Osteoarthritis (OA) is the progressive loss of normal joint function and is characterized by the degeneration of articular cartilage leading to pain, stiffness, and loss of joint mobility. OA is the world's leading cause of physical disability in adults, in the UK 20% of adults aged between 50–59 and 50% aged between 80–89 have symptomatic disease in one or both knee joints (Lacey et al., 2008). Current treatments for OA generally involve managing pain and inflammation, physiotherapy and lifestyle modifications in order to slow the progression of disease. When conservative forms of therapy are exhausted, patients and their rheumatologists may elect to have surgery.

Joint cartilage degeneration has multiple causes; biomechanical stress, metabolic and genetic defects, and, in many cases first manifests as a focal chondral or osteochondral lesion. Lesions can spread over the rest of the joint surface as the edges of the defect are more susceptible to degenerative change due to inequalities in biomechanical stress (Lefkoe et al., 1993; Braman et al., 2005). Much research has focused on repairing focal surface defects in order to spare the joint from further degeneration and recover pain free movement. A fundamental barrier to restoration of cartilage integrity is its slow intrinsic repair capacity that is largely attributable to a lack of vascularity, relatively low cellularity in adult tissue, and the presence of a dense hydrated extracellular matrix that inhibits cellular migration to and from the site of injury. Repair of lesions above a critical size of 3–5 mm in diameter must be initiated through extrinsic mechanisms such as microfracture of the subchondral bone plate to release blood and marrow into the defect to form a clot and fibrocartilaginous repair tissue (Pridie, 1959). A related technique, autologous chondrocyte implantation (ACI), is a two-part procedure, first the isolation and expansion of chondrocytes from intact non-weight-bearing regions of the joint and secondly their transplantation into an osteochondral defect in a weight-bearing region (Grande et al., 1989; Brittberg et al., 1994). The limited expansion capacity of dedifferentiated chondrocytes and their increasing inefficiency at redifferentiation during extended culture has led to the use of mesenchymal stem cells (MSCs) for autologous cartilage repair of larger defects (Wakitani et al., 1994). MSCs are found in numerous human tissues including bone marrow and adipose tissue. These MSCs have been shown to differentiate into bone, cartilage, muscle, and adipose tissue. Further studies have also shown cells displaying the properties of adult-derived MSCs can be isolated from the synovial joint tissues such as articular cartilage and synovial membrane (Alsalameh et al., 2004; Dowthwaite et al., 2004; Williams et al., 2010; McCarthy et al., 2012).

Cell-based therapy is becoming an established element of modern healthcare and is predicted to grow as knowledge and implementation of cell biology, biomaterials, and regenerative medicine increases. The aim of this article is to provide an overview of mesenchymal stem cell-based therapies for the treatment of lesion in articular cartilage. In the absence of disease modifying pharmacological agents (Mobasheri, 2013a) or biological therapies (Mobasheri, 2013b), cell-based therapies are one of the few available treatments for focal cartilage and osteoarthritic defects. In this review article we describe the significance of this topic in the context of the biology of the joint and the osteoarthritic disease process, and summarize key concepts and developments in the area of MSC-based therapies.

