Expression of collagen type I, II and III in loose body of osteoarthritis

J Orthop Sci (2000) 5:288–293

Expression of collagen type I, II and III in loose body of osteoarthritis Ming Pei1, Changlong Yu2, and Mianyu Qu2 1

Division of Health Sciences and Technology, Massachusetts Institute of Technology, E25-342, 45 Carleton Street, Cambridge, MA 02139, USA 2 Institute of Sports Medicine, The Third School of Beijing Medical University, Beijing, P.R. China

Abstract: The expression of collagen type I, II, and III was investigated to evaluate phenotypic change in chondrocytes in loose bodies related to osteoarthritis. We assessed collagen type I, II, and III production in loose bodies from knee joints of ten osteoarthritic patients, using an immunohistochemical method with monoclonal antibodies. Collagen type III expression was identified in all ten loose bodies and was mainly located in cartilage, including chondrocytes and matrices, as well as in a layer of fibroid tissue on the surface. No positive signal for collagen type III was observed in necrotic osteocytes. There was weakly positive staining for collagen type I in chondrocytes. No positive staining for collagen type II could be seen in the cartilage of loose bodies. Cartilage from the non-osteoarthritic knee joints of four people was negative for the expression of collagen type I and III, and positive for the expression of collagen type II. Collagen type I and III expression suggested the dedifferentiation status of chondrocytes in loose bodies. Key words: collagen type I, II, and III, loose body, osteoarthritis, dedifferentiation

Introduction From the studies of cartilage development in vitro and in situ, it has become evident that several distinct chondrocyte phenotypes can be distinguished on the basis of the collagen type expression: (a) chondrocyte precursor cells (eg, limb bud or somite mesenchyme cells synthesizing type I collagen),7,13,15 (b) differentiated hyaline chondrocytes (eg, embryonic cartilage anlagen,

Offprint requests to: M. Pei Received for publication on May 26, 1999; accepted on Oct. 20, 1999

resting zone of fetal epiphyseal cartilage, or articular cartilage expressing collagen types II, IX, XI, and VI),8 and (c) hypertrophic chondrocytes in the growth zone of fetal epiphyseal or rib cartilage, synthesizing predominantly type X collagen.2 A fourth chondrocyte phenotype resembling the chondrocyte precursor cells can be induced in vitro by culture and by repeated passage of chondrocytes in a monolayer, or stimulated by 5-bromouridine deoxyriboze (BUdR), embryo extract, fibronectin, or retinoic acid. This modulation of the collagen expression pattern has been compared with so-called “dedifferentiation” and involves the transition to a fibroblastic cell shape, the formation of focal contacts and stress fibers, and the synthesis of types I, III, and V collagen. Many studies are available on the expression of these collagens by articular chondrocytes in vitro and by normal fetal and adult cartilage in vivo.12,14 Several studies described the stimulation of collagen synthesis as the response to cartilage degradation in experimentally induced osteoarthritic models3,4 and in human osteoarthritis.1,6 However, there is still a controversy on the types of newly formed collagen. Some authors reported that the newly synthesized collagen seemed to be predominantly type II collagen,3,6 as well as type IX, XI, and VI collagen.11 Other authors reported that in osteoarthritic cartilage, type I and III collagen were detected by immunohistochemical and biochemical methods,10 indicating “dedifferentiation” of chondrocytes similar to the findings in chondrocytes in culture. These results could not be confirmed by several other biochemical studies, which were not able to detect any type I collagen in osteoarthritic cartilage, except in fibrocartilage.6 To date, there is no literature reporting the phenotype of chondrocytes in loose bodies from osteoarthritic joints. In this study, we investigated the expression of collagen types I, II, and III in chondrocytes of loose bodies from osteoarthritic joints

