To determine whether oxidative damage to cartilage proteins can be detected in aging and osteoarthritic (OA) cartilage, and to correlate the results with the local production of interleukin-1β (IL-1β) and the responsiveness of isolated chondrocytes to stimulation with insulin-like growth factor 1 (IGF-1).
The presence of nitrotyrosine was used as a measure of oxidative damage. Histologic sections of knee articular cartilage, obtained from young adult and old adult cynomolgus monkeys, which develop age-related, naturally occurring OA, were evaluated. Each cartilage section was graded histologically on a scale of 0–7 for the presence of OA-like changes, and serial sections were immunostained using antibodies to nitrotyrosine and IL-1β. Chondrocytes isolated and cultured from cartilage adjacent to the sections used for immunostaining were tested for their response to IGF-1 stimulation by measuring sulfate incorporation in alginate cultures. For comparison with the monkey tissues, cartilage sections from human tissue donors and from tissue removed at the time of OA-related joint replacement surgery were also immunostained for nitrotyrosine and IL-1β.
The presence of nitrotyrosine was associated with aging and with the development of OA in cartilage samples from both monkeys and humans. All sections that were highly positive for IL-1β also showed staining for nitrotyrosine. However, in a few sections from older adult monkeys and humans, nitrotyrosine was present but IL-1β was absent, suggesting that some age-related oxidative damage is independent of IL-1β. In chondrocytes that were isolated from monkey cartilage positive for nitrotyrosine or IL-1β, the response to stimulation with IGF-1 was significantly reduced. In some samples from older adult monkeys, IGF-1 resistance was seen in cells isolated from tissue that did not stain for nitrotyrosine or IL-1β.
Oxidative damage due to the concomitant overproduction of nitric oxide and other reactive oxygen species is present in both aging and OA cartilage. This damage can contribute to the resistance of chondrocytes to IGF-1 stimulation, but it is unlikely to be the sole cause of IGF-1 resistance in these chondrocytes.
Osteoarthritis (OA) is a multifactorial disease that increases in prevalence with age in humans (for review, see ref. 1) as well as in other animal species, including non-human primates (2). Several aging-related changes that could make tissue more susceptible to development of OA have been observed in articular cartilage, including a reduced response of the chondrocytes to stimulation by growth factors (1). We recently showed that knee articular chondrocytes isolated from young and old adult cynomolgus monkeys have an age-related decline in their anabolic response to insulin-like growth factor 1 (IGF-1) (3). Chondrocytes from monkey cartilage exhibiting histologic OA-like changes also have a reduced response to IGF-1, as was previously demonstrated in human chondrocytes from cartilage removed at the time of joint replacement surgery for OA (4). Further understanding of the mechanisms responsible for a reduced growth factor response in the setting of aging and OA would be important for the development of novel interventions to improve cartilage homeostasis.
Focal articular cartilage damage, which is likely to be driven by biomechanical factors, occurs during development of OA because of an imbalance between anabolic and catabolic activity. Resistance of chondrocytes to stimulation by growth factors such as IGF-1 could play an important role in the pathogenesis of OA by contributing to this imbalance. Increased production of cytokines, including interleukin-1 (IL-1), by chondrocytes also appears to contribute to this imbalance in an autocrine/paracrine manner (5, 6). IL-1 has been shown both to inhibit chondrocyte anabolic activity, including proteoglycan synthesis (7), and to stimulate catabolic activity, including production of metalloproteinases (8). IL-1 has also been shown to stimulate chondrocyte expression of inducible nitric oxide synthase (iNOS), which results in increased production of nitric oxide (NO) (9, 10). Importantly, a previous study in a mouse model of arthritis, using wild-type and iNOS knockout animals, suggested that overproduction of NO can contribute to IGF-1 resistance in chondrocytes (11). Likewise, IGF-1 resistance was noted in chondrocyte cell culture studies when cells were exposed to exogenous NO or were induced by gene transfer to overexpress iNOS (12). Thus, IL-1–driven NO production could contribute to chondrocyte resistance to IGF-1.
