Increased apoptosis in human osteoarthritic cartilage corresponds to reduced cell density and expression of caspase-3.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 50, No. 2, February 2004, pp 507–515 DOI 10.1002/art.20020 © 2004, American College of Rheumatology Increased Apoptosis in Human Osteoarthritic Cartilage Corresponds to Reduced Cell Density and Expression of Caspase-3 Mohammed Sharif, Anne Whitehouse, Patrick Sharman, Mark Perry, and Mike Adams analysis of variance showed that the differences between groups for both TUNEL-positive cells and expression of caspase-3 were statistically significant (P < 0.0001). There was a significant positive correlation between TUNEL-positive cells and expression of caspase-3 (r ⴝ 0.654, P< 0.01). Conclusion. The data suggest that apoptosis is increased, on average, 2–4-fold in OA cartilage. Considering that OA develops over many years, such an increase in the rate of apoptosis in the articular cartilage could play an important role in the disease process. Objective. Chondrocyte apoptosis has been described in both human and experimentally induced osteoarthritis (OA), but its importance in the etiopathogenesis of OA is uncertain. The aims of this study were to determine the rate of chondrocyte apoptosis using different methods, and to investigate the relationship between this process and cartilage cellularity, expression of proapoptotic molecules, and expression of antiapoptotic molecules in articular cartilage obtained from patients with OA and from nonarthritic controls. Methods. We examined the extent of apoptosis in OA and nonarthritic control cartilage using expression of caspase-3, an enzyme that mediates the final stage of cell death by apoptosis, as well as the TUNEL method. We used immunohistochemistry to analyze the expression of a panel of proapoptotic and antiapoptotic molecules that regulate apoptosis in articular cartilage, in order to determine whether the rate of apoptosis is associated with the expression of these molecules. Results. The median (range) percentage of TUNEL-positive chondrocytes in knee OA cartilage (n ⴝ 10 specimens), hip OA cartilage (n ⴝ 9), and control cartilage (n ⴝ 7) was 3.11 (1.67–3.67), 1.86 (1.22–2.89), and 0.39 (0.00–1.78), respectively. When all cartilage samples were pooled, apoptosis showed a strong inverse correlation with cellularity (r ⴝ ⴚ0.74, P < 0.0001). The percentage (range) of cells expressing caspase-3 in the 3 groups was 15.70 (7.40–20.50), 15.77 (7.42–20.5), and 7.40 (5.90–8.00), respectively. One-way Osteoarthritis (OA) is the most common joint disease in the elderly population (1), causing significant pain and disability. The etiology of OA is complex and diverse, consisting of a range of biomechanical, biochemical, and genetic factors (2) that converge in a final common pathway (3) characterized by a progressive focal loss of articular cartilage. Normal cartilage homeostasis and structural integrity depend on chondrocytes, which account for only ⬃5% of the total cartilage volume (4). Chondrocytes maintain the dynamic equilibrium between production of the extracellular matrix and its enzymatic degradation. Loss of this balance in favor of catabolic events results in the loss of articular cartilage seen in OA. Thus, chondrocyte viability is essential for maintaining the integrity of articular cartilage, and reduced cellularity (attributable to either necrosis or apoptosis) may predispose the aging individual to matrix degeneration and may be associated with the onset and/or progression of OA. Necrosis is a pathologic form of cell death and is associated with acute cell injury. Apoptosis (programmed cell death) is a physiologic process that is vital in normal development, cell turnover, and removal of damaged or potentially carcinogenic cells (5). Chondrocyte apoptosis has been described in both human (6,7) and experimentally induced (8) OA, but its importance Dr. Sharif is a recipient of a Career Scientist Award from the NHS Executive South West R&D Directorate, Bristol, UK. Mohammed Sharif, PhD, Anne Whitehouse, BSc, Patrick Sharman, BSc, Mark Perry, PhD, Mike Adams, PhD: University of Bristol, Bristol, UK. Address correspondence and reprint requests to Mohammed Sharif, MD, Department of Anatomy, Southwell Street, University of Bristol, Bristol BS2 8EJ, UK. E-mail: Mo.Sharif@bristol.ac.uk. Submitted for publication April 29, 2003; accepted in revised form October 1, 2003. 