Calcium pyrophosphate dihydrate (CPPD) crystal deposition disease is caused by CPPD crystals formed in the extracellular matrix of articular cartilage (1). These crystals may cause acute attacks of pseudogout, but more importantly, CPPD deposits are intimately associated with and may cause osteoarthritis (OA) of weight-bearing joints (2). CPPD crystals initiate or amplify cartilage destruction by stimulating mitogenesis of synovial lining cells as well as synthesis and secretion of proteases, prostanoids, and cytokines that have been implicated in cartilage matrix degeneration (3). Prevention or reversal of CPPD deposition would be expected to have a salutary effect on the degree of degenerative joint disease associated with CPPD. However, current understanding of the mechanisms involved in CPPD crystal formation is insufficient to devise effective therapeutic or prophylactic interventions for the associated degenerative arthritis.
Local accumulation of excess extracellular inorganic pyrophosphate (ePPi), the anionic component of CPPD crystals, promotes CPPD crystal formation. Previous studies indicate that synovial fluid PPi concentrations are consistently elevated in patients with CPPD deposition (4–10). Articular cartilage chondrocytes uniquely elaborate large amounts of ePPi (11, 12). Although intracellular PPi (iPPi) is a by-product of many synthetic intracellular reactions (13), it does not diffuse across healthy biomembranes (14). To explain ePPi elaboration by chondrocytes, one must invoke de novo ePPi formation by ectoenzymes or a transport mechanism for iPPi to reach the matrix where CPPD crystals form.
The ecto-NTPPPHs hydrolyze extracellular nucleoside triphosphates into their monophosphate esters and ePPi (15, 16). These enzymes are enriched in plasma membranes (17). Several studies indicate that the bulk of ePPi involved in CPPD crystal formation in joints is derived from the hydrolysis of extracellular ATP by ecto-NTPPPH (9, 10, 15, 18, 19). Excess NTPPPH in synovial fluid correlates with the elevated synovial fluid ePPi levels reported in CPPD crystal deposition disease (8, 9). In cultured chondrocytes and articular cartilage, reduction of substrate (ATP) or ectoenzyme activity (by trypsinization) markedly reduces ePPi formation (19, 20).
Two chondrocyte cell membrane proteins may exhibit NTPPPH activity. One is cartilage intermediate-layer protein (CILP), which has recently been described as a human homolog of the putative porcine 127-kd NTPPPH (21, 22). CILP distribution is restricted to articular tissues including hyaline cartilage, fibrocartilage, tendon, ligament (23), and synovial membrane (24), which are also the only tissues that spontaneously elaborate ePPi in organ culture (11). Although CILP has not been proven to have intrinsic NTPPPH activity to date and there is a possibility that CILP may not be the enzyme itself, the original isolation of CILP was from NTPPPH-enriched fractions (25), and its expression correlates with ePPi generation by porcine chondrocytes in response to growth factors and with aging (26). CILP protein content in human OA articular cartilage increases with age (27). Increased CILP expression with aging was confirmed by immunohistochemistry and in situ hybridization studies (28), paralleling the increased prevalence of CPPD deposition disease with aging.
A definitive NTPPPH is plasma cell membrane glycoprotein 1 (PC-1), a member of the phosphodiesterase nucleotide pyrophosphatase family that includes PC-1, autotaxin, and B10 (29). PC-1 is expressed broadly, including on skin fibroblasts, osteoblasts, and chondrocytes (30). Overexpression of PC-1 increases iPPi, ePPi, and matrix vesicle PPi in several cell types (31, 32). Both PC-1 and CILP may be directly involved in chondrocyte ePPi formation or have a regulatory role in ePPi elaboration.
ANK was recently proposed to be an important factor in transport of iPPi across the cell membrane (33). The ANK protein is a multipass transmembrane protein that serves either as an anion channel or as a regulator of such a channel. ANK may control egress of iPPi across the cell membrane, or possibly egress of intracellular substrates for NTPPPH. Ho et al have reported that the ank gene mutation in mouse fibroblasts increases iPPi concentration and reduces ePPi concentration (33). Overexpression of wild-type ANK in mutant ank/ank mouse fibroblasts reversed the alterations in ePPi and iPPi levels, indicating an important role for ANK in regulating PPi trafficking. The effect of ANK was blocked by probenecid, a general inhibitor of organic anion transport previously reported to decrease ePPi elaboration by articular chondrocytes (34).
