Osteoarthritis (OA) is the most common form of joint disease that affects humans. The incidence of OA increases during every decade of life, and by the age of 65 years, almost one-third of the population has OA of the knee joints. The economic burden attributed to the joint pain and disability of OA amounts to billions of dollars each year (1). As the population demographic in the US changes to a predominantly older generation, the increasing prevalence of OA will be a major public health problem.
There is currently no treatment available that will prevent or cure OA. Pharmacologic and nonpharmacologic agents used for OA provide only symptomatic relief of pain. The lack of specific therapy for this disease is perhaps due to our limited understanding of its pathogenesis. Understanding the molecular mechanisms involved in the development of OA will help us to develop ways to prevent or reverse the degenerative process of the disease.
Current concepts of the pathogenic mechanisms of OA suggest that there is a shift in the homeostatic balance between the destruction and synthesis of bone and cartilage, with a net progressive destruction of these tissues (2). The destructive process is mediated by the production of aggrecanases and matrix metalloproteases that selectively degrade cartilage matrix (3–5). In addition to the role of proteases in matrix destruction, recent studies of the biology of chondrocytes show that these cells actively produce reactive oxygen species (ROS), including superoxide anions, hydrogen peroxide, hydroxyl radicals, and nitric oxide (6–13). ROS are capable of inducing apoptotic cell death in chondrocytes, but more important, they can result in the degradation of aggrecan and collagen (14–20). Degradation of aggrecan and collagen by ROS has been studied in some detail (15–20). However, most of these studies have been performed in vitro using purified matrix components and ROS. The in vivo role of ROS in the matrix degradation of cartilage during aging and OA remains to be determined. Immunochemistry studies show the presence of 3-nitrotyrosine residues in human articular cartilage that correlate with aging and OA, which suggests that oxidative damage from reactive nitrite radicals may contribute to the development of age-related OA (21).
In other age-related degenerative diseases, such as those of the brain and cardiovascular systems, lipid peroxidation has been implicated as the key source of oxidative stress (22). It is possible that lipid radicals could also be a source of oxidative stress in age-related OA. Lipid peroxidation generates a variety of hydroperoxide and aldehyde products that are highly reactive with components of the cell and the extracellular matrix (23, 24). Lipid radicals also function as intracellular signaling molecules, influencing various cellular functions (25, 26). There is a paucity of knowledge about the role of lipid-induced oxidative stress in cartilage aging and OA.
We have shown that chondrocyte-derived ROS mediate aggrecan degradation and that various antioxidants prevent this degradation process (27). In ensuing studies, we investigated the role of chondrocyte-dependent ROS-mediated collagen degradation (28). For those investigations, we developed an in vitro model of chondrocyte-dependent collagen degradation. Using this model, we showed that chondrocyte-derived lipid peroxidation mediates collagen degradation and that vitamin E inhibits this degradation process (28). More recently, we demonstrated that malondialdehyde (MDA), a toxic aldehydic end product of lipid peroxidation, mediates the oxidation of cartilage collagen (29). MDA and hydroxynonenal (HNE) are specific and major aldehydic products of lipid peroxidation that are believed to be largely responsible for the cytopathologic effects observed during the oxidative stress of lipid peroxidation (23, 24). Taken together, these observations suggest a role of lipid peroxidation in cartilage aging and OA.
In the present study, we investigated the role of lipoxidation in collagen degradation by human articular chondrocytes. We also investigated the presence of in vivo molecular imprints of lipid peroxidation in human OA cartilage. Our findings are presented herein.
- Top of page
- MATERIALS AND METHODS
In the present study, we developed a human OA chondrocyte-dependent model of collagen degradation, similar to one we previously described in which we used normal lapine chondrocytes (28). The human chondrocyte model showed exquisite sensitivity. The release of 3H-labeled material into the medium corresponded to the potential oxidative damaging effects of hydrogen peroxide. Of the various agonists tested to initiate a respiratory burst in OA chondrocytes, only calcium ionophore and LPS rapidly and in a dose-dependent manner induced the release of 3H-proline–labeled material into the medium. The release of labeled matrix was not a result of cell death, but demonstrated characteristics of a cell-dependent oxidative degradative process in the extracellular matrix, such as the rapid cause-and-effect relationship of matrix catabolism. Physiologic concentrations of vitamin E inhibited collagen degradation (data not shown). These observations confirm that this model can be a useful tool with which to study the mechanisms of collagen degradation in OA.
To demonstrate the role of chondrocyte lipid peroxidation, we directly measured lipid peroxidation activity in cis-parinaric acid–treated primary OA chondrocytes. These studies showed constitutive and inducible lipid peroxidation activity in OA chondrocytes. However, 4 of the 7 samples tested showed maximal levels of lipid peroxidation activity, which was not enhanced by treatment with LPS or calcium ionophore, whereas the remaining samples showed a minimal response to the activating agent. Thus, lipid peroxidation activity in OA chondrocytes was maximally elevated in more than 50% of the OA chondrocyte samples tested, to the extent that further stimulation with agonist was not possible.