507 508 SHARIF ET AL in the etiopathogenesis of OA is uncertain. In a recent study of OA cartilage, 6% apoptosis was observed using the TUNEL method (6), while other investigators (9) using a similar method observed ⬍1% chondrocyte death. Previous studies using the TUNEL method found apoptosis rates of ⬍1% to 28% in OA cartilage, and 0% to 7% in nonarthritic cartilage (7,9–11). Not surprisingly, authors of studies in which high levels of apoptosis were found concluded that the process is more important in the etiopathogenesis of OA than did the authors of studies in which a very low percentage of apoptotic chondrocytes in OA was observed (9,12). Recent investigations have questioned the true extent of apoptosis in OA cartilage and the specificity of the widely used TUNEL method itself (9,13). Therefore, we determined the extent of apoptosis in OA cartilage using the TUNEL method, as well as expression of one of the key enzymes, caspase-3, that mediate the final stage of cell death by apoptosis (14). In addition, we used immunohistochemical analysis to examine the balance between panels of proapoptotic and antiapoptotic molecules in articular cartilage obtained from both patients with OA and nonarthritic controls. PATIENTS AND METHODS This study was approved by the regional ethics committee as part of a study titled “Cartilage Destruction in Osteoarthritis.” A written information sheet and a consent form were presented to patients prior to surgery, and only tissue samples obtained from patients who consented were used in the study. Collection and preparation of cartilage samples. Osteoarthritic cartilage on bone was obtained from patients undergoing elective total knee replacement surgery (n ⫽ 10) or total hip replacement surgery (n ⫽ 9) at Avon Orthopaedic Centre, Bristol, UK. Nonarthritic tissue was obtained from the hip joints of 7 patients undergoing surgery for osteoporotic femoral neck fracture at Bristol Royal Infirmary, Bristol, UK. The mean ⫾ SD age of patients was 67 ⫾ 11 years for those with knee OA and 68 ⫾ 8 years for those with hip OA. The mean ⫾ SD age of nonarthritic controls was 68 ⫾ 14 years. The male-to-female ratio in the 3 groups was also very similar. All OA cartilage was obtained ⬃5 mm from the site of the OA lesions. In order to minimize the topographic variations in apoptosis and expression of the proapoptotic and antiapoptotic molecules, control tissue was obtained from sites similar to those from which the OA samples were obtained. Tissue samples were stored under sterile conditions in a refrigerator and processed in the laboratory within 24 hours of surgery. Full-depth samples of cartilage were cut from the underlying subchondral bone, snap frozen in liquid nitrogen, and then stored at ⫺70°C until required for sectioning. Frozen cartilage specimens were embedded in CryoMatrix (Shandon, Pittsburgh, PA), and then consecutive 7-m sections were cut (at ⫺24°C), using a cryostat (Bright Instru- ment Company, Cambridgeshire, UK). For each patient, 3 sections were mounted onto each of 12 labeled SuperFrost Plus microscope slides (BDH Chemicals, Poole, UK). The sections were air dried at room temperature for 2 hours and then wrapped with silica in tin foil and stored at ⫺70°C until required for TUNEL or immunohistochemical staining. One set of slides from each patient was used for toluidine blue staining, for the assessment of general cartilage condition and cell counts. Toluidine blue staining protocol. Toluidine blue stains proteoglycan blue and enables a general assessment of cartilage condition. Toluidine blue staining was performed with a standard protocol, and the stained sections were used for determination of cellularity, empty lacunae counts, and the presence of fibrillation. TUNEL method. TUNEL is a specific immunohistochemical technique that enables sensitive and specific staining of the high concentrations of DNA 3⬘-OH ends that are localized in apoptotic bodies (15,16). Apoptotic nuclei were stained as described in the protocol provided by the manufacturer (ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit; Intergen, Oxford, UK). Immunohistochemical analysis. The slides were defrosted for at least 30 minutes, and immunostaining was performed at room temperature. A standard indirect immunohistochemistry method was used to stain for caspase-3, proapoptotic molecules (p53 and Bag-1), antiapoptotic molecules (Bcl-2, Bcl-x, and Bax), and a chondrocyte proliferation marker, Ki-67. All of these antibodies are monoclonal antibodies to the human protein raised