We previously found that synovial explants[13](bovine synovium model) are able to form more abundant cartilage matrix than isolated synovial cells, even when these were cultured under 3D-conditions (e.g. in alginate[8, 14]). The human synovial explants investigated in this study confirm this finding. And surprisingly, not only the synovium of the normal donors were found here to show a high chondrogenic activity, but also synovium of joints with a major disease, both in young adults (FAI) and in elderly individuals (OA). - The direct use of synovial tissue (autotransplantation) for cartilage repair purposes in a clinical setup has the great advantage to avoid the need for cell isolation and preculturing (as e.g. in approaches based on autologous chondrocyte implantation[15], and thus could be a promising route for the repair of articular cartilage lesions, in particular also in diseased joints.
Possible reasons for the abundant formation of cartilage matrix in these explants, and in particular in the groups treated with BMP-2 alone and in the combined BMP-2/ TGF-β1 stimulation group, may lay in the preserved presence of the physiological pericellular matrix scaffold (ECM) that could significantly facilitate chondrogenic differentiation than other, non-joint associated and/or synthetic scaffolds (such as type I collagen, etc[16, 17]); or by the use of simply naked cells (mesenchymal stem cells (MSCs), chondrocytes, etc.) in the absence of any ECM[18]. And it had been shown before that indeed synovial tissue, under both clinicopathological and experimental conditions, is able to form spontaneously cartilage tissue in the form of tumors[19] or as cartilage-bone-like tissues (osteophytes) [20, 21].
The trend of the synovium of the diseased joints, both in FAI and OA, to form cartilage even more readily, even though not on a significant level, than in the synovium of the normal human adults, is surprising. A speculative interpretation may be that the investigated synovia are from joint pathologies that can be associated with inflammatory episodes, which may be associated with the presence of higher numbers of cells from the monocyte/macrophage lineage, and which, in turn, is well known to be able to facilitate and/or enhance a growth factor mediated tissue differentiation potential [22], such as those with TGF-β1 and/or BMP-2. And the presence of such local activities may indeed facilitate chondrogenic differentiation effects [23] [24].
In a recent study we found[9] that in terminal OA (before total joint replacement) the chondrogenic potential of the synovium is maintained compared to that of the young OA patients, and thus was found to be independent of the age of the patient. However, we did not know what the normal chondrogenic differentiation potential of the synovium of the healthy adult person is, i.e. is it reduced/impaired in both young and old OA patients ? Or does its potential of differentiation remain on a level comparable to that of the healthy synovium of a normal adult person ? In short, the normal healthy reference level was missing. And here we provide for the first time the data for the chondrogenic potential of the synovial membrane of normal human (middle-aged) persons; and surprisingly, this potential is the same as that in young patients with FAI and is also that of elderly patients with severe OA, i.e. it is not reduced under a major joint disease condition. Given now the availability of these baseline data of the normal population we now have available a basis of comparison (and a reference data pool) that allows us to analyze other patient groups suffering from different forms of joint diseases. This is a great advantage since access to normal human data are a generally difficult task; and it was possible here only thanks to the special trauma center cases where appropriate donors could be identified.
From the point of view of clinical practice it is encouraging to see that beside the absence of a decline of the chondrogenic potential of the isolated human synovial MSCs with age [7, 8] there is also an absence of such a decline for the synovium under major joint disease conditions, as found in this study.
For example, in FAI it was found that isolated MSCs from human synovium maintain a high proliferation potential[25, 26]which may vary as a function of local topological origin (niche effect)[25] . However in these studies it was not known if the activity level of this differentiation potential relates to a level of differentiation as it is present in a normal human synovial joint, or not; the appropriate control group was obviously not available. Moreover, the synovial tissue of FAI joints tends to form spontaneously cartilage-like tumors[27-29], which may also have been a confounding factor.