Articular cartilage is a smooth, thin and opalescent layer of hyaline tissue that covers the surface of bones of synovial joints (Figure 1). It is a highly organized and specialized tissue that permits free articulation, painless movement, and transmission of force through the skeleton due to the avascular, alymphatic and aneural composition. Compared to other tissues, articular cartilage has a low rate of metabolic activity (Pearle et al., 2005). The tissue is maintained by a single sparsely distributed cell, the chondrocyte (Archer and Francis-West, 2003a), that is surrounded by a highly hydrated extracellular matrix (ECM). The structure of cartilage is primarily composed of a highly organized collagen network that supports aggregating proteoglycans (PGs) and hyaluronan (HA) that is the most abundant non-PG forming glycosaminoglycan (GAG) in the cartilage matrix (Jeffery et al., 1991). Articular cartilage is divided into four zones; the superficial (tangential), transitional, radial, and calcified zones (Eyre, 2002). The physical and biochemical differences between the zones is important to allow cartilage to resist both extrinsic and intrinsic forces (Knudson and Knudson, 2001). In addition to the pseudo-stratified appearance of the matrix, it can be further subdivided by its composition around chondrocytes as pericellular, territorial and inter-territorial matrices. The collagen fibril arrangement and orientation, cellular density and proteoglycan composition vary with depth to provide the diverse mechanical properties required across the joint. The main collagens (type II, IX, and XI) provide tensile strength and the main proteoglycan, aggrecan, delivers compressive stiffness. Small proteoglycans, including decorin, biglycan, and fibromodulin, bind to other matrix macromolecules and thereby help to stabilize the matrix (Buckwalter and Mankin, 1998). Chondrocyte-matrix interactions are mediated through the pericellular and territorial matrices, facilitated by collagen type VI and non-collagenous proteins such as anchorin CII, tenascin, and fibronectin (Poole et al., 2001). This allows mechanical, electrical, and physicochemical signal transduction to the chondrocytes, directing the synthetic and degradative activity of chondrocytes (Mobasheri et al., 2002; Millward-Sadler and Salter, 2004). Cartilage tissue contains a large proportion of water (65–80% by wet weight). Chondrocytes themselves constitute 5–10% of the tissue's total volume, collagens form 10–30%, whilst proteoglycans and other molecules consist of 5–10% of the tissue's wet weight (Eyre, 2002; Archer et al., 2003b; Hunziker et al., 2007).

In adult cartilage, the prime function of articular chondrocytes is to synthesize and maintain the ECM required for tissue function. As the tissue is not vascularized it relies on diffusion from the articular surface for nutrient and metabolite exchange; therefore, chondrocytes are adapted to operate at a low oxygen tensions (ranging from 5% at the surface to <1% in the mid/deep layers) with the majority of the cell's energy requirements coming from glycolysis (Lafont, 2010). Despite mature articular chondrocytes having a relatively low metabolic rate, they display a remarkable synthetic capacity. The cartilage ECM is continually remodeled as chondrocytes replace matrix macromolecules lost through degradation as part of normal turnover (Goldring and Marcu, 2009). ECM maintenance depends on the ability of chondrocytes to detect alterations in the macromolecular composition and organization of the matrix, such as the presence of degraded macromolecules, and to respond by synthesizing appropriate types and amounts of new ECM components. The balance between catabolism and anabolism of the ECM maintains the healthy cartilage structure; a loss of equilibrium can lead to degeneration (Goldring and Marcu, 2009; Bader et al., 2011). During aging and disease, tissue homeostasis is disrupted; the rate of loss of ECM exceeds the rate of deposition of newly synthesized molecules (Goldring and Marcu, 2009). Although articular cartilage can tolerate repetitive physical loading, it has a low capacity for self-repair, therefore, an inability to quickly heal minor injuries makes cartilage particularly sensitive to progressive damage and degenerative diseases such as OA.

Articular cartilage defects caused by trauma or chronic injury are classified into two categories: Partial-thickness and full-thickness (Hunziker, 1999). Partial-thickness or chondral defects do not penetrate to the subchondral bone and therefore rely on healing through intrinsic processes that are limited by the tissue's inherent characteristics. There is evidence of transient chondrocyte proliferation at wound sites, however, there is little or no cellular migration into the injured sites to repair the defect (Hunziker and Rosenberg, 1996; Zhang et al., 2009). Furthermore, the reparative response usually terminates before the cartilage defect is fully healed, resulting in a lasting defect of mainly acellular tissue that reduces tissue function leading to further tissue degeneration (Hunziker, 1999; Zhang et al., 2009).

Full-thickness osteochondral defects penetrate the entire thickness of articular cartilage to the subchondral bone. Unlike partial-thickness defects, full-thickness defects are accessible to MSCs, macrophages, and blood cells, if there is a break in the underlying subchondral bone plate. In the latter case, a fibrin clot fills the defect and MSCs from bone marrow migrate into the repair matrix. Gradually MSCs replace the fibrin clot, stem cells differentiate into chondrocytes and secrete a proteoglycan-rich ECM to remodel the damaged area (Hunziker, 1999). However, it is frequently reported that a fibrous repair tissue is formed, which has weaker biomechanical properties and higher permeability (Kreuz et al., 2006). Consequently, the spontaneous repair process in full-thickness defects is a transient phenomenon and tissue degeneration eventually re-occurs. At this point, complete resurfacing is rarely observed leading to the development of secondary OA.