Nitric oxide is a reactive species with a very short half-life, usually measured in seconds (13). Therefore, NO cannot be directly and accurately measured in cartilage samples in order to determine whether it is overproduced in aging or OA tissue. Reaction or degradation products of NO have been used as a measure of NO production in tissues and cultured cells. NO can react with reactive oxygen species (ROS) such as superoxide anion. When NO reacts with superoxide, peroxynitrite is formed, which, in turn, can react with tyrosine residues in nearby proteins to produce 3-nitrotyrosine (14, 15). The stable 3-nitrotyrosine in cell and matrix proteins, which likely remains for the life of the protein, can therefore serve as a marker for reactions involving NO and superoxide in the tissue (14).
Given the potentially important role of chronic damage to cell and matrix proteins caused by NO and oxygen radicals in cartilage, and given the possibility that this damage could impair the cellular response to growth factors including IGF-1, the objective of the present study was to determine whether nitrotyrosine is present in normal cartilage with aging, in OA cartilage, or in both. In addition, we examined the relationship between the presence of nitrotyrosine in tissue and the ability of chondrocytes isolated from the same tissue to respond in vitro to IGF-1. Because IL-1 appears to play an important role in OA and has been shown to increase NO production, the tissue samples were also evaluated for the presence of IL-1β.
MATERIALS AND METHODS
Monkey and human cartilage samples.
Articular cartilage from the knee joints of 21 cynomolgus macaques (Macaca fascicularis), ranging in age from 7 years to 27 years (mean 15.7 years), was obtained at the time of necropsy. The details of the source of the animals and the methods used for cartilage acquisition and histologic grading have been previously published (3). Briefly, cartilage for use in cell culture was removed separately from the medial and lateral tibial plateaus and femoral condyles; a central, intact horizontal strip of cartilage was left at each of these sites until processing for the histologic studies. Toluidine blue–stained sections were scored on a scale from 0 (normal) to 7 (severe damage) by a single investigator (CSC), who was blinded to the cell culture results.
Human knee and ankle articular cartilage samples were obtained within 24 hours of death from tissue donors who had no known history of joint disease; these samples were provided by the Regional Organ Bank of Illinois. Each joint was evaluated macroscopically using a modified Collins scale of 0–5, as described previously (16). Human OA cartilage samples were obtained from the knee joints of patients with OA, at the time of joint replacement surgery. All cartilage samples were obtained in accordance with institutional protocol, with review board approval.
For the monkey tissues, serial sections of paraffin-embedded tibia and femur, from the same blocks as those from which the routine histologic sections were prepared, were cut and immunostained using antibodies directed against IL-1β and nitrotyrosine. The same procedures were followed for immunostaining human tissues. Briefly, sections were adsorbed to poly-L-lysine–coated and oven-dried glass slides, deparaffinized, and hydrated. The slides were then stained using horseradish peroxidase (HRP) or alkaline phosphatase (AP) methods.
For the HRP method, the slides were blocked with 3% H2O2. After blocking of nonspecific binding sites with normal goat serum (diluted 1:10 in Tris buffered saline [TBS]), the sections were incubated with polyclonal rabbit antiserum to either IL-1β (1:50 dilution; Endogen, Woburn, MA) or nitrotyrosine (1:2,000 dilution; Chemicon, Temecula, CA) for 1 hour at room temperature, followed by incubation with biotinylated goat anti-rabbit IgG (LSAB2 kit; Dako, Carpinteria, CA) for 30 minutes at room temperature. After washing with TBS, the sections were incubated with streptavidin–HRP complex (LSAB2 kit; Dako) for 30 minutes at room temperature. Sections were developed with diaminobenzidine (Dako) and counterstained with Mayer's hematoxylin (Sigma, St. Louis, MO). Nitrotyrosine antibody specificity was evaluated by incubating selected sections in the presence of excess exogenous nitrotyrosine before the immunostaining procedure. This approach resulted in substantially reduced immunostaining or, in most samples, eliminated immunostaining.
