA in rats by performing a meniscectomy (MSX). The role of chondromodulin-1 (ChM-1) in chondrocyte maturation and OA progression through the HIF-2α pathway was explored using this procedure-induced OA model [64]. It further showed that chondrocytes are abnormally hypertrophic and highly expressed COL10A1, MMP13 and VEGFA in surgically induced OA, while this study showed that ChM-1 can inhibit hypertrophic differentiation of chondrocytes and slow OA progression [64]. Surgical induction of OA can be used to study the molecular mechanism of OA and provide a new therapeutic mechanism for OA [65-68]. The surgically induced OA model is also preferred in many preclinical studies of new treatment option [51,69,70].

Table 2.

Surgical induction models of OA

Surgical induction models	Authors	Changes of the model	Molecular mechanisms
Destabilization of the medial meniscus (DMM)	Li et al. [47]	Medial meniscus displacement	↑COL10A1, MMP13, ADAMTS 5, AMPKα1
		Osteophyte formation and synovial hyperplasia occurred at week 6
	Fang et al. [150]	Gait change
		Proteoglycan loss, chondrocyte aggregation in medial articular ventricle and osteophytes formation at week 2
		Mild to moderate cartilage damage at week 5
		Intrachondral ossification in osteophytes at week 10
Anterior cruciate ligament transection (ACLT)	Barbosa et al. [56]	Paw print area reduced	↑IL-1β, IL-17, TNF-α
		Synovial inflammation
		Joint swelling
	Aizah et al. [57]	Proteoglycan loss
		Subchondral trabecular bone thickness decreased
	Go et al. [58]	Femoral cartilage surface erosion	↓Col II
		Osteophyte formation
	Jia et al. [59]	Cartilage lesions and loss	↑SDF-1, MMP-13
		Chondrocyte distribution irregular	↓Col II, ACAN
Ovariectomized (OVX)	Andersen et al. [63]	Gain weight	↑CTX-I, CTX-II
		Cartilage erosion	↓Col II
		Cartilage turnover increased

Mechanical load models

Appropriate stimulation may cause responsive stress of the body, leading to chondrocyte proliferation and synthesis of extracellular matrix. OA can develop when the intensity of stimuli increases beyond the body’s ability to self-regulate and damage occurs [71,72]. Therefore, such animal models with mechanic force change play an important role in studying the interaction between mechanical load and the occurrence and progression of OA (Table 3). In humans, overweight may promote the progression of OA people with OA are at increased risk of incident disability or disability progression [73,74]. Many studies have shown that proper exercise is beneficial for cartilage development and improves joint symptoms, but excessive stress can destroy cartilage [74,75]. At the same time, studies have observed that changes in the medial load position of the knee joint in early OA patients promote the development of OA [76]. Mechanical load is one of the factors in the occurrence of OA, so various animal load models are essential to study the association of machinery with various joint health and diseases [77,78]. The first load test was carried out in rabbits. The effect of the load test on the knee joint of rabbits was investigated by simulating the jumping force and repeating the impulse load [79]. Blandine Poulet et al. characterized a new noninvasive mouse joint load model, showing that increased load can lead to osteophyte formation, meniscus ossification, and cruciate ligament injury, the severity of which depends on the load regimen [80]. When an appropriate degree of load was applied, subchondral bone thickened, epiphyseal trabecular thickness and mass increased, and gait changed [81], cells in the meniscus show loss of fibrocyte-like shape, like hypertrophic chondrocyte changes. In the study of temporomandibular cartilage in response to mechanical loading, the results found that increased loading caused increased mineralized cartilage areas and increased expression of COL10A1 [82]. Cartilage thinning and osteopontin production in a periodic joint load rabbit model [83]. The model can be used to study the effect of cyclic load on joint. In the study of OA growth plate dynamics, increased expression of COL10A1 and MMP13, markers of chondrocyte hypertrophy, were observed in articular cartilage and tibial growth plate in a mouse load model [11]. One study has shown that mechanical joint loading can be used as a suitable model to study mechanical OA pain, effectively avoiding the risk of postoperative pain and infection that may result from surgical intervention [84]. Interestingly, obesity can increase mechanical load-induced cartilage damage, and animal load models are critical to study the association between obesity and metabolic syndrome with OA [85,86]. The combination of diet and exercise will significantly reduce the load on the hips and knees, and improve the clinical and biomechanical outcomes of obese OA patients [87,88]. The mice received running training from 2 to 18 months of age, and the results showed that moderate and sustained running training accelerated the development of OA in the mice [89]. In previous studies, repeated intra-articular injection of TGFβ-1 and treadmill overrunning into Col10a1-Runx2 transgenic mice to induce knee OA caused joint damage [41]. Mechanical loading models are also widely used in the study of mechanical-related molecular mechanisms. One study suggests that interleukin-6 (IL-6), a mechanosensitive cytokine, may play a key role in the biomechanical regulation of OA bone remodeling in mechanical load models [90]. Mechanosensitive PIEZO1 channels can drive mechanical damage in chondrocytes and may provide pathways for the prevention and treatment of OA [91]. In short, mechanical loading models are crucial for biomechanical study of OA, the correlation study of obesity and metabolism, and the study of molecular mechanisms.