OA is a disease that affects the whole joint, including cartilage, subchondral bone, synovium, tendons, and muscles (Sellam and Berenbaum, 2010; Goldring, 2012; Loeser et al., 2012; Berenbaum et al., 2013; Mobasheri et al., 2014). The disease is characterized by degeneration of articular cartilage, low-grade synovial inflammation, and alterations in the joint soft tissues and subchondral bone (Goldring and Goldring, 2007; Sellam and Berenbaum, 2010; Mobasheri et al., 2014). Most commonly, OA occurs as primary or idiopathic OA; risk factors include obesity, overuse, joint instability, genetic or anatomical irregularities, metabolic disorders, muscle weakness, and various disorders of bone turnover (Mobasheri et al., 2014). The disease progresses in three stages; firstly there are changes in or loss of the ECM, then chondrocytes attempt to remodel and repair the tissue and finally, the degeneration occurs to a greater extent than synthesis causing incremental loss of cartilage (Hendren and Beeson, 2009). Cartilage degeneration results in bone-to-bone articulation, inflammation, significant pain, and disability. A major component of OA is inflammation, there is an increased activity of many cytokines, chemokines, and adipokines (Loeser et al., 2012). The inflamed synovium produces catabolic and pro-inflammatory mediators such as cytokines, nitric oxide, prostaglandin E₂ (PGE₂), and neuropeptides. These molecules unbalance the matrix degradation and repair equilibrium resulting in excessive secretion of the proteolytic enzymes and cartilage ECM breakdown (Goldring and Goldring, 2007; Sellam and Berenbaum, 2010). Cartilage alterations induce further synovial inflammation, creating a progressive loop that exacerbates clinical symptoms and stimulates further joint degradation.

As cartilage ages, there are a number of biochemical and morphological changes that occur within the tissue that alter the mechanical properties and decrease the ability of chondrocytes to maintain articular cartilage (Martin and Buckwalter, 2002). Damage-induced cellular alterations result in a senescence-associated secretory phenotype characterized by the production and secretion of cytokines, chemokines, and proteases (Loeser, 2013; Mobasheri et al., 2014). Oxidative stress and alterations in mechanical loading can further promote changes in the chondrocytes. OA is an active, progressive, degenerative disease that is the common pathway of age related damage, acute injury and inflammatory degradation of synovial joints.

As mentioned previously cartilage has a poor intrinsic repair capacity, tears and defects larger than 3 mm rarely heal and potentially can lead to OA of the joint. Poor repair is partly due to the lack of blood vessels, which are required for efficient repair responses, and, because of a low cell density with chondrocytes embedded within lacunae, unable to easily migrate to damaged areas (Henrotin et al., 2005). Therefore, extrinsic intervention is often necessary to repair damaged tissue and to prevent further damage. The most effective treatment for end-stage OA is arthroplastic prosthetic replacement. However, due to the finite lifespan of prostheses, this is unsuitable for patients younger than 45 years. A variety of treatment methods have been used to repair defects and stimulate the formation of new articular cartilage. These include lavage, microfracture, osteochondral allograft transfer system (OATS), ACI and osteotomy (Bhosale and Richardson, 2008). Each procedure has varied success and many factors have to be considered including defect size and patient age when selecting the correct treatment.

The most widely used method that surgeons adopt for cartilage repair involves penetration of the subchondral bone causing the release of multipotent MSCs in a process defined as marrow stimulation. Previous iterations of this method have included Pridie drilling, spongialization and abrasion therapy, though the most popular modern method is microfracture (Ronn et al., 2011). Under arthroscopy, tiny fractures are made using an awl in the subchondral bone, inducing bleeding and a fibrin clot that is infiltrated with stem cells from the bone marrow. Stem cells fill the defect, differentiate and form a repair tissue (Yen et al., 2008) that is fibrocartilaginous and not suitable to withstand the demands of everyday activities, and thus, at higher risk of future breakdown (Redman et al., 2005). Microfracture, however, does offer symptomatic relief; in a follow up study it was reported that pain was decreased and function improved in 95% of the study population after up to 17 years post-surgery (Steadman et al., 2003).