For the AP procedure, Tris buffer containing 0.5% casein was used for the washes. The sections were incubated with the primary antibody in Tris buffer at 4°C overnight and were then incubated with biotinylated goat anti-rabbit IgG (1:20 dilution; Super Sensitive kit; BioGenex, San Ramon, CA) for 30 minutes at 30–35°C. The sections were labeled with streptavidin–AP for 30 minutes at 30–35°C (1:20 dilution; Super Sensitive kit; BioGenex), developed with Vector Red I (Vector, Burlingame, CA) and counterstained with Mayer's hematoxylin (Sigma).
Immunostaining for IL-1β in the monkey tissues was graded by an observer who was blinded to the other results of the study (CSC). The immunostaining was confined to chondrocytes. Therefore, after counting equal numbers of cells in each section, each tibial plateau and femoral condyle was graded on a scale of 0–11 as follows: 0 (no or slight staining of ≤5 cells), 1 (6–25 positive cells), 2 (26–75 positive cells), 3 (76–125 positive cells), 4 (126–175 positive cells), 5 (176–225 positive cells), 6 (226–275 positive cells), 7 (276–325 positive cells), 8 (326–375 positive cells), 9 (376–425 positive cells), 10 (426–475 positive cells), and 11 (476–525 positive cells).
Similarly, immunostaining for nitrotyrosine in the monkey cartilage was evaluated by an observer who was blinded to the other results of the study (CSC). Because these sections had positive staining in both chondrocytes and cartilage matrix, they were assessed as being either positive (positive staining in chondrocytes and/or cartilage matrix) or negative (no staining). Each site (medial or lateral tibial plateau, medial or lateral femoral condyle) was evaluated separately. For the studies on human tissues, only randomly selected samples of cartilage, rather than the entire joint site, were available for immunostaining, which was scored as either positive or negative.
Monkey chondrocyte cell culture and IGF-1 stimulation.
Monkey articular chondrocytes were isolated by enzymatic digestion and cultured in alginate beads as previously described in detail (3). The alginate beads were placed in 24-well plates (4 beads per well) in 0.5 ml of serum-free Dulbecco's modified Eagle's medium/Ham's F-12 supplemented with “mini”–insulin–transferrin–selenium (3). After a 3-day recovery period, cells were treated with either recombinant human IGF-1 (R&D Systems, Minneapolis, MN) at 100 ng/ml or vehicle (control) for 24 hours in the presence of 50 μCi/ml of 35SO4. Some of the cultures (n = 9) were also treated with 100 ng/ml des(1–3) IGF-1 (GroPep, Adelaide, Australia) for comparison with IGF-1. After incubation, the media were removed, and the alginate beads were dissolved in sodium citrate, followed by centrifugation to separate the cell pellet from the alginate–matrix fraction. The amount of radiolabeled sulfate incorporated into alcian blue–precipitable material and the amount of DNA in the cell pellets were measured as previously described in detail (3). IGF-1–stimulated sulfate incorporation was calculated as the counts per minute/μg DNA to normalize for cell numbers, and results were expressed as a percentage of the unstimulated controls.
The results were analyzed using Statview software (SAS, Cary, NC). The relationship with positive or negative immunostaining for nitrotyrosine in monkey cartilage was analyzed for age, OA score, IL-1β score, and response to IGF-1 using the Mann-Whitney test. Spearman's rank correlation was used to analyze the relationship between the IL-1β score and response to IGF-1.
Presence of nitrotyrosine in monkey and human articular cartilage.