Table 3.

Mechanical load models of OA

Mechanical load model	Authors	Load application frequency	Changes of the model
Stimulated jump	Radin et al. [79]	Be subjected to 1/2 the animal’s body weight 40 times a minute for 20-40 minutes per day	Subchondral osteosclerosis
			Cartilage injury
Applied knee load	Poulet et al. [80]	Three times a week for two weeks or five weeks	Osteophyte formation
			Meniscus ossification
			Synovial hyperplasia and fibrosis
Right knee weight	Poulet et al. [81]	Once a week;	Subchondral bone thickness of the femur and tibia increased
		Three times a week for two weeks; Three times a week for five weeks	Increased trabecula of epiphysis
	Heegde et al. [84]	9N/11N load, three times a week for two weeks	Mechanical sensitivity of paws decreased
			Articular cartilage injury
			Load a week synovial lining thickens
Repeated joint bucking load	King et al. [83]	Repeat bending of the front paw toe at 1 Hz for 2 hours per day for 60 hours	Thickness of calcified cartilage reduced
			Production of osteopontin increased
Lifelong moderate running training	Lapveteläinen et al. [89]	Speed 13.3 meters/min, distance 1000 meters/day	The incidence of medial and lateral tibial condyle OA increased

Chemical induction models

Monosodium iodoacetate (MIA) is a metabolic inhibitor that disrupt glycolysis by inhibiting the activity of glycolysis-3-phosphate dehydrogenase in chondrocytes and ultimately leads to chondrocyte death [92,93]. Intra articular injection of MIA is often used to induce OA models. A recent study showed that MIA injection in the joint of rats caused immediate inflammation, resulting in histological destruction of functional knee units and ultimately OA pain [94]. The MIA-induced OA model has also been applied to the treatment of OA [95,96]. Curcumin nanoparticles can improve the progression of MIA-induced OA in rats, and the content of MMP13 in articular cartilage is significantly increased in OA models [97]. p53 is a key factor in OA pain. Sirtuin 1 (SIRT1) mediates MIA-induces OA pain by upregulating p53 expression, which makes targeting SIRT1/P53 pathway a possible new therapeutic option [98]. In a study of early biomarkers of OA induced OA in donkeys using MIA, significant up-regulation of miR-146b, miR-27b and Col10a1 expression in serum and synovial fluid was observed in early OA in this animal model [99]. It can be seen from the above studies that the MIA-induced OA model is often used in the pain research of OA, and it can be selected in the pain treatment. Papain is a proteolytic enzyme that can degrade proteoglycans and cause cartilage damage [100]. Different concentrations of papain induced knee OA in rabbits shown different severity of symptoms [101]. To elucidate the effect of many developed tissue engineering methods and novel biomaterials on osteochondral defects in OA in an inflammatory environment, papain was used to construct a model of persistent cartilage defects in rabbits with OA [102]. In this model, compared with non-OA group, COL I and COL II in the regenerated cartilage of OA group were significantly reduced, while MMP13 protein expression was increased [102]. At the same time, some studies compared the pig OA model induced by surgery and papain, and found that joint injection of papain can produce more and faster cartilage erosion, synovial inflammation, similarly to human OA [59,103]. However, both OA models showed cartilage lesions and irregular distribution of chondrocytes at 8 weeks of induction [59]. Papain induced animal models have also been used in the study of aerobic exercise combined with glucosamine in the treatment of OA [104]. After the intervention of aerobic exercise in the OA model, the cartilage lesions were improved, the chondrocyte arrangement was more regulate, and the aggregation was significantly reduced [104]. Cyclooxygenase-2 (Cox2) inhibitors are used to treat pain associated with OA [105]. To investigate the mechanism of action of COX2 on chondrocytes, papain induced OA model was used. The results showed that COX2 prevented the terminal differentiation of articular chondrocytes by up-regulating the expression of PTHRP in the early stage of OA and promoted the mRNA expression of Col10a1 and Mmp13 [106]. Collagenase destroys the joint structure containing type I collagen, leading to cartilage degradation and eventually OA [107,108]. After a single injection of collagenase in the knee joint, the knee joint of mice developed degeneration, subchondral osteosclerosis, osteophyte formation, and then knee deformity [109]. The collagenase-induced OA mouse model can also be selected in the study of OA pain treatment [110]. The animal model of chemically induced OA is less traumatic, simple to operate, and more economical, and can also applied to the validation of some treatment schemes of OA and the mechanism research (Table 4).