However, it cannot be excluded that some joint diseases are indeed associated with such a decline, in particular when the disease affects primarily the synovium itself, such as in rheumatoid arthritis[30] or in special forms of FAI[31, 32] that differ from the more frequent degenerative forms of FAI[33], or also in OA. The diversity of the local pathology in FAI in the labro-acetabular complex illustrates the high degree of locally-contained niche-type of physiological processes[34]. This may be one of the reasons why surgical treatment of this disease is not a straight-forward simple approach [35], but requires sophisticated evaluations and case-specific measures.
It is conceivable that the great potential of the synovium to undergo chondrogenesis in a FAI affected joint that is suffering simultaneously from significant inflammatory episodes[32], compared to the immunogenically-induced inflamed synovium, as present in various chronic autoimmune diseases and/or rheumatoid arthritis, the situation respecting its chondrogenic potential may be different. This remains to be clarified. Nevertheless, on the basis of the great potential of synovium-derived mesenchymal stem cells (MSCs) to induce the formation of novel tissues in regenerative medicine, in particular in orthopaedics and in rheumatology[36], respecting patients with typically localized or circumscribed articular cartilage defects (such as they occur after trauma, in focal forms of OA, in osteochondritis dissecans, etc.), and such patients in particular could potentially greatly benefit from an autologous synovial MSC-based cartilage repair approach.
OA often is associated with the formation of osteophytes (OSP) that form preferentially in the joint periphery. They often are associated with the periosteum and/or the synovial membrane. Generally the OSPs are localized in the synovial tissue space[37, 38]: And it is the MSCs of these tissues that form the cartilage and bone tissue of the osteophytes[39]. Experimental data point out that the main cell source for their formations lie indeed in the synovium[40, 41], and are associated with the local internal release of TGF-β1 and/or BMP-2[42-44] during the process of the active disease. And even though the physiological role of the OSPs remains unclear (desirable role: stabilizing the joint biomchanically? Undesirable role: limiting the range of motion? [21, 45]), it is well documented that they are formed initially by cartilage tissue formation [42, 43], a cartilage that undergoes a continuos growth, initiating subsequently enchondral ossification processes and the formation of bone tissue[40, 42, 43]. The positive aspects of these data are that the synovium of OA joints maintains also physiologically a strong chondrogenic potential during the disease process, even though it may be of an inflammatory nature. However, the use of this synovium with OSP-activity is associated with the danger that the newly formed chondrocytes continue their physiological downstream differentiation process into terminal hypertrophy, cartilage mineralization and enchondral bone tissue formation, which would be very undesirable activities when using the synovium of OA joints for tissue engineering purposes where a stable articular cartilage tissue must be generated, and the downstream differentiation process needs to be arrested in a pre-terminal phase.
During the ontogenic maturation process the articular cartilage tissue fulfills physiologically a dual function: it acts on one side as articular cartilage proper and provides the practically frictionless movement in the joints between the two adjacent bones and it transfers load from one skeletal element to the another one, but it also acts, on the other hand, during postnatal growth as a superficial growth plate, regulating the bone growth of the epiphysis of long bones, i.e. fulfills a dual functionality at the same time[46]. When reaching adulthood, the growth plate-like activity of the articular cartilage is arrested, and the columnar anisotropic tissue organization is maintained. The articular cartilage radial zone corresponds to the maturation zone and early hypertrophic zone of the growth plate [47] [48], and chondrocytes of these zones cease further differentiation processes into terminal hypertrophy at this point in time of skeletal development. And in this way the extensive terminal hypertrophy process, needed for epiphyseal growth, is arrested as well as are tissue mineralization processes and activities of new bone formation. The radial zone cells of the adult human articular cartilage[49] are characterized by a specific cell size and volume, and this is significantly smaller than the chondrocyte cell volumes attained during terminal hypertrophy. Given this background measurements of cell size and the histochemical staining for calcification were included in this study. And indeed, it was only possible to arrest the chondrocyte differentiation process at the appropriate cell size stage of adult human articular cartilage[50] by the use of the combined stimulation protocol with BMP-2/TGF-beta 1. The effective use of this combined growth factor pair for this purpose had been shown earlier with bovine materials [13], and it seems to work also for the human tissue, as documented here.