M. Pei et al.: Collagen expression in OA loose bodies

289

at the protein level to check for phenotype change in chondrocytes. Materials and methods Reagents Two monoclonal antibodies (MoAbs), directed against collagen type I and III, respectively, were made in our own laboratory (data not shown). A Moab directed against collagen II, the second antibody (biotinylated goat anti-mouse IgG), and avidine-biotin-peroxidase complex were purchased form NewMarker (Union City, CA, USA). Other chemicals were obtained from Sino-American Biotechnology (Beijing, China). Specimen preparation Loose bodies were harvested during operations carried out for ten osteoarthritic patients who had knee-joint locking symptoms. The patients were aged from 35 to 46 years (average, 39 years). There were 7 men and 3 women. All loose bodies had existed within the joint cavities for at least 6 months, according to the clinical data and X-ray films. Fresh fracture fragments and loose bodies which were attached to other tissues, such as synovium or joint surface, were excluded from our study. Four specimens of normal cartilage, used as controls, were obtained from donors who were killed in traffic accidents. Specimens were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS), decalcified with 0.3 M ethylenediaminetetra acetic acid (EDTA, pH 7.2), dehydrated, and embedded in paraffin; and sections 5µm in thickness were made. For all specimens, apart from the immunochemical staining, hematoxylin and toluidine blue staining were performed on parallel sections.

Fig. 1. Partial fibrosis occurred in the exterior layer of the cartilage of loose bodies, chondrocytes had disappeared in necrotic areas, and, in certain areas, proliferating chondrocytes were also found. H&E, 3100

Fig. 2. Hyperplastic fibrocartilage-like tissue (arrow) existed on the opposite side to primary hyaline-like cartilage, next to necrotic marrow cavity. H&E, 340

Results Immunostaining for collagen type I, II, and III Deparaffinized sections were treated with 3% hydrogen peroxide/methanol for 10 min to block endogenous peroxidase, and then incubated with testicular hyaluronidase (2 mg/ml in PBS, pH 5) for 30 min at room temperature. After treatment with 10% normal goat serum for 10 min, the sections were consecutively incubated with primary antibody, biotinylated goat anti-mouse IgG, and avidin-peroxidase complex. Each incubation step was performed for 60 min. Then sections were stained with 0.05% diaminobenzidine (DAB) in PBS with 0.009% hydrogen peroxide. Nuclei were counterstained with hematoxylin. The primary antibody was replaced with normal serum as a negative control.

Histopathological characteristics of loose bodies Loose bodies were ivory-white by the naked eye and of varied size. Some displayed a smooth appearance, which was designated as ovoid configuration. Others showed a bumpy gross appearance, which was designated as the mulberry configuration. H&E staining showed that partial fibrosis had occurred in the exterior layer of the cartilage of the loose bodies, chondrocytes had disappeared in necrotic areas, and in some areas, some proliferating chondrocytes were also found (Fig. 1). In some specimens, hyperplastic fibrocartilage-like tissue existed on the opposite side to primary hyalinelike cartilage, next to necrotic bone marrow cavity (Fig. 2). The majority of osteocytes in the osseous tissue were

290

M. Pei et al.: Collagen expression in OA loose bodies

Fig. 3. Most of the osteocytes in the osseous tissue were dead, and this usually occurred in the layers of bone (arrow) separated by calcified cartilage. H&E, 340

Fig. 5. The surviving hyperplastic chondrocytes showed an amaranth appearance in the cytoplasm. Toluidine blue, 340

Fig. 4. Bone tissue in the loose bodies was negative for toluidine blue staining (arrow). Toluidine blue, 340

Fig. 6. Areas of calcified (arrow) and uncalcified cartilage appeared by turn. Toluidine blue, 340

dead, and this usually occurred in the layer of bone separated from the surface by newer formed calcified cartilage (Fig. 3). Bone and fibrotic tissue in the loose bodies were negative for toluidine blue staining (Fig. 4), and surviving hyperplastic chondrocytes showed as amaranth appearance in cytoplasm (Fig. 5). Chondrocyte death was also noted in the calcified tissue, where the calcification occurred in the matrix around the lacunae of the dead chondrocytes. It seemed that the areas of calcified and uncalcified cartilage appeared by turn (Fig. 6).

chondrocytes and matrices of the loose bodies (Fig. 7), as well as in a layer of deteriorated fibroid tissue on the surface (Fig. 8). No positive signal was observed in necrotic osseous tissue. As for collagen type I expression, there was weak positive staining in the chondrocytes (Fig. 9). No collagen type II positive signal was found in the cartilage of the loose bodies. Cartilage from the control group was negative for the expression of collagen type I and III, and positive for the expression of collagen type II (Fig. 10).