An alternative intervention developed by Brittberg and colleagues, ACI, is currently regarded as the “gold standard” for cartilage repair (Brittberg et al., 1994). Cartilage is harvested from a low weight-bearing region, chondrocytes are isolated and expanded in vitro before being injected into a full-thickness articular defect under a periosteal patch stitched over the defect and sealed in with fibrin glue (Brittberg et al., 1994, 2003; Redman et al., 2005). Implanted chondrocytes begin the process of producing neo-cartilage through the production of ECM. ACI has been shown to produce effective and durable repair tissue, relieving symptoms and clinical success remains high, even after 20 years post-implantation (Peterson et al., 2010). The repair tissue produced by ACI has been shown to be varied but in general is more hyaline-like than produced using microfracture. However, there is often an abundance of type I collagen which is also characteristic of fibrocartilage (Roberts et al., 2002). Improvements in the procedure have led to second generation ACI techniques; synthetic collagen membranes have replaced the periosteal flap, and several biomaterial and natural scaffolds have been developed into which the chondrocytes are seeded (Redman et al., 2005). Despite the encouraging clinical outcomes ACI has a number of disadvantages; it requires multiple surgeries and is more invasive than microfracture, treatable defect size is limited by the finite amount of harvestable donor tissue and the restricted in vitro expansion of chondrocytes before de-differentiation makes their use redundant (Barbero et al., 2003). In follow-up studies, it has been shown that 1 year post-operatively, ACI offers significantly improved repair compared to microfracture (Visña et al., 2004); however, after 2–5 years randomized trials show no significant difference in repair efficiency between ACI and microfracture (Knutsen et al., 2007; Van Assche et al., 2010). The limitations of current surgical strategies have led to investigations into the use of adult stem cells from various tissue sources in an endeavor to improve hyaline-like cartilaginous repair and increase the treatable defect size.

Friedenstein first characterized clonogenic fibroblast-like cells extracted from bone marrow via attachment to tissue culture plastic (Friedenstein et al., 1976). These marrow-derived stromal cells were found to be inherently osteogenic but displayed plasticity being capable of differentiating into multiple cell types of the mesodermal lineage. MSCs have been shown to form cartilage, bone, adipose tissue, intervertebral disc, ligaments, and muscle (Prockop, 1997; Pittenger et al., 1999). Therefore, MSCs are typically defined as adherent, self-renewing, fibroblastoid-like cells that can differentiate to osteoblasts, adipocytes, and chondrocytes in vitro (Barry and Murphy, 2004; Phinney and Prockop, 2007). Self-renewal refers to the biological pathways and mechanisms that preserve the undifferentiated stem cell state. In MSCs this capacity for self-renewal is in part due to telomerase reverse transcriptase (TERT) activity (Kolf et al., 2007). Additionally, leukemia inhibitory factor (LIF), fibroblast growth factors (FGFs), Wnts and other growth factors and cytokines, have been implicated in maintenance of the MSC phenotype (Tsutsumi et al., 2001; Metcalf, 2003; Kléber and Sommer, 2004; Kolf et al., 2007). These factors have also been shown to be critical for self-renewal and maintenance of undifferentiated embryonic mesenchymal tissue.

It is widely accepted that primary MSC cultures are a heterogeneous population of cells with varying capacities of self-renewal and differentiation (Ho et al., 2008; Phinney, 2012). Their heterogeneity means no singular unique marker is available for identification and isolation (Table 1). Therefore, a panel of positive and negative markers must be used for the selection criteria. MSC populations commonly express surface proteins including CD29, CD44, CD49a–f, CD51, CD73, CD90, CD105, CD106, CD166, and Stro1 and must be negative for hematopoietic lineage markers including CD11b, CD14, and CD45 (Halfon et al., 2011). The optimal panel of marker for selection is frequently debated and numerous additional markers have been reported in the literature. The International Society of Cellular Therapy recommended that MSCs must meet a minimum criteria including: Adhesion to plastic, expression of surface markers CD73, CD90, CD105 (≥95% expression), the absence of hematopoietic markers CD34, CD45, CD14 or CD11b, CD79α or CD19 (≤2%), the absence of HLA Class II molecules and tripotent differentiation into chondrogenic, osteogenic, and adipogenic phenotypes (Horwitz et al., 2005; Dominici et al., 2006).