Chondrocyte ePPi elaboration is a bioregulable process, responsive to growth factors and some cytokines. Transforming growth factor β1 (TGFβ1) is the major growth factor that elevates ePPi production by normal chondrocytes (35). Aged chondrocytes are much more responsive to TGFβ1 than are chondrocytes from young animals (26, 36). Insulin-like growth factor 1 (IGF-1) is a negative modulator of ePPi elaboration by articular chondrocytes (37).
The purpose of this study was to evaluate perturbations of chondrocyte NTPPPH and ANK expression and of ePPi accumulation in chondrocytes from diseased human cartilage. First we measured ePPi elaboration into conditioned media from CPPD, OA, and normal chondrocytes. Subsequently, we studied CILP, PC-1, and ANK expression in each group, and investigated the response to growth factors in these diseased chondrocytes.
- Top of page
- MATERIALS AND METHODS
Chondrocytes from CPPD-containing cartilage elaborate more ePPi than do chondrocytes from OA cartilage without crystals, or normal chondrocytes. This is the first reported study comparing ePPi production by human chondrocytes from degenerative hyaline cartilage with and without CPPD crystals. Since CPPD crystals form extracellularly and since excess ePPi accumulates locally in synovial fluid in CPPD crystal deposition disease (4, 6), we measured ePPi levels in relation to CPPD crystal deposition. Previous studies of iPPi levels in chondrocytes (48) and skin fibroblasts (48, 49) and ePPi levels in synovial fluids (6, 50, 51) have shown higher values in samples from patients with CPPD crystal deposition than from normal controls. In addition, studies comparing CPPD and OA groups indicate that levels of iPPi in chondrocytes (48) and skin fibroblasts (48, 49) and ePPi in synovial fluids (4–6, 8, 9) were higher in patients with CPPD deposition disease than in OA patients. In some studies, the elevation of ePPi levels in synovial fluids from patients with CPPD compared with OA did not reach statistical significance (7, 10). Consistent with the latter reports, ePPi elaboration by CPPD chondrocytes was higher than that by OA chondrocytes in the present study, but the difference did not reach statistical significance.
The activity of the ePPi-generating enzyme NTPPPH is elevated in conditioned media surrounding CPPD chondrocytes. NTPPPH activity in cartilage extracts (52, 53), in synovial fluids (50, 51), and in cultures of skin fibroblasts (49) from CPPD patients is higher than that in similar preparations from normal controls. Despite the fact that our studies were performed in vitro, our findings regarding NTPPPH activity in conditioned media are consistent with those of these previous studies done mostly in vivo. In addition, studies have generally shown significantly higher NTPPPH activity in CPPD cartilage extracts (52, 53), in CPPD synovial fluids (8, 9), and in cultures of CPPD skin fibroblasts (49) compared with similar preparations from OA patients. In other studies, NTPPPH activity in cartilage extracts (17) and in synovial fluids (10, 17) from CPPD patients was elevated, but values were not significantly higher than in OA samples. In the present study, NTPPPH activity in the media from CPPD chondrocytes was higher than that from OA chondrocytes, but the difference was not significant.
Our results demonstrate that chondrocytes from diseased tissues are subject to many of the regulatory factors that affect ePPi production by normal chondrocytes. In tissue and organ cultures, ePPi accumulation is regulated by growth factors, cytokines, and other bioactive mediators. In particular, TGFβ1 enhances ePPi elaboration by porcine articular chondrocytes (35). In this study, TGFβ1 enhanced, and IGF-1 inhibited, ePPi elaboration by normal human chondrocytes, consistent with the results observed with porcine chondrocytes (26, 35). Similar responses of ePPi levels to growth factors were observed in CPPD and OA chondrocytes. These findings confirm that growth factors, especially TGFβ1, regulate ePPi levels in diseased human chondrocytes. Furthermore, TGFβ1-induced ePPi accumulation in both CPPD and OA chondrocyte cultures was higher than that with normal chondrocytes. TGFβ1 and IGF-1 are present in higher concentrations in OA and CPPD fluids than in normal fluids (54–57). Taken together, these findings indicate that higher sensitivity and higher levels of growth factors in chondrocytes affected with degenerative disease than from normal chondrocytes might cause increased generation of ePPi.
Traditionally, steady-state levels of individual RNA transcripts have been measured by Northern blot and ribonuclease protection assays. RT-PCR has revolutionized the analysis of RNA and has proven to be a very useful method for quantifying even a few molecules of mRNA in tissue samples, although it is considered a semiquantitative technique because of amplification of the RNA amounts. We assessed the quality and quantity of reverse-transcribed cDNA, using primer pairs for CILP, PC-1, ANK, and GAPDH. Our previous study of CILP, PC-1, and GAPDH expression in porcine chondrocytes showed that the results of RT-PCR correlated well with those of Northern blot analysis (26). With regard to NTPPPH expression, we confirmed that CILP mRNA expression paralleled ePPi elaboration in all subsets of chondrocytes and in response to growth factors. These findings are consistent with those of our previous study using porcine chondrocytes (26) and support the notion that CILP might promote CPPD crystal formation by enhancing ePPi production.