Table 4.

Chemical induction models of OA

Drugs	Authors	Changes of the model	Molecular mechanisms
Monosodium iodoacetate (MIA)	Kwon et al. [94]	Knee swelling
		Contralateral hind leg load increased
		Articular cartilage and subchondral bone disorganized
		Osteophyte formation
		Cartilage erosion
	Hamdalla et al. [97]	Knee joint space stenosis	↑TNF-α, IL-6, IL-1β, TGF-β, NF-κB, iNOS, Caspase-3, MMP13
		Osteophyte formation	↓IL-10, Col II
		Cartilage erosion
		Chondrocytes reduced
Papain	Meng et al. [102]	Cartilage injury	↑MMP13
		Knee swelling	↓COL I, COL II
		Subchondral cyst
		Subchondral sclerosis
		Synovial inflammation
	Jia et al. [59]	Cartilage injury	↑SDF-1 in synovium, MMP-3
		Synovial inflammation	↓Col II, ACAN
		Fracture of cartilage
Collagenase	van der Kraan et al. [109]	Cartilage erosion
		Subchondral osteosclerosis
		Osteophyte formation
		Knee deformity

There is no single optimal experimental animal model for OA research. The selection of animal models should be considered from multiple perspectives, such as research objectives, economic budget, modeling feasibility and so on. The molecular mechanism of OA disease is more suitable to use the transgenic model, the pathological mechanism is more suitable to choose the surgical model, and the MIA induction model is preferred to study the pain mechanism and analgesic treatment. In the summary of the various models mentioned above, most of the OA models showed abnormal chondrocyte hypertrophy and changes in the expression of hypertrophy markers (Col10a1, Mmp13, Runx2) (Figure 3). Thus, chondrocyte hypertrophy may be a necessary process for the development of OA. Therefore, an understanding of each animal model is essential.

Figure 3.

Expression of COL10A1 in various OA-induced models. A. Immunohistochemical results of COL10A1and MMP13 in articular cartilage of transgenic OA model and control group [43]. B. Expression of COL10A1 and DLX5 in cartilage of DMM induced mouse OA model and control group [48]. C. Immunohistochemical results of COL10A1 and MMP13 in tibial articular cartilage of knee joint in mechanical load model and control group [11]. D. Immunohistochemical results of COL10A1 and MMP13 in articular cartilage in papain induced OA model and control group [106].

