Mesenchymal chondroprogenitor cell origin and therapeutic potential

Mesenchymal progenitor cells, a multipotent adult stem cell population, have the ability to differentiate into cells of connective tissue lineages, including fat, cartilage, bone and muscle, and therefore generate a great deal of interest for their potential use in regenerative medicine. During development, endochondral bone is formed from a template of cartilage that transforms into bone; however, mature articular cartilage remains in the articulating joints, where its principal role is reducing friction and dispersing mechanical load. Articular cartilage is prone to damage from sports injuries or ageing, which regularly progresses to more serious joint disorders, such as osteoarthritis. Osteoarthritis is a degenerative joint disease characterized by the thinning and eventual wearing of articular cartilage, and affects millions of people worldwide. Due to low chondrocyte motility and proliferative rates, and complicated by the absence of blood vessels, cartilage has a limited ability to self-repair. Current pharmaceutical and surgical interventions fail to generate repair tissue with the mechanical and cellular properties of native host cartilage. The long-term success of cartilage repair will therefore depend on regenerative methodologies resulting in the restoration of articular cartilage that closely duplicates the native tissue. For cell-based therapies, the optimal cell source must be readily accessible with easily isolated, abundant cells capable of collagen type II and sulfated proteoglycan production in appropriate proportions. Although a cell source with these therapeutic properties remains elusive, mesenchymal chondroprogenitors retain their expansion capacity with the promise of reproducing the structural or biomechanical properties of healthy articular cartilage. As current knowledge regarding chondroprogenitors is relatively limited, this review will focus on their origin and therapeutic application.


Introduction
Osteoarthritis (OA) is a chronic disease of the joints characterized by progressive destruction of articular cartilage resulting in painful, limited joint movement. In the European Union over 39 million people exhibit symptoms of OA, a number anticipated to double in the next decade, creating an imperative for the timely development of eff ective treatments for the disease [1]. Current clinical therapies such as pharmaceutical inter ventions, bone marrow stimulation techniques or micro fracture do not result in regeneration of healthy cartilage tissue [2,3], but focus on the short-term relief of OA symptoms. When pharmaceutical intervention fails, clinicians regularly revert to invasive and permanent solutions.
Th e fi rst widely accepted regenerative treatment for cartilage repair was autologous chondrocyte transplantation. Despite its initial therapeutic promise, chondrocyte transplantation has associated complications, such as donor site morbidity, repair cell de-diff erentiation with expansion in vitro and restricted cellular life span upon implantation [4]. Immature progenitor cells with the potential to develop into mature tissues in response to appropriate cues have therefore become a primary focus of cartilage repair strategies as an alternative to chondrocyte-based methods [5].
Th e application of chondroprogenitors, cells that are specifi cally pre-disposed to diff erentiate into mature chondrocytes, to repair articular lesions and subsequently inhibit the onset of OA is a current focus of research eff orts. As the mature articular joint develops from embryonic mesodermal precursors that diff erentiate into chondroprogenitors and ultimately into mature adult chondrocytes or synoviocytes, it is hypothesized that progenitors retained in these adult articular tissues provide a potential reservoir of chondroprogenitors.

Development of the chondroprogenitor
Th e development of the embryonic appendicular skeleton, whereby undiff erentiated limb mesenchyme matures into a cartilaginous precursor and subsequently into bone, is dependent upon precursor exposure to specifi c combinations of morphogens and mechanical stimuli. Presumably a residual chondroprogenitor in the adult has been similarly primed and, when further stimulated, will respond by undergoing chondrogenic diff erentiation. Of critical importance during this developmental process are the transforming growth factor (TGF)-β, fi broblast growth factor (FGF), Wnt and Notch signaling pathways.