In contrast, PC-1 mRNA expression changed little, regardless of the source of chondrocytes or stimulation with growth factors. PC-1 is the predominant NTPPPH involved in mineralization by osteoblasts (32), and PC-1–deficient mice (ttw/ttw) develop hyperossification in early life, which produces progressive ankylosis of peripheral joints (58). Moreover, PC-1 deficiency has been identified recently in a male human infant with idiopathic infantile arterial calcification, in which calcification (hydroxyapatite deposition) in the media of large muscular arteries and smooth muscle cell proliferation occur, associated with periarticular calcification (59). These observations support the contention that PC-1 plays an important role in repressing apatite mineral formation. Increased expression of PC-1 was recently reported as a potential pathogenic factor in knee meniscal fibrocartilage matrix calcification in chondrocalcinosis (60).
Positive CILP immunostaining in both CPPD crystal aggregates and hypertrophic/metaplastic chondrocytes of meniscus from CPPD patients has also been reported (61). Moreover, in our preliminary immunohistochemistry and in situ hybridization studies using hyaline cartilage, CILP expression was strong at the periphery of CPPD crystal deposits (Yamakawa K, et al: unpublished data). These observations and the current report suggest that CILP may be a crucial factor in CPPD crystal formation not only in meniscal fibrocartilage, but also in articular hyaline cartilage.
ANK is another important factor that regulates ePPi levels. ANK is a putative transporter for egress of iPPi from cells. It is reportedly up-regulated in degenerative joint disease (62). Our results confirm that ANK mRNA expression is higher in chondrocytes and cartilage extract from CPPD and OA patients than in normal chondrocytes and cartilage extract. Moreover, chondrocytes and extract from cartilage with both OA and CPPD crystal deposition express higher levels of ANK than do similar samples with OA but no CPPD crystals. In response to growth factors, ANK mRNA expression was increased by TGFβ1, which parallels the effects of TGFβ1 on ePPi accumulation and CILP expression. These findings support the concepts that ANK participates in controlling ePPi elaboration and that up-regulated ANK mRNA expression promotes CPPD crystal formation. Recent reports on mutations of the ANK gene in British (63) and French (64) kindreds of patients with CPPD deposition disease also support the notion of a causative role for ANK in CPPD crystal deposition. No significant link was found between human CILP gene mutation and some familial forms of CPPD deposition disease (65).
Recently, Johnson et al reported that ANK and PC-1 cooperatively regulate both iPPi and ePPi (62). In our current study, ANK expression was closely associated with CILP expression. Coordinated dysregulation of ANK, PC-1, and CILP may result in the increased ePPi levels that promote CPPD crystal formation in cartilage.
There is a discrepancy between elevated ePPi levels and unaltered NTPPPH activities in response to growth factors. Unlike the case in normal chondrocytes, the NTPPPH activity in CPPD and OA chondrocytes did not increase in response to growth factors. This lack of response is consistent with stable mRNA and protein expression of PC-1. The NTPPPH activity in chondrocytes is mostly accounted for by PC-1 (>50%) (66), in contrast to CILP (∼5%) (23). In CPPD and OA cultures and extracts, CILP and ANK protein expression did not differ, nor did they respond to growth factors, although mRNA levels did change. The discrepancy between the mRNA and protein expression may be caused by a regulatory mechanism acting at a posttranscriptional level (alternative splicing, repress translation, etc.), feedback destabilization, or regional differences in the translation rate of the transcript (67, 68). It is also possible that this CILP is so stable that short-term alterations in protein synthesis may not be detectable.
In conclusion, we report that chondrocytes from CPPD patients generate more ePPi and NTPPPH enzyme activity than those from degenerative cartilage without CPPD crystals or from normal chondrocytes. CILP and ANK mRNA expression is also up-regulated in chondrocytes and cartilage extracts from patients with CPPD compared with those from OA patients or from normal cartilage. Furthermore, CILP and ANK mRNA expression and ePPi elaboration are concomitantly stimulated by TGFβ1 and inhibited by IGF-1 in chondrocytes from normal, OA, or CPPD sources. These findings suggest that increased CILP and ANK expression participate in excess accumulation of ePPi and promote CPPD crystal formation in articular cartilage.