in mice. Ki-67 was purchased from Becton Dickinson (Oxford, UK), and all other monoclonal antibodies were obtained from Neomarker (Fremont, CA). For each patient, 8 slides were used (1 each for the 7 antibodies, and 1 as a negative control). To obtain the dilutions required for optimal staining, each primary antibody was tested at a range of concentrations in phosphate buffered saline. The concentrations (2 g/ml for the majority of the primary antibodies) at which positive cells were sufficiently stained to be clearly visible but at which nonspecific background staining was minimized were used for all subsequent staining of the cartilage sections. Microscopic examination and scoring of cartilage sections. All slides were prepared and examined in a blinded manner, in order to reduce bias during scoring. Each patient was given a code letter (upper case). The different antibodies for each patient were then randomly allocated a code letter (lower case). There was no intentional relationship between the codes allocated for a given antibody in different patients. Slides were observed using a DMRB light microscope (Leica, Milton Keynes, UK) at ⫻20 or ⫻40 magnification. Images of sections were digitally captured using Optimas software (MediaCybernetics, Silver Spring, MD). The toluidine blue– stained sections were scored as follows: for quality of section, 0 (very poor) to 3 (good); for fibrillation, 0 (none) to 4 (fissured); for overall cellularity and cellularity of each cartilage zone, 0 (no cells) to 4 (abundant); and for empty lacunae, as a percentage of all lacunae counted. Following immunohistochemical staining, sections were scored for intensity of staining on a scale of 0–3 (0 ⫽ none, 3 ⫽ strong) and for the percentage of positive cells in each zone. The TUNEL sections INCREASED APOPTOSIS IN OA CARTILAGE 509 Figure 1. A, Femoral head from a patient who underwent hemiathroplasty following osteoporotic subcapital fracture of the femoral neck, representative of the gross appearance of typical nonarthritic control cartilage used in this study. B, Section of cartilage obtained from femoral head shown in A, stained with toluidine blue. C, Femoral head from a patient with osteoarthritis (OA) who underwent total hip replacement, representative of the gross appearance of typical OA cartilage used in this study. D, Section of cartilage obtained from femoral head shown in C, stained with toluidine blue. were scored for the percentage of cells positive for TUNEL in each zone. Validation of counting of positively stained cells and empty lacunae. To ensure that assessment of the proportion of cells staining positive for apoptotic and proliferative markers was consistent with methods used in previous work, a second observer recounted a random selection of 10 slides from 10 different individuals. The proportion of cells staining positive was reassessed, and the first and second readings were compared. Statistical analysis. The data generated in this study are nonparametric. Therefore, the Kruskal-Wallis test was used, followed by Dunn’s test (for the posttest statistic). One-way analysis of variance (ANOVA) was used to investi- gate the differences between the 3 specimen groups for cellularity and the percentage of apoptosis. Spearman’s rank correlation coefficient was calculated for correlations between the percentage of apoptosis and cartilage fibrillation, caspase-3 expression, and proapoptotic and antiapoptotic molecule expression for each zone, as well as for the cartilage as a whole. RESULTS The gross appearance of typical control and OA cartilage specimens and the corresponding histologic sections stained with toluidine blue, as used in the study, are shown in Figure 1. Safranin O staining showed loss 510 SHARIF ET AL Table 1. Fibrillation, cellularity, and empty lacunae in articular cartilage* Cartilage type Fibrillation score (range 0–4) Cellularity score (range 0–3) % empty lacunae Knee osteoarthritis Hip osteoarthritis Nonarthritic control 1.6 ⫾ 0.71 1.60 ⫾ 0.46 0.43 ⫾ 0.53† 1.61 ⫾ 0.18 1.50 ⫾ 0.20 2.86 ⫾ 0.66† 35 ⫾ 20 31 ⫾ 23 25 ⫾ 13 * Values are the mean ⫾ SD. † P ⬍ 0.001 versus knee osteoarthritis and hip osteoarthritis. of glycosaminoglycan from all OA sections compared with controls (data not shown). Validation studies for the toluidine blue and immunohistochemical staining showed minimal intraobserver and interobserver variation. All correlation coefficients for intraobserver and interobserver variation for cell count, percentage