These data illustrate also the usefulness of the bovine in vitro model and studies [13, 14] that are able to help to reduce the number of animal experiments in the interest of the 3R philosophy [51].The use of only BMP-2 lead to the formation of hypertrophic chondrocytes in their large terminal size, and eventually matrix mineralization may occur (not yet observed after 6 weeks culturing). Tissue organization of such huge terminal chondrocytes is, on the other hand, also associated with biomechanical tissue properties that are inadequate and insufficient to fulfill the demands of the adult human articular cartilage (much too low volume density of the cartilage matrix, being mainly responsible for the mechanical tissue properties[52]); these implications are another important reason to achieve a controlled arrest of the downstream differentiation of the newly generated cartilage tissue.
The gene expression activity levels demonstrate the general positive correlation of anabolic cartilage gene activities with the formation of the new cartilage tissue. And the catabolic gene activities stayed quite low in all three donor groups. Obviously these gene activity profiles and patterns do not allow to differentiate between different cartilage types and/or degrees of differentiation. A useful example to illustrate this limitation is the activities of the type X collagen gene, associated generally with hypertrophy and terminal differentiation. Even though the three experimental groups were found in this study to be morphologically and histomorphometrically significantly different from one another, the collagen X gene expression levels remained practically on the same relatively low activity level in all three stimulation groups over the whole time period (6 weeks) investigated (no significant differences between them). The morphological, morphometrical and histochemical analyses are thus indispensable to provide the desired information of degree of tissue differentiation and type.
Alkaline Phosphatase is likewise a marker of terminal chondrocyte hypertrophy (and matrix mineralization) [53]. And the activities of this gene remained also on very low levels in all three experimental groups, i.e. not reflecting the tremendous differences in degrees of hypertrophy attained under the three different stimulation protocols. On the other hand, this finding can also be interpreted positively in the sense that further downstream differentiation of the newly generated chondrocytes most likely will not happen so that the unwanted stage of tissue mineralization may never be reached. However, without additional long-term investigations this remains uncertain.
An analysis of the posttranslational expression of gene products would be another useful attribute to possibly provide additional information on the gene products and their available quantities. However, also these, would not be able to provide the very essential information on the structural tissue organization, on the histophysiology and the ultimate result respecting required biomechanical tissue properties. Even though such additional experiments are desirable to complete our picture of the attainable tissue quality, they may be undertaken only now in a more targeted way on the basis of the structural and functional tissue organization data provided in this study.
The really encouraging gene activity results obtained here are those for type II collagen that seemed to be the dominating gene activities in the three donor groups. Also the fact that type I collagen gene activities remained low throughout the investigated time period is quite an encouraging finding from the point of view to ultimately employ this newly generated tissue for clinically applied tissue engineering purposes.
The high gene expression activity levels found for aggrecan, a typical and essential structural and functional component of the cartilage tissue matrix, nicely point in the same direction as the collagen II gene activities. However, these activity levels are not as high as expected as for the type II collagen gene. And again, the fact that the TGF-β1 stimulated groups show gene activity levels for aggrecan similar to those of the other two groups, even though the structural and morphometrical data are fundamentally different, illustrates the limited usefulness of the gene activity analysis if taken on an isolated basis without the different additional structural and quantitative investigations and characterizations. Clearly, the TGF-β1 stimulated groups showed practically no differentiation effects into cartilage, but the aggrecan gene activity levels failed to indicate that.
An encouraging finding was indeed that the catabolic gene activities all remained at low levels throughout the experimental time investigated. This finding can be interpreted in a way that the usual tissue remodeling activities start early on and consistently, even in initial phases, in novel tissue formation stages and under growth. The low catabolic gene activity levels may indicate that the newly formed tissue is a stable product, undergoing physiological remodeling only, and is not driven towards a dominating degradation and tissue destruction pathway.
In conclusion, our data reveal that in young and elderly patients alike, suffering from different major joint diseases, their synovium can be equally well induced to undergo chondrogenesis as in normal individuals. These findings are encouraging to further develop this concept for clinical applications for articular cartilage repair in patients suffering from various joint diseases.