In embryonic limb development, FGF-4 stimulates Sonic hedgehog (Shh) expression in a positive feedback loop that coordinates proximal-distal and anterior-posterior patterning of the cartilaginous anlagen, as was historically demonstrated in an avian model [6]. Shh, in turn, initiates a cascade of stimulatory molecules such as those of the TGF-β superfamily, thereby inducing mesenchymal diff er entiation into chondrocytes, as was originally demon strated in murine limb mesenchyme [7]. Similarly, FGF-18 promotes cartilage formation in murine limb progenitor in vitro micromass cultures [8]. Th is eff ect is currently under clinical investigation for cartilage repair [9]. Th is phase I safety study assesses the ability of FGF-18 to stimulate chondrocyte development leading to the repair and regeneration of articular cartilage in patients undergoing knee replacement surgery. As FGF receptor isoform expression is highly regulated during each stage of embryonic human limb chondrogenesis [10], conserved FGF receptor expression patterns have been identifi ed in embryonic chondrogenesis in vivo and progenitor cell diff erentiation in vitro, enabling a mechanism to compare tissue engineered cartilage with natural development [11].
Mesenchymal cell diff erentiation into chondrocytes, and the associated regulation of extracellular matrix (ECM) deposition, is minutely co ordinated by paracrine factors. TGF-β and bone morphogenetic protein (BMP) signal ing, often through Sox9 as a transcriptional mediator, are responsible for initiating expression of carti laginous ECM such as aggrecan, collagen types II and XI, fi bronectin and tenascin in in vitro murine micro mass cultures [12,13]. Opposing roles for the involve ment of the mitogen-activated protein kinase signal ing components ERK-1 and p38 downstream of TGF-β superfamily stimulation have been identifi ed during in vitro chondrogenesis [14]; suppres sion of Erk-1/2 resulted in enhanced chondrogenesis whereas inhibition of p38 suppressed cartilage formation [15]. More specifi cally, inhibition of p38 signifi cantly repressed formation of pre-chondrogenic nodules in vitro, sustained N-cadherin expression and increased expression of fi bronectin and α5β1 integrin, while ERK inhibition showed no eff ect on nodule formation, reduced expression of N-cadherin and accelerated reduc tion of fi bronectin and α5β1 integrin expression during late stages of chondrogenesis [15]. Th is delicate balance between signaling mechanisms is sustained in mature cartilage, ensuring maintenance of a healthy articular tissue [16].
Chondrogenic diff erentiation of adult human progenitor cells in vitro through TGF-β1 is mediated intracellularly by both Smad3 and Wnt-associated β-catenin [17]. By increasing nodule formation, Wnts 5a and 5b have been shown to promote early chondrogenesis in vitro [18] and play an important role in chondrocyte diff er entiation and proliferation in an in vivo mouse model by controlling expression of cell cycle regulators such as cyclin D1 and p130 [19]. However, expression of Wnts 14, 7a and 1 have been demonstrated to inhibit the chondrocyte maturation in in vivo chick models and in in vitro micromass cultures [20].
A distinct population of murine embryonic precursors located at the prospective joint will develop into adult articular cartilage and synovial tissues, in contrast to the remaining cartilaginous template that ossifi es to form the appendicular skeleton [21]. As the embryonic development of articular cartilage proceeds from the preliminary cartilaginous template to an articulated joint with striated cartilage and underlying subchondral bone, Notch signaling plays a fundamental role as demonstrated in vivo with an avian model [22]. Notch signaling via JAG1 is required to initiate chondrogenesis in adult human progenitor cells in vitro [23]; however, maintained Notch signaling in adult human progenitor cell micromass cultures suppresses diff erentiation by inhibiting Sox9 binding of the collagen type II promoter [24], indicating that temporal regulation of Notch signaling is paramount to proper development.