of apoptosis, and positive immunohistochemical staining were between 0.89 and 0.96 (P ⬍ 0.0001). However, although both observers found fewer empty lacunae in nonarthritic cartilage compared with OA cartilage, the correlation coefficients for intraobserver and interobserver variations for empty lacunae counts were low. The results for cartilage fibrillation, cellularity, and empty lacunae counts are shown in Table 1. Fibrillation scores for knee and hip OA were very similar to each other but were greater than the score for nonarthritic control cartilage. OA cartilage was hypocellular compared with nonarthritic control cartilage. There appeared to be more empty lacunae in OA cartilage than in nonarthritic control cartilage, but this difference did not reach statistical significance. Apoptosis (according to the TUNEL reaction) was more frequent in knee and hip OA cartilage than in nonarthritic control cartilage (Figure 2). One-way Figure 2. Mean and SEM percentage of apoptosis in cartilage obtained from 10 patients with knee osteoarthritis (OA), 9 patients with hip OA, and 7 nonarthritic (NA) control subjects. ANOVA showed the differences between groups to be highly significant (P ⬍ 0.0001). However, the difference between knee OA and hip OA was not statistically significant (P ⬎ 0.05). There were also significant differences in the percentage of apoptosis in different zones within each group and between groups (Table 2). Within each of the cartilage groups, the percentage of apoptosis was highest in the superficial zone and lowest in the deep zone. In both knee and hip OA cartilage, the percentage of apoptosis was significantly higher (P ⬍ 0.01) in the superficial zone compared with the deep zone. In all 3 zones, knee OA cartilage showed higher cell death by apoptosis compared with hip OA cartilage and nonarthritic control cartilage. Spearman’s correlation coefficient for the correlation between overall cellularity and apoptosis for all cartilage specimens used in the study was ⫺0.74 (95% confidence interval ⫺0.88, ⫺0.48) (P ⬍ 0.0001) (Figure 3). Correlations between cellularity scores and apoptosis within each group of specimens (knee OA, hip OA, and nonarthritis) were ⫺0.537, ⫺0.509, and ⫺0.721, respectively. However, these correlations did not reach statistical significance. A typical OA cartilage section immunostained for caspase-3, and the corresponding negative control (all reagents except the primary antibody to caspase-3) are shown in Figure 4. The percentage of chondrocytes positive for caspase-3 in nonarthritic and OA cartilage and the zonal changes in the expression of the enzyme are shown in Figure 5. Because there were no significant differences between either zonal or overall expression of caspase-3 between knee OA and hip OA cartilage, the 2 OA groups were pooled in Figure 5. There were significantly (P ⬍ 0.001) more cells expressing caspase-3 in both knee and hip OA cartilage compared with nonarthritic control cartilage, and the expression of caspase-3 correlated with the mean percentage of apoptosis in all cartilage groups (r ⫽ 0.654, P ⬍ 0.01). The percentage of chondrocytes staining for caspase-3 in different zones of the cartilage was highest in the superficial zone and lowest in the middle zone in all 3 cartilage groups INCREASED APOPTOSIS IN OA CARTILAGE Table 2. 511 Apoptosis in each zone in articular cartilage* Cartilage zone Cartilage type Superficial Mid Deep Overall Knee osteoarthritis Hip osteoarthritis Nonarthritic control 4.12 (2.67–5.33) 3.34 (1.67–3.67) 1.00 (0.00–3.33)† 3.00 (0.17–4.33) 1.50 (0.67–4.00) 0.17 (0.00–2.00)† 2.34 (0.17–3.00) 0.75 (0.33–1.67) 0.00 (0.00–00)† 3.11 (1.67–3.67) 1.86 (1.22–2.89) 0.39 (0.00–1.78) * Values are the median (range) percent. † P ⬍ 0.001 versus knee osteoarthritis and hip osteoarthritis. (Figure 5). Expression of caspase-3 in the superficial zone was significantly higher than that in the middle and deep zones in all 3 groups (P ⬍ 0.001). However, the difference between the deep and middle zones was significant only in the nonarthritis group (P ⬍ 0.05). The percentage of chondrocytes positive for the proliferative marker Ki-67 was higher in both knee OA and hip OA compared with nonarthritic control cartilage. However, the difference was significant only between knee OA and nonarthritic cartilage (Table 3). There were no significant differences in the expression of Ki-67 between the zones within knee OA, hip OA, or nonarthritic cartilage. Chondrocytes in both nonarthritic