Whilst MSC capacity for multipotent differentiation represents a great potential for regenerative medicine, beneficial effects are also achieved by MSC-mediated improvements in native cell viability and proliferation, reduction in apoptosis and anti-inflammatory immunomodulatory effects (Caplan, 2009; Glenn and Whartenby, 2014). These additional reparative effects are achieved through MSC secretion of paracrine growth factors and cytokines, direct cell-cell interactions through tunneling nanotubes and release of extracellular vesicles containing peptides, mRNA and microRNAs (Figure 2); for a detailed review of the various MSC repair mechanisms please see Spees et al. (2016). Treatment of articular cartilage with intra-articular injection of MSCs has shown that some regeneration of joint tissues can occur; however, studies have shown that the source of repair tissue is primarily derived from native cells (Murphy et al., 2003). In support of the latter studies, Horie and colleagues showed that inter-articular injection of human MSCs into the rat knee increased the expression of type II collagen in resident cells (Horie et al., 2012). These findings indicate that MSCs also coordinate and enhance the reparative response rather than exclusively replacing lost and injured tissue (Wyles et al., 2015). MSC-enhanced repair is induced via secretion of various anabolic paracrine factors, such as transforming growth factor beta (TGF-β), vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF) and other bioactive molecules that modulate reparative responses (Tille and Pepper, 2002; Sorrell et al., 2009; Freitag et al., 2016). Chondrocytes from end stage diseased cartilage remain metabolically active and continue to express genes for cartilaginous matrix molecules, corroborating the hypothesis that MSCs support repair and modulate these chondrocytes (Freitag et al., 2016).

Recent studies have revealed that MSCs interact with immune cells and eliminate survival factors, suppress inflammatory T cell proliferation, inhibit maturation of monocytes and myeloid dendritic cells, and re-calibrate the chemokine gradient (Glenn and Whartenby, 2014). MSCs inhibit several aspects of B cell activity, including activation, proliferation, chemokine receptor expression, and differentiation into antibody-secreting plasma cells (Glenn and Whartenby, 2014). MSCs have also been shown to induce nitric oxide (NO) in response to inflammatory cytokines to suppress CD8+ T cell proliferation, cytokine production, and cytotoxicity (Glenn and Whartenby, 2014). MSCs also produce many soluble factors including TGF-β, hepatocyte growth factor (HGF), PGE₂ and interleukin 10 (IL-10) that suppress CD4+ T cell proliferation and polarize them toward T-helper cells (TH1/TH2) and anti-inflammatory T-regulatory cells (Tregs) (Glenn and Whartenby, 2014), promoting T cell apoptosis and reduce inflammation, and providing a favorable environment for tissue repair. After in vivo administration, MSCs can induce peripheral tolerance and migrate to injured tissues where they have the capacity to exert immunosuppressive properties (Glenn and Whartenby, 2014) and inhibit the release of pro-inflammatory cytokines and promote the survival of existing cells and the repair of damaged tissue (Uccelli et al., 2008). Using a murine in vivo model of OA it has been shown that MSCs deactivate macrophages stimulated by cartilage breakdown products, preventing further catabolism of the cartilage and also elevating macrophage production of anabolic growth factors (ter Huurne et al., 2012). Due to their strong immunomodulatory phenotype MSCs are being explored clinically for treatment of a variety of immune-mediated diseases such as motor neuron disease and Sjögren's syndrome (Parekkadan and Milwid, 2010).

The versatility of MSCs has led to the development of cellular therapies for tissue engineering and regenerative medicine. According to data reported by the US National Institutes of Health (http://www.clinicaltrial.gov/) there are approximately 500 clinical trials currently utilizing MSCs as therapy for a diverse range of disease (Squillaro et al., 2016). MSCs seem to provide clear benefits over chondrocytes when treating degenerative conditions such as OA. MSCs can be expanded for a prolonged period ex vivo due to their capability of self-renewal and also maintain their phenotype to a greater degree than dedifferentiated chondrocytes in culture. Additionally, MSCs have the potential to form all joint tissues and this hypothetically enables them to repair both lesions in the articular cartilage or osteochondral defects as well as other joint structures such as ligaments.