Recent advances indicate a strong association between hypoxia-inducible factors (HIFs) in chondrogenesis and the development of OA. While HIF-1α stimulates chondrogenic diff erentiation, HIF-2α is involved in endo chondral ossifi cation and cartilage degradation. Amarilio and colleagues [25] inactivated HIF-1α in mouse limb bud mesenchyme, resulting in abnormal joint and carti lage formation in vivo, indicating HIF-1α regulates the diff erentiation of pre-chondrocytes. Further in vitro studies demonstrated reduced cartilage formation in HIF-1α-depleted micromass cultures and reduced ex pres sion of key chondrogenic markers Sox9, Sox6, aggrecan and collagen type II [25]. In vivo functional studies by Yang and colleagues [26] demonstrated that HIF-2α is a critical transcription factor that catabolically regulates cartilage degradation through matrix metallo proteinase and ADAMTS (a disintegrin and metallo proteinase with thrombospondin motifs) expression.
Th e delicate balance of paracrine factors discussed above, as well as several other factors not reviewed here, results in the diff erentiation of an embryonic mesenchymal progenitor cell into a mature articular chondrocyte. Presumably chondroprogenitors residing in the adult have also been primed by these signals and will therefore respond similarly by undergoing chondrogenic diff erentiation when further stimulated, making them an attractive cell source for the regeneration of cartilage. Th e successful identifi cation of a pool of chondroprogenitors in vivo, followed by their isolation and in vitro expansion, is therefore a pre-requisite to their clinical application.

Progenitor cell tissue distribution
Adult progenitor cells have been described in vivo and in vitro as mobile cells with both diff erentiation and selfrenewal potential that reside in a niche adjacent to mature, diff erentiated cells [27]. Progenitors have been successfully retrieved from several sources, including adipose, synovium, synovial fl uid, perichondrium and bone marrow [28][29][30][31] (summarized in Table 1). Depending on the tissue source from which they are isolated, progenitors harbor distinct diff erentiation potential and occur at variable frequencies.
Eff orts by Friedenstein and colleagues [31] and Pittenger and colleagues [32] unequivocally demon strated the presence and potential of bone marrow-derived mesenchymal stem cells (MSCs) with the innate ability to diff erentiate into multiple mesenchymal lineages in vitro. Despite their rare occurrence (0.001 to 0.01% of cells in the stromal compartment [32]), bone marrow-derived MSCs may be effi ciently isolated and expanded in culture without associated donor site morbidity.
Bone marrow-derived MSCs are isolated nonselectively by exposing the mononuclear cell aspiration to tissue culture plastic. Adherent progenitors, termed colony forming unit-fi broblasts (CFU-Fs), initiate the generation of a clonal population of CD105-positive, CD73-positive, CD90-positive and CD44-positive cells that are then expanded in monolayer [33]. Clonal analysis of bone marrow-derived MSCs describes 20 to 50% of the total population of cells to be truly tri-potent, but most importantly has identifi ed a subpopulation of MSCs that retains only chondrogenic potential, or chondropro genitors [34]. As the resultant population of MSCs is heterogeneous, eff orts continue to identify and isolate a homoge nous chondroprogenitor cell population from marrow that would more effi ca ciously repair diseased cartilage.
Similar to those in marrow, adult periosteal-derived multipotent progenitors are tri-potent [35]. Regardless of donor age, periosteal progenitors are clonogenic and have signifi cant in vitro expansion potential [36] with con tinued positive expression of traditional bone marrow-derived progenitor cell markers such as CD105, CD166, CD13, CD73 and D7-FIB [28]. Most importantly, identifi cation of specifi cally chondro-and osteogenic precursors within the periosteum has been identifi ed, making this tissue an attractive tissue source for chondroprogenitor isolation [37].
Th e mature synovial membrane is composed of a fi brous external layer and an inner secretory layer that produces synovial fl uid. Synovial tissue-derived multilineage progenitors may have the greatest overall therapeutic potential to regenerate damaged cartilage due to their impressive capacity for proliferation and their superior chondrogenic diff erentiation potential [38]. Interestingly, during the early stages of OA there is an increase in progenitor cell numbers in synovial fl uid, presumably as a result of synovial membrane degradation leading to their release into the synovial fl uid [39]. Synovial membrane derived mesodermal progenitors uniquely are not depleted in number or potential with donor age [38], making them an attractive cell source for autologous therapy.