Nitrotyrosine, detected by immunostaining, was present in samples of articular cartilage from older adult cynomolgus monkeys and older adult humans (Figure 1). In general, nitrotyrosine was present both within cells and in the cartilage matrix and was seen mainly in the more superficial regions, although some sections showed staining in deep-zone chondrocytes as well (Figure 1B). Immunostaining was most apparent in sections of cartilage that showed histologic changes consistent with OA. All 6 human cartilage sections obtained at the time of joint replacement surgery for knee OA were positive for nitrotyrosine. Knee (n = 8) or ankle (n = 17) cartilage samples from tissue donors without a known history of joint problems, and which appeared macroscopically normal, showed age-related immunostaining for nitrotyrosine. Samples positive for nitrotyrosine were from donors who were significantly older than donors whose cartilage samples stained negative (mean ± SEM age 66 ± 2 years and 43 ± 7 years, respectively; P = 0.01) (Figure 2D). There was no apparent difference in nitrotyrosine staining between knee and ankle cartilage.
We previously demonstrated that in monkeys, as in humans, the medial tibial plateau is the site most commonly affected by histologic OA-like changes, and that the lateral femoral condyle is least affected (2, 17). The histologic OA scores were significantly higher in cartilage sections that were positive for nitrotyrosine (P = 0.01) (Figure 2A). There was also a relationship between age and nitrotyrosine formation, with a significantly greater monkey age associated with the nitrotyrosine-positive samples (P = 0.002) (Figure 2B). Because aging and development of OA are related, the presence of nitrotyrosine was also evaluated in sections of monkey cartilage that were histologically normal or showed only early aging-related degenerative changes (histologic OA score <4). In this set of samples, those positive for nitrotyrosine were associated with a significantly greater mean age (P = 0.01) (Figure 2C), indicating that the appearance of nitrotyrosine with aging can occur without overt histologic changes of OA.
IL-1β immunostaining and correlation with nitrotyrosine.
Because autocrine or paracrine production of IL-1β can mediate increased NO production by chondrocytes, serial sections of monkey cartilage from animals with a range of ages and OA scores were also immunostained for IL-1β. Unlike immunostaining for nitrotyrosine, immunostaining for IL-1β was mainly restricted to cells (Figure 3A); therefore, staining was scored based on the number of positive cells in each section. The IL-1β score was significantly higher in the group of samples with histologic OA scores ≥4 (P = 0.03) (Figure 3B) and in the nitrotyrosine-positive samples (P = 0.016) (Figure 3C). Among cartilage sections from young adult monkeys, all that were nitrotyrosine-positive were also highly positive for IL-1β. Among cartilage sections from old adult monkeys, some samples were nitrotyrosine-positive but IL-1β–negative (Figure 3D). The latter sections were most commonly obtained from tissue without histologic evidence of OA, because it was very unusual to find IL-1–positive cells in normal-appearing tissue.
Similar results for IL-1β staining were noted in human cartilage samples (8 samples from tissue donors and 4 OA samples). All 4 OA samples were positive for both IL-1β and nitrotyrosine. Only 1 cartilage section from a normal tissue donor was positive for IL-1β, and this section was obtained from a joint scored as Collins grade 2 (deep fibrillation and fissuring and possible osteophyte formation). The other joints, which were IL-1–negative, were scored as grade 0 or 1.
IGF-1 response of chondrocytes isolated from cartilage with immunostaining for nitrotyrosine and IL-1β.
In a previous study, we showed that as monkey age increased, the anabolic response of isolated chondrocytes to IGF-1 declined (3). As already discussed, there is evidence that increased production of NO can reduce the ability of chondrocytes to respond to IGF-1 (11, 12). Because many of the monkey knee joints that were immunostained for nitrotyrosine and IL-1β in the present study were the same as those from which we had isolated chondrocytes for the study of aging and IGF-1 response (3), we had the opportunity to determine whether there was a relationship between the IGF-1 response (measured by sulfate incorporation) and the presence of nitrotyrosine or IL-1β. In fact, there was a highly significant difference in IGF-1–stimulated sulfate incorporation (P = 0.002) between cells isolated from joints in which there was positive cartilage immunostaining for nitrotyrosine and those in which there was not (Figure 4A). Chondrocytes from nitrotyrosine-positive cartilage did not show a response above that of unstimulated controls, while cells from nitrotyrosine-negative cartilage had an average sulfate incorporation that was 130% of control.