Expression of collagen type I, II, and III in the chondrocytes of loose bodies Collagen type III expression was identified in all loose bodies examined and was primarily located in the

Discussion After loose bodies had remained free within a joint for at least 6 months (as determined by clinical history), proliferative change was then observed on the outer

M. Pei et al.: Collagen expression in OA loose bodies

291

Fig. 7. Collagen type III expression was detected and was primarily located in the chondrocytes and matrices of loose bodies. Diamino benzidine (DAB), 3100

Fig. 9. There was weak positive staining for collagen I in the cartilage of loose bodies. DAB, 3100

Fig. 8. Collagen type III expression was found in a layer of deteriorated fibroid tissue on the surface. DAB, 3100

Fig. 10. Collagen type II showed positivity in control articular cartilage. DAB, 3100

surfaces of most specimens.9 Our study showed that the outer surface of the loose bodies was a layer of degenerated fibroid tissue which, presumably, resulted from the denatured superficial cartilage. When viewed grossly, many old loose bodies appeared to have a surface with pits and crevices. Some authors have pointed out that this type of surface was due to osteoclastic remodeling, and it was a phenomenon that occurred in the majority of loose bodies of all etiologies.9 Nutrition for the viable osteoclasts on the surfaces of the specimens was probably derived from the synovial fluid.9 Osteoclastic activity, like proliferative activity, was not uniform on all parts of the surface of any single loose body.14 However, in our study, we failed to confirm any osteoclastic activity on the surface of loose bodies. We presume that the difference in the surface of loose bodies is probably caused by stress, lack of nutrition, and cell phenotype

changes. It is hard to believe that osteoclasts would exist on the surface of loose bodies and remodel the surface without any reliable nutrition supply. Within the loose bodies the fat and hematopoietic cells of bone marrow, the blood vessels, and the osteoblasts, osteoclasts, and osteocytes are always necrotic because of the lack of blood supply. However, we observed that viable proliferating chondrocytes existed on the undersurface close to necrotic bone marrow spaces. This chondrous tissue replacement extended into the bone marrow spaces for varying depths from the surface, but malnutrition in these proliferating chondrocytes may have limited this activity, so this phenomenon was mainly noted in the peripheral zone. In older specimens, some chondrocytes in cartilage extended into bone marrow spaces, lost their nutritional support, and had died. The cartilage matrix then underwent calcification. This phenomenon

292

indicated that chondrocytes may still be able to survive and keep their proliferative capacity in loose bodies because of the nutrition supplied by synovial fluid. Although theoretically, osseous tissue within a loose body is always necrotic, in our experiment, we observed that some fractions of osseous tissue were still alive, and this phenomenon possibly resulted either from the short period since the loose bodies had become loose, or that endochondral bone formation could occur directly without blood supply. Our study showed that not only cartilage but also a layer of new bone could survive in loose bodies without any blood supply. As to the outer layer of tissue of the loose bodies, we observed fibrillation and fraying of the unmineralized cartilage layer on the surface. It was felt that such change was due to the loose body bouncing inside the joint and being trapped between the joint surfaces. The cartilage surface would then degenerate because of the repeated violence, and the matrix would be degraded and fibers exposed, so that the cartilage looked like a layer of fibroid tissue. Another possibility is that, as loose bodies are within as osteoarthritic environment, it is inevitable that the chondrocytes change their phenotype as do the chondrocytes in osteoarthritis cartilage; the matrix and collagen fibers synthesized by chondrocytes would no longer be normal and would become fragile and readily deteriorate. Our experiment showed that collagen type III expression was detected in all loose bodies examined and was primarily located in the chondrocytes and matrices, as well as in a layer of fibroid tissue on the surface. And there was weak positive staining for collagen type I in the chondrocytes. No collagen type II positive signal was found. Cartilage from controls was negative for the expression of collagen type I and III, and positive for the expression of collagen type II. Therefore, it is reasonable to say that chondrocytes in loose bodies have changed their phenotype, and they should no longer be regarded as chondrocytes from hyaline cartilage. The switch of phenotypes in chondrocytes is a common phenomenon in osteoarthritic cartilage. Our previous study has shown that, in experimental osteoarthritis, there was a switch of collagen types, from type II to III. In the early stage of osteoarthritis, fibrosis on the surface of cartilage was the main manifestation. Under a scanning electron microscope, many collagen fibrils, of varied thickness, could be seen accumulated on the surface of the articular cartilage. Some had seasonal cross striations and had the characteristics of collagen type I and III.5 Immunohistochemistry and in situ hybridization also showed that collagen type I and III occurred in the matrix around the chondrocytes in osteoarthritic cartilage. In the early stage of degradation of articular cartilage, only a few chondrocytes expressed