and OA cartilage showed positive staining for the proapoptotic molecules (p53 and Bax) and the antiapoptotic molecules (Bcl-2, Bcl-x, and Bag-1), but no significant differences were observed in the expression of these molecules between the cartilage groups (Table 3). However, in all cartilage samples, the percentage of cells expressing p53 was higher in the superficial zone than in the middle and deep zones (P ⬍ 0.01), as shown in Figure 3. Association between cellularity and the mean percentage of apoptosis in articular cartilage obtained from nonarthritic (NA) control subjects, patients with knee osteoarthritis (kOA), and patients with hip OA (hOA). Figure 6. The percentage of cells expressing Bax and Bcl-2 also appeared to be higher in the superficial zone compared with the middle and deep zones, but the differences were not significant. Spearman’s correlations between the expression of these molecules in different Figure 4. A, Representative osteoarthritis cartilage section immunostained for caspase-3. B, Corresponding negative control. (Original magnification ⫻ 200 in A; ⫻ 20 in B.) 512 SHARIF ET AL Figure 5. Percentage of chondrocytes expressing caspase-3 in nonarthritic (NA) and osteoarthritic (OA) cartilage. Bars show the mean and SEM. zones of the cartilage and the presence of apoptosis showed that only Bcl-2 was significantly associated with apoptosis in the superficial zone (r ⫽ ⫺0.526, P ⬍ 0.05). DISCUSSION The data obtained in this study show that apoptosis (TUNEL-positive chondrocytes) in OA cartilage was, on average, ⬃4-fold higher than that in nonarthritic control cartilage. Expression of caspase-3 was ⬃2-fold higher in OA cartilage and correlated with apoptosis as determined by the TUNEL method. Moreover, cellularity was inversely proportional to apoptosis when all specimens were pooled. Taken together, these results suggest that a small but significant proportion of chondrocytes die by apoptosis in OA cartilage compared with normal tissue, leaving the tissue with a reduced density of living cells. To date, there have been several reports on cell death in articular cartilage, and most investigators agree that there is increased cell death by apoptosis in OA cartilage compared with nonarthritic tissue. What is debatable is the extent of apoptosis and its role in the etiopathogenesis of OA. Using in situ TUNEL staining of OA knee cartilage, Blanco and colleagues (6) found, on average, 6% cell death, while similar studies by Hashimoto et al (7) and Heraud et al (11) showed 22.3% and 18%, respectively. These discrepancies may be attributable to topographic variations in apoptosis as well as slight differences in the TUNEL protocols used. Indeed, a higher rate of apoptosis has been reported at the site of OA lesions when compared with a nonlesion site within the same OA knee joint (10). The relatively low apoptosis rates in the current study compared with those in several previous studies could be attributable to the sampling of the cartilage ⬃5 mm from the site of focal lesions. In a recent study (9), following careful characterization of a TUNEL detection kit (Oncor, Gaithersburg, MD) and using fetal growth plate cartilage as positive control, Aigner et al observed a very low rate of apoptosis in OA cartilage. The TUNEL kit used in the current study (ApopTag Plus Peroxidase In Situ Apoptosis Detection) had been characterized by the manufacturer in a manner similar to that in the latter study and was used exactly as suggested in the manufacturer’s protocol. Low rates of apoptosis appear to be more realistic, because a high level of chondrocyte death would rapidly lead to OA cartilage becoming acellular. However, it is not known how much time is required for a chondrocyte to die by apoptosis and for its remains to be cleared from the cartilage matrix. In the current study, expression of caspase-3 was Table 3. Chondrocytes positive for Ki-67, caspase-3, and proapoptotic and antiapoptotic molecules in cartilage* Cartilage type Ki-67 Caspase-3 Proapoptotic p53 Bax Antiapoptotic Bcl-2 Bcl-x Bag-1 Nonarthritic Knee osteoarthritis Hip osteoarthritis 0.00 (0.00–2.78) 7.40 (5.90–8.00) 6.67 (3.33–10.00)† 15.70 (7.40–20.50)† 1.12 (0.00–18.89) 15.77 (7.42–20.5)† 6.17 (3.33–11.67) 5.00 (0.00–8.33) 5.42 (0.00–11.11) 3.33 (0.00–13.33) 8.89 (2.22–11.11) 2.22 (0.00–3.33) 9.17 (6.67–18.89) 5.00 (0.00–7.22) 0.00 (0.00–2.20) 6.67 (4.40–23.33) 0.84 (0.00–10.00) 1.11 (0.00–6.67) 8.34 (3.33–16.67) 0.00 (0.00–3.33) 4.44 (1.11–10.00) * Values are