The “niche” incorporates all the factors required to regulate stem cell self-renewal and differentiation including differentiated or stromal cells in direct or indirect contact with stem cells, ECM molecules and soluble mediators within the milieu (Schofield, 1978; Kolf et al., 2007). It is believed that certain signals enter the niche or that stem cell leave the niche in order to initiate differentiation for regeneration or repopulation of a tissue (Kolf et al., 2007). Multiple MSC niches have been described within various tissues (Figure 3) and stem cells housed in them isolated and utilized for chondrogenic repair. MSCs have been found in the bone marrow, adipose tissue, articular cartilage, synovium, skeletal muscle, dental pulp, circulatory system, heart, brain, umbilical cord tissues (including Wharton's jelly), and other connective tissues (Gronthos et al., 2000; Zuk et al., 2001; Alsalameh et al., 2004; Dowthwaite et al., 2004; Crisan et al., 2008; Gronthos and Zannettino, 2011; Batsali et al., 2013; Mobasheri et al., 2014; Steward and Kelly, 2015).

Articular cartilage originally develops from multipotent progenitor cells in the presumptive skeletal embryonic mesenchyme. Under the influence of external and intrinsic factors, chondroprogenitor MSCs transform into chondroblasts and start depositing cartilage-specific ECM. Chondrogenesis of chondroprogenitor mesenchymal cells starts with the condensation and nodule formation and is highly dependent on the interplay between differentiating chondrocytes and their ECM. Understanding the mechanisms underlying the molecular control of chondrogenesis has important implications for the development of novel, effective treatment approaches and improving cartilage tissue engineering.

Although initially considered as a tissue with a simple structure, reproducing the stratified appearance of cartilage has proven to be difficult. The task can be subdivided into two main phases, regeneration of cartilage tissue, achieved through mimicking embryonic chondrogenesis, and then remodeling of this repair tissue to generate the adult form and hence function. In an analogy to skin wound healing, which forms the longest part of the repair process, remodeling of the repair cartilage is predicted to take months or years, but, only through reconstitution of the original form can we expect to attain long-lasting and functional repair of osteochondral defects. We know that joint formation and maturation are driven by progenitor cells that exhibit the characteristics of MSCs, and therefore, these represent the best means to enable us to direct extrinsic repair processes through surgical intervention.

Historically, bone marrow-derived MSCs have been used to heal cartilage defects through implantation, however there are several important reasons why they are not optimal for durable repair strategies. First, BM-MSCs induce fibrocartilagenous repair, defined as matrix production that has a higher ratio of collagen type I to II than native tissue, and secondly, there is no evidence that this cartilage can be remodeled to form stratified hyaline cartilage. The presence of type I collagen is per se not deleterious, it is in fact normal for immature cartilage to be rich in type I cartilage in the superficial zone. Type I collagen expression is lost as the tissue matures and the protein cannot be immunolocalized in adult articular cartilage. Collagen type I and associated collagens types III and V reappear in injury and disease. The inherent problem with BM-MSCs is that collagen type I expression is an inherent part of their phenotype and transcriptional profile, hence the formation of fibrocartilagenous repair tissue. The most logical use of BM-MSCs in cartilage repair strategies would be in their native context, as a mesenchymal stromal cell, supporting repair of the local parenchymal cells, chondrocytes, through the supply of paracrine factors and generation of a provisional matrix. Similarly, there is scant evidence that other non-articular derived MSCs are capable of generating cartilage that can be remodeled to form stratified tissue, though each are capable of chondrogenic differentiation, and, producing paracrine factors that enable endogenous repair, particularly in fibrocartilagenous tissues such as menisci.