M. Pei et al.: Collagen expression in OA loose bodies

procollagen α1 (III) mRNA, and simultaneously collagen type III appeared in the matrix. In the later stage of osteoarthritis, the content of collagen type III in cartilage increases. With the process of pathological changes, the content of collagen type I increases.2 This is an important embodiment of dedifferentiation of chondrocytes in osteoarthritic cartilage. A possible mechanism is due to the high homology between the gene and amino acid sequences of various kinds of collagen (especially fibrous collagen), and the loss of control of expression of collagen by stimulating factors.16 When articular chondrocytes were cultured in vitro, the chondrocytes secreted collagen I and III to a greater extent with increased numbers of passages. In conditions of monolayer growth, as obvious change occurs in the third generation, and the differential phenotype of chondrocytes (collagen II) is lost, mainly in the fifth generation, and the levels of collagen I and III increase remarkably. The mechanism of the phenotype switching remains unknown, but this phenomenon is similar to the dedifferentiation of chondrocytes in osteoarthritic cartilage.

References 1. Adam M, Deyl Z. Altered expression of collagen phenotype in osteoarthrosis. Clin Chim Acta 1983;133:25–32. 2. Eerola I, Salminen H, Lammi P, et al. Type X collagen, a natural component of mouse articular cartilage. Arthritis Rheum 1988; 41:1287–95. 3. Eyre DR, McDevitt CA, Billingham MEJ, et al. Biosynthesis of collagen and other matrix proteins by articular cartilage in experimental osteoarthrosis. Biochem J 1980;188:823–37. 4. Floman Y, Eyre DR, Glimcher M. Induction of osteoarthrosis in the rabbit knee joint: biochemical study on the articular cartilage. Clin Orthop 1980;147:278–86. 5. Gattone VH 2nd, Saul FP, O’Connor BL, et al. Scanning electron microscopic study of age-related surface changes in rat femoral head articular cartilage. J Submicrosc Cytol 1982;14:99– 106. 6. Goldwasser M, Astley T, van der Rest M, et al. Analysis of the type of collagen present in osteoarthritic human cartilage. Clin Orthop 1982;167:296–302. 7. Linsenmayer TF, Toole BP, Trelatrad RL. Temporal and spatial transitions in collagen types during embryonic chick limb development. Dev Biol 1973;35:232–9. 8. Mayne R. Cartilage collagens: what is their function, and are they involved in articular disease? Arthritis Rheum 1989;32:241–6. 9. Milgram JW. The development of loose bodies in human joints. Clin Orthop 1976;124:292–303. 10. Nimni M, Deshmukh K. Differences in collagen metabolism between normal and osteoarthritic human articular cartilage. Science 1973;181:751–2. 11. Ronzière MC, Ricard-Blum S, Tiollier J, et al. Comparative analysis of collagens solubilized from human foetal, and normal and osteoarthritic adult articular cartilage, with emphasis on type VI collagen. Biochim Biophys Acta 1990;1038:222–30. 12. Sandberg M, Vuorio E. Localization of types I, II, and III collagen mRNAs in developing human skeletal tissues by in situ hybridization. J Cell Biol 1987;104:1077–84.

M. Pei et al.: Collagen expression in OA loose bodies 13. Von der Mark K, von der Mark H. Immunological and biochemical studies of collagen type transition during in vitro chondrogenesis of chick limb mesodermal cells. J Cell Biol 1977;73:736–47. 14. Von der Mark K, von der Mark H. The role of three genetically distinct collagen types in endochondral ossification and calcification of cartilage. J Bone Joint Surg Br 1977;59:458–64.

293 15. Von der Mark K, von der Mark H, Gay S. Study of differential collagen synthesis during development of the chick embryo by immunofluorescence. Dev Biol 1976;53:153–70. 16. Young RD, Lawrence PA, Duance VC, et al. Immunolocalization of type III collagen in human articular cartilage prepared by highpressure cryofixation, freeze-substitution, and low-temperature embedding. J Histochem Cytochem 1995;43:421–7.