the median (range) percent. † P ⬍ 0.05 versus nonarthritic cartilage. INCREASED APOPTOSIS IN OA CARTILAGE Figure 6. Percentage of chondrocytes expressing p53 in different zones of nonarthritic (NA) and osteoarthritic (OA) cartilage. Bars show the mean and SEM. highest in the superficial zone, where a greater number of TUNEL-positive chondrocytes were also found. There are several possible explanations for increased chondrocyte death by apoptosis in the superficial zones of the articular cartilage. First, cells in the superficial zone undergo greater deformation than those found in deeper zones (17,18). Second, the alignment of the superficial zone collagen network parallel to the articular surface may be less protective, or may cause even further damage, to superficial cells as compared with the perpendicular alignment found in deeper zones (18). As expected, cartilage fibrillation was significantly higher in all OA cartilage compared with nonarthritic control cartilage, but no association was found between the degree of fibrillation and the number of chondrocytes staining positive for apoptosis. Because age-related decreases in cartilage cellularity (19) are associated with an increased frequency of cartilage fibrillation (20), it would seem that reduced cellularity (due primarily to necrosis) is associated with cartilage fibrillation, but that apoptosis is not. However, it is worth noting that in other studies, investigators reported increasing numbers of apoptotic cells with increasing histologic grading (more fibrillation) of the OA cartilage (7,21). Reduced cellularity is a characteristic feature of OA cartilage (22), and apoptosis has been proposed as an underlying cause of the hypocellularity (20,23). In this study, we found that the cellularity scores in both knee and hip OA cartilage were markedly reduced compared with those in nonarthritic control cartilage 513 (Table 1), and that cellularity scores correlated negatively with the percentage of apoptotic chondrocytes, suggesting that reduced cellularity in OA cartilage may be at least partially attributable to cell death by apoptosis. Empty lacunae were counted in order to provide further evidence of hypocellularity in OA cartilage and to determine any association with apoptosis. However, although there appeared to be more empty lacunae in OA cartilage compared with nonarthritic control cartilage, the differences were marginal, and no association was found with apoptosis. Empty lacunae are often difficult to count in thin sections, and the large number found in the control cartilage (on average 25%) suggests that many empty lacunae may represent artifactual loss of cells to the neighboring sections. In an attempt to explore the mechanisms of cell death by apoptosis in human articular cartilage, expression of panels of proapoptotic molecules (p53 and Bax), antiapoptotic molecules (Bcl-2, Bcl-x, and Bag-1), and a chondrocyte proliferative marker (Ki-67) was analyzed by immunohistochemistry. The data demonstrate large variations (as indicated by the wide ranges of values shown in Table 3) in expression of these molecules in all 3 cartilage groups, and there were no significant differences between groups. However, when expression of these molecules was examined in relation to cartilage zones, expression of the proapoptotic molecule p53 was found to be significantly higher in the superficial zone in all 3 groups. This is rather intriguing, because we also found more apoptosis in the superficial zone. Nonetheless, there was no correlation between expression of p53 in the superficial zone and the percentage of apoptosis in this zone. In a previous study (24), Yatsugi et al observed increased expression of p53 in the superficial zone, which correlated positively with apoptosis. In addition, in the current study expression of the other proapoptotic molecule, Bax, in OA cartilage appeared to be considerably lower than that reported by Kim et al (10) (0–13% versus 65%). These apparent differences in the expression of p53 and Bax may be attributable to topographic variations in the cartilage used and/or use of different primary antibodies by the investigators. Among the antiapoptotic molecules, expression of Bcl-2 was generally higher than that of Bag-1 and Bcl-x in all 3 groups (Table 3), and the expression correlated negatively with apoptosis in the superficial zone, suggesting that chondrocytes express more Bcl-2 in an attempt to reduce the level of apoptosis in this zone. In addition