When the data were separated based on chondrocytes from young adult monkeys (7–14 years) and those from old adult monkeys (>14 years), it was clear that cells from nitrotyrosine-positive cartilage from either age group lacked a sulfate incorporation response to IGF-1 above that of control (Figure 4B). Most chondrocyte cultures from young monkeys that came from nitrotyrosine-negative tissue responded to IGF-1, while cells from old monkeys did not, suggesting that factors in addition to nitrotyrosine formation may also contribute to the age-related decline in responsiveness to IGF-1. The presence of IL-1β also was associated with a reduced chondrocyte response to IGF-1, with a significant negative correlation between IL-1β score and IGF-1 response (r = −0.61; P = 0.001) (Figure 4C). The lack of response to IGF-1 in the nitrotyrosine-positive tissue did not appear to be due to excess production of potentially inhibitory IGF binding proteins (IGFBPs), because these cells also did not respond to des(1–3) IGF-1, which has substantially lower affinity for IGFBPs (Figure 5).
These results support a role for oxidative damage to cartilage in the setting of both aging and development of OA. Oxidative stress resulting in the accumulation of proteins that have undergone oxidative damage has been a long-standing theory for the cause of aging-related changes in a number of tissues (for review, see ref. 18). Because of lack of study, however, little evidence was available suggesting that such a process occurs in cartilage. Chondrocytes have been shown to be capable of producing ROS (19) and NO (9) when stimulated in vitro, but because of the highly reactive nature of these compounds and their short half-lives, it has been much more difficult to demonstrate their presence in vivo. Recently, the formation of nitrotyrosine has been used as evidence for oxidative and nitrative damage in aging muscle (20) and aging brain (21), in patients and animals with lung injury (22), and in the setting of chronic neurodegenerative disorders (23). The present results, demonstrating nitrotyrosine in articular cartilage obtained from older adult monkeys and older adult humans, as well as in monkey and human OA cartilage, suggest that local accumulation of proteins altered by the reaction between ROS and NO may be important in the pathogenesis of OA as well.
The presence of nitrotyrosine in OA cartilage is consistent with a large body of data showing that overproduction of NO is important in the pathogenesis of OA (for review, see ref. 10). Inhibition of NO production in animal models of OA has been found to attenuate development of OA lesions (24). Despite extensive evidence that NO is involved in OA, the exact mechanisms by which NO acts are not clear. Furthermore, it is not clear whether NO acts alone or whether some of the affects of NO in OA occur when NO reacts with other ROS such as superoxide (25). Also not known are the driving forces responsible for the increased production of NO in OA cartilage, although IL-1 is a prime candidate.
As shown here and in other studies (for review, see ref. 26), IL-1 clearly is increased in OA cartilage. In the present study, we noted an association between nitrotyrosine formation and the presence of IL-1β. IL-1β can increase production of both NO and ROS, providing a potential source for the generation of peroxynitrite (9, 27). The generation of peroxynitrite from the reaction of NO with superoxide has been hypothesized to be responsible for at least a portion of the tissue-damaging effects of NO (28, 29). The detection of nitrotyrosine in monkey cartilage and human OA cartilage, as demonstrated in the present study, and in rabbit cartilage in an anterior cruciate ligament transection model (30), provides strong but indirect evidence that peroxynitrite is formed in aging and OA cartilage. Additional evidence for peroxynitrite generation in human OA is provided by the recent description of nitrotyrosine in chondro-osteophytes from joint tissues removed at the time of hip replacement surgery for OA (31).