Thus far, articular cartilage-derived CPCs appear to have the attributes to produce repair tissue conducive to long-term repair. CPCs not only produce hyaline cartilage but also do not undergo terminal differentiation to produce calcifying cartilage. Most significantly, these cells are critical determinants in the maturational transition of cartilage from one that has an immature isotropic structure to one that is anisotropic and mature. In general, it has been predicted that maturation is essentially complete by puberty, which in rabbits is 3 months, and consequently much longer in humans. MRI imaging of ACI treated patients predicts that the remodeling phase is complete in 2 years, which gives a lower time estimate, although biopsies of these repair tissues indicate that the morphology of normal adult cartilage is seldom achieved. Work from a number of groups has shown that osteoarthritic cartilage contains a viable pool of CPCs that display all the qualitative attributes of MSCs. The question arises why do CPCs persist in diseased cartilage, and, can they be reactivated? We do not know whether the chondrogenic activity of CPCs from diseased cartilage is impaired in vitro compared to progenitors from normal cartilage. Similarly we are in the dark to whether epigenetic changes in OA CPCs resulting as a consequence of metabolic disease, chronic overloading, or genetic deficits affect their function. If this latter is shown to be the case, various reprogramming strategies to alter the epigenetic profile of these cells may induce reactivation of their chondrogenic phenotype. There is data to suggest that CPCs reside in a stem cell niche which is demarcated by unusual chondroitin sulfate epitopes. The loss of this niche through proteoglycan depletion following injury and disease may be a factor as to why CPCs become dysregulated and unable to differentiate, though data indicates self-renewal is not affected. Again reactivation of CPCs will have to take into account reestablishment of the niche, either through genetic reprogramming or presentation of the niche in biomaterials.

The presence of viable CPCs in diseased cartilage suggests that attempts to induce repair through endogenous mechanisms may be still within reach. MSCs from different origins may be useful in combination therapies where CPCs produce a hyaline repair tissue and adipose or bone marrow-derived MSCs provide supportive paracrine factors that enhance cell viability and reduce inflammation. When selecting the appropriate MSC source the balance will be between expediency and effectiveness. If we look at the options then almost every cell type has its' own advantages; ease of isolation, expansion, immunomodulatory effects, and chondrogenic capacity. In addition there are many challenges associated with isolating, expanding and differentiating MSCs for subsequent implantation into degenerate joints. The physiological microenvironment of diseased joints is hypoxic, acidic, deprived of nutrients, and contains high concentrations of pro-inflammatory cytokines and reactive oxygen species. Furthermore, MSCs may be exposed to abnormal physical loads as the joint has already been biomechanically compromised. Thus, future regenerative medicine strategies will need to address these remaining concerns possibly through addition of surgical interventions such joint realignment. As next steps, focusing efforts toward achieving standardized methods of MSC isolation, characterization, and administration has great potential to provide powerful new treatments. In addition, bioprinting technology offers the possibilities of delivering cells, growth factors, and biomaterials precisely to the required position and may be key to engineer hyaline cartilage. Given that patients present with different problems it is unlikely that there will be a simple “one size fits all” answer but probably an array of solutions that need to be carefully aligned to the individual's requirements.

CF drafted the article. CM, RZ, IK, and AM contributed to the interpretation and critical appraisal of the literature, contributed text, and edited the manuscript. CM and CF prepared the original artwork. All authors have made substantial intellectual contributions to the evolution of the concepts presented in the manuscript and approved the final version submitted.

The authors' work is supported by the European Commission and Arthritis Research UK. AM is the co-ordinator of the D-BOARD Consortium funded by European Commission Framework 7 programme (EU FP7; HEALTH.2012.2.4.5-2, project number 305815, Novel Diagnostics and Biomarkers for Early Identification of Chronic Inflammatory Joint Diseases). CM is supported by the New National Excellence Program of the Ministry of Human Capacities in Hungary, and the European Commission through a Marie Curie Intra-European Fellowship for Career Development grant (project number 625746; acronym: CHONDRION; FP7-PEOPLE-2013-IEF) awarded to AM. AM is also a member of the Applied Public-Private Research enabling OsteoArthritis Clinical Headway (APPROACH) Consortium, a 5-year project funded by the European Commission's Innovative Medicines Initiative (IMI). APPROACH is a public-private partnership directed toward osteoarthritis biomarker development through the establishment of a heavily phenotyped and comprehensively analyzed longitudinal cohort. The research leading to these results has received partial support from the Innovative Medicines Initiative (IMI) Joint Undertaking under grant agreement no. 115770, resources of which are composed of financial contribution from the European Union's Seventh Framework programme (FP7/2007-2013) and EFPIA companies' in kind contribution. AM has received funding from Arthritis Research UK (grant number 21076) and financial support from the Deanship of Scientific Research (DSR), King Abdulaziz University (grant no. 1-141/1434 HiCi). He is also member of the Arthritis Research UK Centre for Sport, Exercise, and Osteoarthritis, funded by Arthritis Research UK (Grant Reference Number: 20194). IK is supported by the UK Regenerative Medicine Platform.