to its antiapoptotic function, Bcl-2 is also thought to be involved in maintaining the differentiated phenotype of chondrocytes (25). Higher expression is 514 observed in young cartilage (26), and because OA cartilage is thought to behave in some ways like young cartilage (27), up-regulation of Bcl-2 may be associated with cartilage repair in OA. However, it is worth emphasizing that the interaction between the proapoptotic molecules and various effector enzymes such as caspase3 that lead to cell death by apoptosis is very complex. In the present study, large variations in the expression of both proapoptotic and antiapoptotic molecules in a relatively small number of cartilage specimens prevent us from drawing any conclusions about the role of these molecules in controlling apoptosis in human articular cartilage. The percentage of apoptotic cells increases with age in the articular cartilage of skeletally mature animals (19), but no sex-related differences have been reported in animal or human tissue. The present study compares apoptosis in cartilage from OA joints and nonarthritic joints matched for sex and age. However, it was necessary for us to use nonarthritic hip cartilage as the control for both OA knee and OA hip cartilage. Therefore, some differences observed between OA knee and nonarthritic hip cartilage may be explained by underlying differences between different joints rather than the disease process itself. Nevertheless, in this study the data showed no significant differences between knee OA and hip OA cartilage. Indeed, for some of the immunohistochemical analyses (e.g., caspase-3), the data for knee and hip OA were very similar. Another potential problem is that nonarthritic tissue from the hip was obtained from patients with osteoporotic femoral neck fractures. This is the usual control used in this type of research, but it remains possible that osteoporotic changes in bone could influence the overlying articular cartilage. Recent studies have shown a positive correlation between the degree of severity of OA and the number of apoptotic chondrocytes in both experimentally induced OA in rabbit cartilage (8) and human OA cartilage (7). Cell death by apoptosis in articular cartilage may contribute to cartilage damage in and/or the pathogenesis of OA via several mechanisms. First, an increased rate of apoptosis in cartilage could lead to a reduction in cartilage cellularity, and, thus, a reduced ability of the cartilage cells to replace any damaged cartilage in the early stages of OA. Second, accumulation of apoptotic bodies and their removal from the tissue can cause further degradation of the cartilage. Cartilage is avascular and contains no mononuclear phagocytes capable of removing apoptotic bodies. However, it is likely that the chondrocytes themselves are phagocytosing the apoptotic bodies, because chondrocytes with condensed mate- SHARIF ET AL rial in their cytoplasm have been observed by electron microscopy (10). Apoptotic bodies in cartilage are structurally (28) and functionally (21) similar to matrix vesicles associated with hydroxyapatite crystal deposition, and they are able to precipitate calcium from solution in similar amounts to matrix vesicles (21) and therefore could be responsible for abnormal cartilage calcification in aging and OA. Also, apoptotic bodies could release their contents (including proteases) into the matrix, promoting enzymatic degradation. Destruction of matrix components, especially in the pericellular region, may affect cell–matrix interactions, so that further apoptosis is triggered, and a vicious cycle is established. In conclusion, the data presented in this report show that at any point in time, 2–3% of chondrocytes in OA cartilage are dying by apoptosis, compared with 0.6% in nonarthritic cartilage, and that an increased level of apoptosis corresponds to reduced cellularity in the cartilage matrix. The levels of apoptosis observed in the current study are lower than those found in several previous studies (6,7,11) but are consistent with results of the recent (meticulous) study by Aigner et al (9). OA usually develops over many years and in most patients progresses rather slowly, so a low level of cell death by apoptosis could play an important role in the disease process. REFERENCES 1. Hamerman D. Aging and osteoarthritis: basic mechanisms. J Am Geriatr Soc 1993;41:760–70. 2. Goldring MB. The role of the chondrocyte in osteoarthritis. Arthritis Rheum 2000;43:1916–26. 