Osteosarcoma and its chondrogenic elements

OS is a rare primary bone malignancy, and most OS occurs in children and adolescents [111]. To date, the exact causes of OS are unknown. OS is largely considered a malignant tumor that originates in connective tissue, and mainly with tumor cells that produce bone or osteoid [112]. Although osteoblasts are the main cells of OS, there is no evidence that osteoblasts can be restored to primitive cells or malignant cells after differentiation from bone progenitor cells [113]. In fact, in most OS, there are also chondrocytes and fibroblasts that present in the matrix. It has been reported that there is cartilage matrix in the tumor tissues of mice transplanted with OS cell lines, and it has ossification similar to the process of endochondral ossification, suggesting that chondrogenesis or endochondral ossification plays a role in the development of OS [19,20]. RUNX2 is a transcription factor belonging to the Runt domain gene family, which is to play a crucial role in osteoblast differentiation and chondrocyte maturation [39,114,115]. Chondrocyte maturation of Runx2^-/- mice was seriously disturbed [116]. RUNX2 is also the main transcription factor that regulates the expression of Col10a1 gene, a specific molecular marker of hypertrophic chondrocytes [117]. RUNX2 is highly expressed in developing human bone and in various tumors, including OS [118,119]. RUNX2 plays a role in the evolution of the OS in several ways (Figure 4). Previous studies have demonstrated that interfering with Runx2 expression in OS cells can further inhibit the invasion of OS cells by inhibiting VEGF, MMP-2, MMP-9 expression [120]. Within chondrocytes, SOX9 is involved in cell cycle and differentiation [121]. In OS, Sox9 interacts with Runx2 and is part of the Runx2-regulated transcriptional circuit that promotes OS cells survival [122]. p53-deficient mice can lead to the development of OS and are enhanced by loss of pRB, suggesting an important role for P53 and pRB in the development of OS [123]. However, high expression of RUNX2 was shown in mouse osteosarcoma models lacking P53 and pRB [124]. Interestingly, RUNX2 has been shown to interact directly with P53 and pRB [23,125]. In normal osteoblast differentiation, RUNX2 interacts specifically with pRB during cell cycle exit [125]. In OS, pRB is absent, and RUNX2-pRB mediated cell cycle cannot withdraw and infinite proliferation occurs [126]. At the same time, the proliferation of bone progenitor cells and the expression of RUNX2 were increased in p53 deficient mice, thus promoting the maturation of osteoblasts and the occurrence of OS [23]. The specific interaction between RUNX2, P53 and pRB not only maintains the normal cycle process of osteoblasts, but also dysregulates it in OS [127]. Wnt/β-Catenin signaling is crucial in determining whether mesenchymal progenitors differentiate into osteoblasts or chondrocytes [128]. Studies have shown that abnormal activation of Wnt/β-Catenin pathway involved in autocrine Wnt signaling drives the proliferation of OS cells [129], so targeting Wnt/β-Catenin signal transduction may provide a new therapeutic approach. The Wnt/β-Catenin pathway has been shown to promote OS metastasis by activating RUNX2 expression in some OS [21]. The specific fusion protein EWS-FLI expressed in Ewing’s sarcoma has been shown to bind the FLI sequence to Runx2, thereby blocking osteoblast differentiation. Disrupting the interaction between RUNX2 and EWS-FLI can promote the differentiation of tumor cells [130]. The interaction between CBFβ and RUNX2 plays a key role in bone development. Mice without CBFβ showed severely delayed bone formation, chondrocyte maturation, and no endochondral ossification [131,132]. In OS, the expression of both CBFβ and RUNX2 is increased, resulting in impaired cell cycle. This control by heterodimer CBFβ-RUNX2 may contribute to the occurrence and development of OS [133]. Notch signal transduction is an old-fashioned cell interaction mechanism and an important regulator of the proliferation and differentiation of mouse chondrocytes [134,135]. Normally, Notch signaling pathway can regulate the expression of MMP13 through RUNX2 and promote hypertrophic differentiation of chondrocytes [135]. At the same time, high level expression of the Notch signal transduction target gene Hes1 has been found in metastatic OS cell lines, and studies have shown that Hes1 is required for the aggressiveness of OS [136]. Osteopontin (OPN) plays an important role in the process of bone formation and absorption, and is an crucial component of bone matrix [137]. Relevant studies have found that the combined expression of RUNX2/OPN promotes OS metastasis and may be a valuable independent predictor of OS metastasis and survival [138]. High mobility group box 1 protein (HMGB1) is a chromatin protein specifically secreted by differentiated chondrocytes to regulate endochondral ossification [139]. The expression of HMGB1 in OS cells can inhibit apoptosis and increase drug resistance [24]. A recent study showed that the chondrocyte derived exosome miR-195 can inhibit tumor cell proliferation and increase drug sensitivity of OS cells [140]. Therefore, chondrocytes play an important role in occurrence, development and treatment of OS, further investigation is needed.

Figure 4.

Regulatory mechanisms of Runx2 in OS. The interaction between SOX9 and RUNX2 promotes the cell survival of OS [122]. The deletion of P53 and pRB increases the expression of Runx2 and promotes the proliferation of OS cells [124,126]. The interaction of CBFβ and OPN with RUNX2 promotes cell proliferation, metastasis, and survival of OS [133,138]. Abnormal activation of Notch and Wnt/β-Catenin signaling pathways can play a role in cell proliferation and invasion of OS by acting to Runx2 [129,135,136]. Runx2 can promote the expression of Mmp13, Mmp2, Mmp9 and Vegf to further promote OS invasion [120,135].

Conclusion

OA is a multifactorial disease that not only causes degeneration of articular cartilage but also affects the subchondral bone, synovium, and surrounding muscles and ligaments. Although the exact pathogenic mechanism of OA is unknown, it is hypothesized that abnormal hypertrophy of articular chondrocytes in OA leads to matrix degradation and production of pro-inflammatory factors, ultimately resulting in inflammation in different animal models of OA. Notably, the suppression of inflammatory markers has been demonstrated by chondrocyte hypertrophy inhibitors [141]. At the same time, abnormal activation of chondrocyte hypertrophy and aging can accelerate the progression of OA, and regulating hypertrophic or senescent chondrocytes may be a potential therapeutic target to alleviate the progression of OA [142]. However, the exact link between cartilage degradation, inflammation, and hypertrophy of chondrocytes remains unclear [143]. Investigating the effects of chondrocyte hypertrophy on OA would thus provide a promising area of research for the development of potential therapeutic targets for OA.

OS is characterized by abnormal bone formation, but chondroid components were also found in OS tumor tissues [19]. Besides RUNX2, an essential transcription factor for collagen X expression and chondrocyte hypertrophy, which is highly expressed in OS and can cause infinite cell cycle proliferation of OS [126], another protein HMGB1, which is specifically secreted by differentiated chondrocytes in OS, can also inhibit apoptosis and promote cell proliferation in OS [24]. However, how chondrogenesis, hypertrophic differentiation or related signaling pathways may play a role in OS, and hold therapeutic potential for OS, remain to be explored.

Disclosure of conflict of interest

None.