Proteoglycan aggregates in normal and osteoarthritic cartilage

Early investigators believed that articular cartilage was an inert material lacking internal structure. During the last decades, studies have however provided strong evidence that articular cartilage was a metabolically active tissue comprising an elaborated framework of macromolecules. It is also clear that this fascinating tissue is heterogeneous at the structural biochemical and cellular levels.
The finding that native PG aggregates can be divided into two well-characterized families- ie the less saturated or slow sedimenting and the more saturated or fast sedimenting aggregates – point out that the phenomenon of PG aggregation is more variable and complex than initially conceived. Further the exact sites in the HA binding domains of both PG and link protein molecules have not yet been defined and the details of the spatial arrangement of HA, link protein and the G1 domain of PGs in the central backbone of aggregated are not yet known.
The clear-cut difference in the size distribution of the two populations of aggregates is related to their dissimilar content on HA and link protein but not to the size of HA molecules. We do not have a final explanation on how the same HA molecules induce the formation of two well-defined families of PG aggregates; however, we provide strong evidence that link protein not only stabilizes PG aggregates against dissociation but also induces the formation of more saturated and larger PG aggregates. These findings emphasize the necessity to better understand the exact mechanisms governing the synthesis and extracellular assembly of aggregates constituents. This requirement is further compounded by the observation that newly synthesized Pgs are no longer incorporated into larger aggregates in the early stages of experimental canine osteoarthritis (OA).
Native PG aggregates are not saturated with link protein and the molar ration link protein: PG monomer obtained in native aggregates is lower than that observed for the whole cartilage matrix. These observations suggest that link protein might have other functions than those related to PG aggregation.
The smaller aggregates are concentrated in the superficial layers of articular cartilage whereas the larger aggregates are concentrated in the middle and deep layers. Although these differences in the topographical distribution of aggregates reflect topographical changes in the metabolic activity of chondrocytes, it has not yet been clearly established whether these metabolic changes are inherent to the cells or secondary to dissimilar physio-chemical and/or mechanical factors in the different cartilage layers.
The topographical variations in the distribution of both aggregates also suggest that the two types of aggregates may have different functions. The large and link-rich aggregates might play an important role in the material properties of cartilage since they dramatically increase the rheological properties of PGs and they are larger and more abundant in areas of maximum than of minimum contact. On the other hand, the formation of the large ahhregates probably allows the cells to achieve an optimal concentration of PGs in the different topographical areas of cartilage matric since with degree of physiological stress and with tissue depth from the articular surface there is a strong correlation between the total PG content and the relative size and abundance of the large aggregates. How is this achieved is however not clear. The relevance of the large aggregates is further compounded by their disappearance in the early stages of osteoarthritis (OA). This event could contribute significantly to the biochemical and biomechanical alterations of OA cartilage. We do not have any clear explanation for this important phenomenon but we provide indirect evidence that an alteration in link protein synthesis and/or link incorporation into aggregates might be relevant.
Another important event in OA is the progressive reduction in the tissue content of HA, although the HA molecules remaining in the tissue are apparently not degraded. On the other hand, there is also a progressive increase in the tissue content of proteinases which contribute to the degradation of cartilage. As several of these enzymes can cleave PGs in the interglobular domain, the G1 domain remains anchored on the matrix whereas the CS-bearing domain, which provides the main physical properties of PGs, can diffuse out the tissue. The accumulation of G1 domains within OA matrix might interfere with the aggravation of newly synthesized PGs.
In OA tissue there is an imbalance between proteinases and their inhibitor(s). However neither the factors responsible for such imbalance nor the prime movers of proteinase activation have been clerly established. Complex interactions between the networks of cytokines and the network of growth factors might be relevant but require further investigations.
Despite the increase in their catabolic activities chondrocytes also exhibits an increase in their anabolic activities. This might explain at least in part why OA develops slowly. The differences in metabolic activities between normal and OA chondrocytes persist in vivo. Whether these alterations represent a primary change or develop after the cells are stimulated by OA initiating factors in unknown.
Interestingly, the progressive changes in the size distribution of PG aggregates in OA are quite similar to those observed in the growth plate during enchodral calcification although they occur at a quite lower pace in the disease process than in the physiological process. This observation together with the finding of type X collagen in OA chondrocytes might progressively become hypertrophic