Alternative, albeit less likely, causes of nitrotyrosine formation include the reaction of tyrosine with nitrogen dioxide that occurs when NO reacts with hydrogen peroxide in the presence of heme peroxidases (14). Whatever the exact mechanism is that results in generation of nitrotyrosine residues in cartilage, it likely involves the reaction of NO with other ROS.
Not only does nitrotyrosine formation serve as a marker of oxidative and nitrative damage, but, because many cell-signaling proteins use tyrosine phosphorylation for activation, formation of nitrotyrosine residues can directly interfere with cell function (29). Although our studies do not directly demonstrate the consequences of nitrotyrosine formation, the association between nitrotyrosine and the lack of a chondrocyte response to IGF-1 and des(1–3) IGF-1 suggests at least one potential negative effect on chondrocyte function.
Consistent with our findings of a potential link between nitrotyrosine and IGF-1 resistance, Studer et al (12) showed that excess NO could reduce the chondrocyte response to IGF-1 by reducing tyrosine phosphorylation of the IGF-1 receptor. When cultured chondrocytes were incubated with peroxynitrite, we noted nitrotyrosine formation in proteins that were in the molecular weight range of the IGF-1 receptor subunits. Although we attempted to immunoprecipitate the IGF-1 receptor and then immunoblot for nitrotyrosine, we could not demonstrate a clearly defined IGF-1 receptor band (data not shown). The inability to detect the IGF-1 receptor after peroxynitrite treatment could have resulted if nitrotyrosine formation interferes with the ability of the IGF receptor antibody to bind to the receptor protein.
Nitrotyrosine and IL-1β were present in almost all of the OA cartilage samples obtained from humans with clinically diagnosed OA and from monkeys with histologic OA. Nitrotyrosine was also present in some samples from older adult monkeys and older adult humans without evidence of overt OA changes. When the response of monkey chondrocytes to IGF-1 was evaluated by age and the presence or absence of nitrotyrosine, it was apparent that some samples from old monkeys that were negative for nitrotyrosine did not respond to IGF-1. This result indicates that mechanisms related to IGF-1 resistance independent of the effects of nitrotyrosine might also be present with age. However, the possibility of sampling error cannot be ruled out.
Although the tissue sections were obtained from sites in very close proximity to those from which cartilage was taken for cell culture, the exact same cells could not be used for immunodetection of nitrotyrosine and for measurement of the IGF-1 response in cell cultures. The relationship between nitrotyrosine formation in the tissue and the lack of IGF-1 response in cell culture was very significant in samples from young monkeys, in which only chondrocytes isolated from nitrotyrosine-positive tissue had a poor response to IGF-1. This is likely related to the finding that in young monkeys the nitrotyrosine-positive tissues also showed histologic evidence of OA, and there is a relationship between OA and loss of IGF-1 response. Thus, histologic OA is strongly associated with both nitrotyrosine formation and a reduced response to IGF-1.
In summary, the demonstration of nitrotyrosine in normal cartilage obtained from older humans and non-human primates and in OA cartilage from both species suggests that exposure to the combined effects of ROS and NO occurs in cartilage. A link to the presence of IL-1β and the known ability of IL-1 to stimulate NO and ROS production indicates that IL-1 may be at least one causative factor. Evidence, albeit indirect, of an association between nitrotyrosine and a lack of response to IGF-1 suggests that one potential consequence of cell damage from nitrotyrosine may be reduced IGF-1 signaling. Clearly, further study is needed to better understand how oxidative damage in cartilage affects chondrocyte function, resulting in changes in cartilage homeostasis that are relevant to aging and the development of OA. Such an investigation may help in the design of interventions aimed at reducing oxidative damage in cartilage.
The authors would like to thank Jan Shivers and Anne Undersander for assisting with immunohistochemistry, Barbara Patten for assistance with grading monkey tissue sections, the Regional Organ Bank of Illinois for providing human-donor tissues, and the Department of Orthopedic Surgery at Rush–Presbyterian–St. Luke's Medical Center for providing OA tissues.