Like synovium and bone marrow, infrapatellar fat pad contains a readily abundant source of CD105-positive, CD44-positive and CD166-positive progenitor cells with the potential to diff erentiate into mesodermal lineages. CD271 expres sion, a putative marker of the in vivo progenitor cell [40], is highly expressed in adipose progenitors and is uniquely maintained with proliferation, indicating retention of their progenitor capacity upon expansion [29].
Although healthy mature cartilage is primarily composed of mature chondrocytes, progenitor cells with chondro genic capacity have been isolated from the super fi cial zone of articular cartilage [41]. Additionally, chondroprogenitors have been identifi ed in arthritic cartilage after their migration from the bone marrow through breaks in the tidemark and into the diseased cartilage [42]. With two potential sources of progenitor cells, one in the healthy superfi cial zone and another generated as a repair response to disease, the development of technologies to harness and retain their potential may off er novel regenerative therapeutics.
Progenitor cells have also been identifi ed in tendon, specifi cally in an ECM-rich niche. Th e tendon-derived stem/progenitor cells (TSPCs) are clonogenic with a higher rate of proliferation compared to bone marrowderived MSCs. TSPCs are multipotent with an enhanced potential for osteo-and adipogenic diff erentiation compared to bone marrow-derived progenitors. With diminished chondrogenic potential and profi cient tenogenic potential, TSPCs are perhaps best suited for tendon repair applications [43].
Comparison of progenitors derived from adult bone marrow, synovium, periosteum and adipose has demon strated distinct diff erences between these cell types [29,30]. Upon initial isolation of progenitor cells, there are signifi cantly fewer CFU-Fs in bone marrowderived cells compared to progenitors from adipose, periosteum or synovium. Synovial fl uid and bone marrow-derived pro genitor cells share a similar chondrogenic or osteo genic potential [44]; however, synovial membrane-derived cells in particular retain signifi cant potential for chondrogenesis [38]. Together, these adult tissue sources off er an opportunity to specifi cally isolate progenitors that when clonally analyzed are predisposed to diff er en tiate into chondro cytes, thereby off ering a source of highly effi cacious repair cells. Figure 1 depicts some possible sources of chondroprogenitors. Th e increased potential of synovium-derived progenitors may be associated with the proximity of the niche they reside in to articular cartilage. A common embryonic genesis for articular cartilage and synovial tissue may also be a factor [21]. Migration of progenitor-like cells from marrow to OA cartilage has been suggested but has not as yet been identifi ed in the healthy joint and the presence of marrow-derived mesenchymal-like progenitors in rheumatoid arthritic pannus highlights an additional route for migration of marrow-derived chondroprogeni tors [45]. However, identifi cation of the optimal chondro progenitor awaits a direct intra-donor comparison of cells isolated from the various sources.

Clinical considerations and therapeutic applications
Articular chondrocytes, when explanted and expanded in vitro, lose their chondrocytic phenotype as indicated by morphometric changes and elimination of collagen type II deposition [46]. Th e re-diff erentiation of these cells in vitro regenerates their articular phenotype while the diff erentiation of a progenitor cell in vitro results in the generation of a transient, pre-hypertrophic chondro cyte, similar to the chondrocyte phenotype in the developing embryonic skeleton [47], highlighting the in herent diff erence between progenitors and native chon dro cytes [48]. Essential to the functional formation of neocartilage through the implementation of progeni tors is the acknowledgement of these evident diff erences in the phenotypic state, and our ability to develop methods to overcome the present gaps. It is crucial that research focuses on not only the signaling events that are known to support chondrogenic diff erentiation in the embryo but also on the molecular events that fail to initiate in progenitor cell diff erentiation.