Why do OA chondrocytes have such high levels of lipid peroxidation activity? It is possible that the enhanced lipid peroxidation activity in OA chondrocytes is the result of an altered lipid profile. Bonner et al (33) first documented that lipids, especially polyunsaturated fatty acids (PUFAs), accumulate with aging of normal cartilage. Adkisson et al (34) showed that normal cartilage expresses low levels of n-6 PUFAs and high levels of n-9 fatty acids. High levels of n-9 PUFAs found in young cartilage are progressively depleted with increasing age, and this depletion of n-9 PUFAs is accompanied by a steady increase in the levels of n-6 PUFAs. Similar changes are particularly pronounced in OA cartilage (34). In several models of degenerative arthritis, lipid accumulation generally precedes local tissue degeneration (35, 36). Silberberg and colleagues (37, 38) demonstrated an increase in the incidence of age-dependent OA in the C57 inbred strain of mice fed a diet high in saturated fatty acids. Dietary lipids also modify the fatty acid composition of cartilage and isolated chondrocytes (39).
The tissues from degenerated joints exhibit an acceleration of metabolism, an effect that can be reproduced in vitro with normal chondrocytes supplemented with exogenous essential fatty acids (40, 41). Lippiello et al (42) observed that the levels of fatty acids were markedly enhanced in OA cartilage in association with an increasing degree of OA lesions. Certain rare forms of osteochondrodysplasia associated with precocious degenerative joint disease are characterized by the accumulation of lipids in chondrocytes (43, 44). Taken together, the evidence suggests that the distribution and content of lipids in cartilage change during aging and OA. Lipids are essential for maintaining the integrity of the cell membrane. Lipid peroxidizability was shown to increase with the PUFA content in model membranes and cells (45). There is a strong body of evidence that supports a tissue-damaging role of lipid peroxidation (22).
Measurement of lipid peroxidation activity provides an assessment of only lipid-derived sources of oxidative stress resulting from the oxidation of lipids in cell membranes (22). To broaden this investigative approach, we used a DCF oxidation method to measure the overall oxidation activity in OA chondrocytes. DCF oxidation primarily measures the intracellular production of reactive oxygen intermediates, such as hydrogen peroxide, peroxidases, ROS other than hydrogen peroxide, and reactive nitrogen species (46–48). We observed DCF oxidative activity in all OA chondrocyte samples examined. In contrast to lipid peroxidation activity, all OA specimens showed spontaneous and inducible DCF oxidation activity. These observations indicate that lipid peroxidation activity may be selectively pushed to near-maximal levels in OA chondrocytes, whereas the overall oxidative activity in OA chondrocytes may show an activation-dependent response pattern. It is possible that the oxidative activity measured by DCF in OA chondrocytes may be abnormal, but we were unable to perform this measurement in normal articular chondrocytes for a comparison because of insufficient tissue.
Because MDA and HNE, which are specific and major toxic products of lipid peroxidation, are responsible for tissue damage, their identification in tissues provides specific molecular markers of in vivo tissue damage mediated by lipid free radicals (22–24). Aldehydic products of lipid peroxidation liberated outside the cells oxidizes the extracellular matrix components, resulting in fragmentation, modification, aggregation, and protein conformation changes and eventually leads to alterations in tissue functioning. Using immunohistochemical methods, we identified MDA and HNE protein adducts in OA tissue sections. The binding of monoclonal reagents to tissue was immunospecific, indicating that these aldehydic adducts were indeed present in OA tissue and were derived in vivo. Furthermore, aldehydic adduct activity was detected at the sites of OA lesions, such as around cysts and erosions, and at fissure sites. Aldehydic adduct immunoreactivity was also present at intra- and intercellular matrix sites, indicating evidence of lipoxidation damage in matrix tissue. In addition, the immunoreactivity was present around individual chondrocytes and chondrocyte clusters, suggesting that chondrocytes may be the source of lipid peroxidation in cartilage. In contrast, normal cartilage sections showed nonspecific reactivity only on the cartilage surface. Normal cartilage chondrocytes showed little immunoreactivity. Taken together, these observations indicate that the MDA and HNE adducts were distributed in a lesion-specific pattern in OA cartilage, suggesting their pathogenic role.
Using specific immunoreagents that detect intact, MMP-specific cleavage or denatured collagen, Hollander et al (49) observed a progressive gradient of cartilage damage from the top to the deep zone. In the present study, no such gradient for aldehydic adduct immunoreactivity was observed. The aldehydic adduct reactivity was detected in all 3 zones (superficial, middle, and deep). The progressive gradient of damage from the top to the deep zone of cartilage was interpreted as indicating that the pathology of OA is initiated at the top of the cartilage and gradually progresses to the deep zone (49). Our observations suggest that lipoxidation-induced damage is initiated throughout the entire cartilage thickness.
The unique distribution of lipids in cartilage, which changes significantly with age, the influence of the dietary intake of lipids, and the consequent lipid peroxidizability suggest a role of lipoxidation in the pathology of OA. The presence of aldehydic adduct markers, which are specific for lipid peroxidation, in OA cartilage, as observed in the present study, provides direct molecular evidence that lipoxidation plays a role in the pathogenesis of OA. Enhanced levels of lipid peroxidation activity in OA chondrocytes also indicate a role of lipid peroxidation in OA. The chondrocyte-dependent oxidative damage to cartilage collagen shown in our in vitro degradation model also supports a role of lipoxidation. In the Framingham Knee Osteoarthritis Cohort study (50), the population with a medium to higher intake of vitamin C, β-carotene, and vitamin E had a reduced risk of progression of knee OA, as assessed radiographically. Antioxidants have not been rigorously tested in the prevention or treatment of OA. The results of our studies (refs. 27–29 and the present study) as well as those of other studies on the role of nitric oxide (21) and redox status (14) in chondrocyte aging indicate that antioxidant therapies may prove useful in the prevention and/or treatment of cartilage aging and OA.