3. Westacott CI, Sharif M. Cytokines in osteoarthritis: mediators or markers of joint destruction? Semin Arthritis Rheum 1996;25: 254–72. 4. Kuettner KE. Biochemistry of articular cartilage in health and disease. Clin Biochem 1992;25:155–63. 5. Kerr JF, Wyllie AH, Currie AR. Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 1972;26:239–57. 6. Blanco FJ, Guitian R, Vazquez-Martul E, de Toro FJ, Galdo F. Osteoarthritis chondrocytes die by apoptosis: a possible pathway for osteoarthritis pathology. Arthritis Rheum 1998;41:284–9. 7. Hashimoto S, Ochs RL, Komiya S, Lotz M. Linkage of chondrocyte apoptosis and cartilage degradation in human osteoarthritis. Arthritis Rheum 1998;41:1632–8. 8. Hashimoto S, Takahashi K, Amiel D, Coutts RD, Lotz M. Chondrocyte apoptosis and nitric oxide production during experimentally induced osteoarthritis. Arthritis Rheum 1998;41: 1266–74. 9. Aigner T, Hemmel M, Neureiter D, Gebhard PM, Zeiler G, Kirchner T, et al. Apoptotic cell death is not a widespread phenomenon in normal aging and osteoarthritic human articular knee cartilage: a study of proliferation, programmed cell death INCREASED APOPTOSIS IN OA CARTILAGE 10. 11. 12. 13. 14. 15. 16. 17. 18. (apoptosis), and viability of chondrocytes in normal and osteoarthritic human knee cartilage. Arthritis Rheum 2001;44:1304–12. Kim HA, Lee YJ, Seong SC, Choe KW, Song YW. Apoptotic chondrocyte death in human osteoarthritis. J Rheumatol 2000;27: 455–62. Heraud F, Heraud A, Harmand MF. Apoptosis in normal and osteoarthritic human articular cartilage. Ann Rheum Dis 2000;59: 959–65. Kouri JB, Jimenez SA, Quintero M, Chico A. Ultrastructural study of chondrocytes from fibrillated and non-fibrillated human osteoarthritic cartilage. Osteoarthritis Cartilage 1996;4:111–25. Kouri JB, Aguilera JM, Reyes J, Lozoya KA, Gonzalez S. Apoptotic chondrocytes from osteoarthrotic human articular cartilage and abnormal calcification of subchondral bone. J Rheumatol 2000;27:1005–19. Janicke RU, Sprengart ML, Wati MR, Porter AG. Caspase-3 is required for DNA fragmentation and morphological changes associated with apoptosis. J Biol Chem 1998;273:9357–60. Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 1992;119:493–501. Gorczyca W, Bruno S, Melamed MR, Darzynkiewicz Z. Cell cycle-related expression of p120 nucleolar antigen in normal human lymphocytes and in cells of HL-60 and MOLT-4 leukemic lines: effects of methotrexate, camptothecin, and teniposide. Cancer Res 1992;52:3491–4. Guilak F, Ratcliffe A, Mow VC. Chondrocyte deformation and local tissue strain in articular cartilage: a confocal microscopy study. J Orthop Res 1995;13:410–21. Clements KM, Bee ZC, Crossingham GV, Adams MA, Sharif M. How severe must repetitive loading be to kill chondrocytes in articular cartilage? Osteoarthritis Cartilage 2001;9:499–507. 515 19. Adams CS, Horton WE Jr. Chondrocyte apoptosis increases with age in the articular cartilage of adult animals. Anat Rec 1998;250: 418–25. 20. Stockwell RA. The interrelationship of cell density and cartilage thickness in mammalian articular cartilage. J Anat 1971;109: 411–21. 21. Hashimoto S, Ochs RL, Rosen F, Quach J, McCabe G, Solan J, et al. Chondrocyte-derived apoptotic bodies and calcification of articular cartilage. Proc Natl Acad Sci U S A 1998;95:3094–9. 22. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and metabolic abnormalities in articular cartilage from osteoarthritic human hips. II. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971;53: 523–37. 23. Vignon E, Arlot M, Patricot LM, Vignon G. The cell density of human femoral head cartilage. Clin Orthop 1976;121:303–8. 24. Yatsugi N, Tsukazaki T, Osaki M, Koji T, Yamashita S, Shindo H. Apoptosis of articular chondrocytes in rheumatoid arthritis and osteoarthritis: correlation of apoptosis with degree of cartilage destruction and expression of apoptosis-related proteins of p53 and c-myc. J Orthop Sci 2000;5:150–6. 25. Feng L, Balakir R, Precht P, Horton WE Jr. Bcl-2 regulates chondrocyte morphology and aggrecan gene expression independent of caspase activation and full apoptosis. J Cell Biochem 1999;74:576–86. 26. Kim HA, Song YW. Apoptotic chondrocyte death in rheumatoid arthritis. Arthritis Rheum 1999;42:1528–37. 27. Pfander D, Cramer T, Weseloh G, Pullig O, Schuppan D, Bauer M, et al. Hepatocyte growth factor in human osteoarthritic cartilage. Osteoarthritis Cartilage 1999;7:548–59. 28. Blanco FJ, Ochs RL, Schwarz H, Lotz M. Chondrocyte apoptosis induced by nitric oxide. Am J Pathol 1995;146:75–85.