Bone marrow-derived MSCs have been directly injected [49] or combined with a scaff old and implanted [50] into the intra-articular space in vivo in an eff ort to assess their potential for effi cacious repair of damaged cartilage tissue or diseased joints. Th ese results have proven to be ambiguous and unsatisfactory as a result of low viability and retention of the cells [49]. Due to inconsistency in results and association with hypertrophy and ossifi cation, there are very few human clinical trials investigating heterogeneous bone marrow-derived MSCs as a therapeutic for cartilage repair.
prelimi nary reports have claimed a statistically and clinically signi fi cant improvement in pain experienced by patients post-injury with application of MSCs [51]. Th e application of Chondrogen was well tolerated by recipi ents and superior to currently available, comparable products on the market. With advances in chondropro geni tor cell isolation and culture techniques, products such as this will be improved upon by replacing large numbers of perhaps minimally effi cacious heterogeneous MSCs with low numbers of highly effi cacious chondroprogenitors.
To treat chondral defects, Advanced Technologies and Regenerative Medicine is currently investigating a cartilage autograft implantation system where autologous healthy cartilage is harvested from non-weight bearing regions, minced and re-distributed on a scaff old for implantation. Initial results have been promising, supporting a phase III clinical investigation [52]. If left untreated, lesions such as these regularly result in the onset of OA. It is possible that the reparative cell responsible for the generation of neocartilage in this trial is indeed the re-implanted chondroprogenitor residing on the superfi cial surface of the harvested cartilage tissue. By identifying the reparative cell in this application, a less invasive methodology for reparative cell isolation could be developed, thereby greatly reducing donor site morbidity as well as enhancing the effi cacy of the therapy.
TissueGene, Inc. is currently investigating the safety of intra-articular application of chondrocytes modifi ed virally to overexpress TGF-β1 as a potential treatment for arthritic lesions [53]. Although preliminary results are not yet available, re-creation of the embryonic environment by supplementing cells with TGF-β1 may significantly contribute to cartilage repair over the application of chondrocytes alone, especially after their in vitro expansion and consequential de-diff erentiation.
Th e benefi ts and disadvantages of allogeneic versus autologous cell-based clinical therapies continue to be debated. In a clinical study supported by the National Institutes of Health, this specifi c question will be addressed by comparing autologous and allogenic MSC therapies in the setting of heart failure [54]. Recent phase I clinical administration of allogenic MSC therapies to myocardial infarct patients validated the safety of allogenic cells [54] and autologous cells [55]. However, their long-term survival, engraftment and mechanism of action remain unclear [56].

Conclusion
Several million people worldwide suff er from OA, a chronic, debilitating disorder of the articulating joints. As current clinical interventions do not stimulate the genera tion of a mechanically sound reparative tissue, but focus on the short-term relief of OA symptoms, the fi eld now looks to progenitor cell-based therapies as our future; specifi cally, progenitor cells primed for chondrogenic diff erentiation. Chondroprogenitors residing in the adult have hypothetically been primed for chondrogenic diff erentiation with Wnts, FGFs and TGF-β superfamily members; just as embryonic limb mesenchyme is infl uenced, and, when stimulated, will diff erentiate into high quality articular cartilage. Recent chondrocyte-or MSC-based clinical trials have failed to meet their primary endpoints, often due to our minimal knowledge of their mechanism of action and fate following delivery. Nevertheless, the clinical application of chondroprogenitors for cartilage regeneration may have a bright future with a greater understanding of their embryonic development and the identifi cation of a reliable tissue source. Th e clinical application of a homologous chondroprogenitor population will eliminate the need for heterogeneous cell therapies [56] and result in an effi cacious, minimally invasive approach to articular cartilage repair. Progenitor cells with chondrogenic potential have been isolated from bone marrow, synovium, perichondrium and cartilage itself. The relative ability of these cells to contribute to normal cartilage homeostasis or repair of diseased tissue may be related to their niche or developmental origin.