Closing ranks: The chondrocytes' move in bad times or biomechanical job sharing?

Cell clustering is an architectural principle that governs many organs. Indeed, it is so common that we tend to overlook its presence after we have had our first experiences with histology in microanatomy classes. Embryogenesis is driven by cell-sorting mechanisms that lead to temporal clustering and secondary segregation as an underlying principle of organ formation. Muscle fibers are clusters, as can be easily understood by looking at cross-sectional views. Tendons as well as nerves, pancreatic islets, renal tubuli, adrenal glands, and so on all are functionally grouped cell aggregates. In contrast, we nurture the legend of the lonesome cowboy called chondrocyte, because our impression from vertical sections suggests the singular appearance to be more or less characteristic for healthy articular human cartilage. The situation is different with osteoarthritis (OA) cartilage, in which the standard type of vertical tissue sections reveal the occurrence of clonal cell clusters as one of the histologic hallmarks for OA pathology.

It is noteworthy that in all these observations, the key is the vertical section. Switch to sectioning parallel to the surface, and the rules change dramatically: clusters everywhere. The first observations were made in the late 1990s and early 2000s by Barbara Schumacher, then at the Department of Biochemistry at Rush University in Chicago. The data made it clear that the clustered organization of the superficial chondrocytes is related to the joint (knee versus ankle) (1). At that time, Bernd Rolauffs, the first author of the study presented in the current issue of Arthritis & Rheumatism (2), was a hopeful young orthopedic postdoctoral fellow eagerly learning everything Barbara Schumacher could teach, including the new findings on spatial cellular organization in cartilage.

Barbara Schumacher moved to Bob Sah's laboratory in San Diego, where the study of spatial chondrocyte organization was brought to a new precision level, clearly influenced by the engineering approach fostered there. The new constellation of investigators published data in 2005 and 2007, including a neighborhood analysis and a description of the angular orientation in bovine articular cartilage from fetus to adult animal (3). Even though some of the figures presented in the 2007 report bear some resemblance to those in a 2005 publication, a clear advancement in the understanding of chondrocyte organizations emerges from those contributions. In the meantime, Bernd Rolauffs brought his knowledge to Alan Grodzinsky at the Massachusetts Institute of Technology in Boston, who happened to have been the PhD supervisor for Bob Sah. The encounter resulted in (among other things, of course) the experimental study presented here. In the meantime, other investigators were busy as well. Jones et al (4) showed the enormous disturbance of the original pattern in advanced stages of OA after publishing initial data sets on this issue.

The new contribution by Rolauffs et al introduces a mathematical algorithm with which to quantitate the cell patterning on the articular surface and presents first evidence for a regrouping of superficial chondrocytes in very early preclinical stages of cartilage degeneration (2). In their previous work analyzing intact cartilage as well as macroscopically intact regions of joint surfaces with focal OA, Rolauffs and colleagues were able to recognize previously unappreciated patterns of cellular order within the superficial chondrocyte organization (5). Interestingly, only one of these patterns dominates each large human joint surface, suggesting biomechanical involvement in their genesis. However, the present study extends beyond this recent work: changes in the spatial order can be assessed by recording the relative angles between 2 neighboring chondrocytes and a reference line. By developing this novel spatial measurement, Rolauffs et al succeed in demonstrating differences between different species. Most importantly, they unravel the finding that the onset and the different stages of OA correlate with characteristic changes in the spatial order of the cells in the cartilage surface.

An innocent outsider may conclude that regrouping of cells might also be generated by cell migration. The degree of biologic truth in such a blasphemous idea remains an unanswered question, because in mainstream science chondrocytes are not considered to be migratory cells. However, there are some hypotheses regarding the modes of cartilage regeneration that require migratory activities for chondrocytes. Peter Simkin (6) proposed the existence of a continuous supply of fresh cells recruited from lateral synovial tissue (i.e., stem cells). Such cells may migrate via the surface and sink into the deeper layers by any kind of active or passive motility. The presence of apoptotic bodies or free DNA in the cartilage extracellular matrix has not been demonstrated thus far, therefore rendering this hypothesis quite unlikely despite its alluring picture.

A different story is presented by the classic cell clusters of advanced OA. There is no doubt that those clusters originate from cell divisions within one chondron, and that these structures contain cells that undergo hypertrophy and apoptosis (7). Thus, the clusters described by Rolauffs et al may distinguish themselves from these pathologic features by harboring viable cells with no features typical for the late stages of OA (2). In yet another review, Teresa Morales points to the primary cilium and to cytoplasmic extension (up to a length of 80 μm) of chondrocytes within the adult cartilage (8). Although these structures indicate some kind of anchored cytoplasmic mobility, they may serve different functions such as mechanical force sensing and matrix remodeling rather than representing still images of migratory activities. Morales rightly points to the fact that the chondrocyte migration out of their matrix observed for explant cultures in vitro may not reflect in vivo behavior, because defects do not heal in vivo.

Nonetheless, the newly described spatial remodeling processes may be used as a diagnostic tool to investigate the early onset of OA or other remodeling processes. Furthermore, this technique may serve to distinguish between OA-affected areas (with their classic clusters) and healthy areas capable of regenerative and adaptive remodeling. Exact analyses of the cellular positioning are a prerequisite for a diagnostic concept. However, current imaging techniques allow such investigations only in vitro, deploying confocal (or similar) fluorescence microscopy; confocal laser arthroscopy or future noninvasive imaging techniques such as high-end magnetic resonance imaging may open opportunities. The clinical applicability may therefore be a question of technical advancement in noninvasive image resolution.

Until the advent of those future imaging technologies in the common laboratory or hospital setting, the study by Rolauffs et al suggests exploring the underlying mechanisms of ordered and unordered cellular spatial placements by a combination of approaches. These methods may include biochemical and biomechanical studies on tissue maturation, at the onset of OA and after incidences of injury. In vitro settings with bioreactors or animal studies may be suitable to obtain data from microscopy and analytical mathematics that allow the development of appropriate models for in vivo cell regrouping. The most pressing issue will be to resolve the chicken versus egg question: is cellular regrouping a cause or a consequence of OA pathology?

AUTHOR CONTRIBUTIONS

Dr. Mollenhauer drafted the article, revised it critically for important intellectual content, and approved the final version to be published.